Instinct Tops $1B Volume, Challenges Ad-Based Apps With AI Agents

Instinct交易额破10亿美元，以AI智能体挑战广告模式应用 随着自主智能体消除浏览摩擦并将各类应用整合为统一界面，消费级软件的变现模式将从争夺广告驱动的屏幕使用时间，转向收取交易抽成。 Noah Shinn深入解析了Instinct背后的架构与运营理念。作为一款自主个人助理，Instinct无需独立App，而是通过短信、语音和电子邮件交互来执行现实任务。讨论详细介绍了Instinct如何处理超过10亿美元的年化商品交易总额、用分发抽成取代注意力驱动的广告模式，并实现智能体之间的直接协同以处理社交与日常事务调度。Shinn还探讨了主动后台计算对基础设施的严苛要求、独立的安全性看门狗架构，以及他所持的观点——环境音频与动态生成的微应用终将取代碎片化的软件生态。 个人AI智能体初创公司Instinct的年化交易总额正超过10亿美元，在仅限邀请的机制下保持10%的日环比增长，主要由高意向的旅游预订驱动。 (17:23) Instinct以100亿美元估值从Sequoia、Benchmark和Coatue筹集了约10亿美元，用于资助庞大的算力需求并保持面向消费者的免费模式，而非陷入订阅制陷阱。 (81:38) 与等待人类提示词的被动响应式聊天机器人不同，Instinct利用自主休眠与唤醒周期来执行后台工作流，虽然需要高出数个数量级的计算Token，但通过异步自定义批处理获得了3到8倍的效率提升。 (62:30) 创始人Noah Shinn摒弃了广告变现与刷卡手续费竞争，转而瞄准分发抽成，因为无摩擦的智能体协同正推动互联网经济学脱离依赖屏幕时间的广告展示模式。 (26:08)

Noah Shinn argues that personal AI agents will replace conventional application dashboards by operating as proactive, socially aware assistants embedded in everyday communication channels. Rather than monetizing user attention through ads or upfront subscriptions, Instinct aims to build a trusted, multi-agent coordination layer sustained by merchant transaction take rates. Ambient Architecture and Proactive Execution (0:49) Instinct intentionally avoids dedicated smartphone applications, instead providing each user's assistant with a unique phone number, email address, and virtual computer environment. Users interact primarily via voice notes, text messages, and existing channels like WhatsApp or iMessage. When high visual fidelity is required for tasks such as complex trip itineraries or event logistics, the agent dynamically generates standalone web applications on the fly. Unlike traditional reactive software that waits for prompts, Instinct operates on independent sleep and wake cycles throughout the day. It anticipates daily obligations—such as scanning incoming materials at 6:00 AM, proactively calling a user to flag an urgent legal document five minutes before a deadline, or lining up ground transportation based on calendar events. Peer-to-Peer Coordination and Permissioned Trust (7:15) To eliminate the friction of scheduling and logistics, Instinct utilizes a peer-to-peer network enabling autonomous agents to communicate directly. Users define high-level intent, after which their respective agents negotiate calendar constraints and lock in appointments without human back-and-forth. This model extends to emergent group planning, such as cross-referencing friends' music preferences on Spotify, selecting concert outings, and routing shared rides. Inter-agent coordination is governed by what Shinn terms a 'trusted person network' featuring weighted, permissioned access tiers. While spouses can configure full data transparency, professional acquaintances may be restricted to calendar availability alone. If a counterpart's agent queries data outside its permitted tier, the system flags the boundary breach immediately. In commerce, Shinn claims this peer architecture will replace first-come, first-served booking snipers with bilateral value matching, pairing high-value celebrations directly with premier restaurant reservation systems. Transaction-Driven Monetization Over Ad Models (32:32) Shinn firmly rejects advertising, sponsored placements, and recurring consumer subscription fees. He contends that advertising creates adversarial incentives that compromise user safety once agents possess superior social intelligence, while charging a $100 monthly subscription represents an early 'local optimum' that limits broad consumer adoption. Supported by a reported $1 billion funding round at a $10 billion valuation led by Sequoia Capital, Benchmark, and Coatue, Instinct plans to remain free to end users. Monetization relies instead on distribution take rates paid by merchants on completed commerce. Shinn reports that Instinct is already pacing at more than $1 billion in annualized gross transaction volume—with travel bookings representing 50%—while compounding at roughly 10% day over day. While acknowledging pushback from incumbents whose ad-driven interfaces are bypassed by headless agents, Shinn argues that reducing booking friction to zero will expand total transaction volumes enough to offset lost ad revenues. Decoupled Safety Filters and Cost Discipline (23:25) Because effective personal assistance requires high-context access to emails, financial accounts, and passwords, Instinct treats security as an operational foundation rather than an afterthought. To prevent hallucinations and prompt injection vulnerabilities from executing real-world harm, the architecture decouples core intelligence from execution oversight. A separate 'watchdog' supervision layer evaluates the model's reasoning chain and proposed tool calls, retaining the authority to pause, inspect, or abort anomalous transactions before execution. According to Shinn, user trust develops empirically over weeks of low-stakes reliability rather than upfront onboarding prompts. Approximately 40% of active users connect a credit card within three weeks; once crossed, retention reaches roughly 80%. To control inference costs, Instinct routes proactive, non-urgent background routines through custom batch pipelines, enabling the company to deliver frontier-grade performance at a fraction of standard cloud endpoint pricing. Compute Runway and Incumbent Pushback (56:00) Managing compute procurement represents Shinn's primary operational focus, consuming roughly 40% of his time. With Instinct's user base compounding at approximately 10% daily and hardware lead times stretching three to four months, forecasting becomes precarious: purchasing double the current capacity lasts only a week, while 10x expansions introduce steep capital exposure and spot-market surcharges. Shinn projects that proactive, continuously running agents will generate far greater token volumes than legacy request-response tools. Simultaneously, the platform must navigate the looming corporate agent conflicts as incumbents protect ad-heavy interfaces. Shinn advises traditional enterprises to isolate risk by running low-volume agent pilots, arguing that platforms built primarily on captive user attention face structural decline, while transaction-aligned businesses will benefit from agentic automation. Instinct's long-term viability hinges on whether users remain comfortable delegating complete personal context to an autonomous agent and whether merchants accept distribution take rates in lieu of captive ad impressions. Significant operational uncertainties remain, including managing exponential compute consumption amid hardware supply constraints and mitigating defensive counter-measures from platform gatekeepers like Apple and Meta.

Instinct 创始人 Noah Shinn 阐述了一种全新的个人 AI 智能体范式：通过接入真实通信渠道与独立虚拟计算环境，智能体能够主动介入现实任务并重塑商业交易模式。他认为个人智能体的核心护城河在于可预测性与用户信任，而非单一的技术参数竞赛。 无独立界面的主动型数字助理架构 (0:49) Instinct 舍弃了传统的独立 App 形式，直接融入用户的日常通信渠道。系统为每个智能体分配独立的电话号码、电子邮箱和虚拟计算机环境，用户通过发短信、发邮件或语音通话即可与其交互。与仅响应提示词的传统被动工具不同，Instinct 具备自主的唤醒与睡眠周期，能够根据上下文做出主动判断。例如，若一份重要商业文件要求在下午 3:00 前完成签署，智能体会在 2:55 主动给用户打电话提醒并置顶相关邮件。 由于配备了完整的虚拟计算能力，用户无需受限于预设的工作流。实际使用中涌现出了多种复杂场景：从在全网寻找现场转播生肉镜头，到结合 3D 人体建模与衣橱扫描实现每日服装搭配与自主电商采购，乃至获取银行账户只读权限后自主登录商家网站、拦截验证码并完成订阅服务的取消流程。 对等智能体网络与动态微应用重构协作 (7:15) 为了解决跨个体的协调摩擦，Instinct 推出了“智能体对智能体”（Instinct to Instinct）协作机制。系统基于带权重的“受信任联系人网络”（trusted person network）运行，不同社会关系对应严格的权限分级，如配偶间可共享大量信息，而职场同事仅能查看日程空闲或特定工作邮件。若外部智能体试图越权读取数据，系统会立即向所属用户发出警报。在群体协同中，六位好友的智能体可调取日程与 Spotify 偏好自动制定周末聚会方案，并自主调用 Uber 规划最优的多点拼车路线。 在交互呈现上，Noah 认为多应用林立的生态正面临终结。面对多城市旅行规划或婚礼筹备等需要高信息密度的复杂任务，智能体不再将用户分流至各个第三方 App，而是根据需求实时生成交互式网页微应用（dynamic micro-apps），完成展示后即弃，从而消除多层界面的摩擦。 交易分润商业模式与对注意力经济的替代 (17:23) Noah 指出，Instinct 拒绝采用广告或流量变现模式，因为依靠赞助商佣金引导用户购买会摧毁个人代理最根本的信任契约。平台目前保持免费，其商业模式类似 Apple Pay 或高价值分发渠道，通过为商家提供合格交易来收取佣金。尽管仍处于受邀测试阶段，Instinct 的年化交易总额（GMV）已超过 10 亿美元且日复合增长率约 10%，其中差旅消费占平台总支出的 50%。Noah 认为精品酒店等行业常规提供高达 30% 的佣金，这为平台提供了巨大的交易抽成空间。 在谈及与传统数字化巨头的竞争时，Noah 分析指出，高度依赖用户在线时长和应用内广告的业务（如外卖平台、社交媒体和传统在线旅游代理 OTA）将面临智能体无界面化的直接冲击。然而，对于能够通过降低下单摩擦来促成交易的核心零售业务，主动型智能体反而会因消除决策阻力而扩大整体真实消费量，使实际交易蛋糕变得更大。 渐进信任积累与解耦的安全监控架构 (21:00) 个人智能体的proactive能力高度依赖隐私数据，面临显著的信任门槛。根据 Instinct 的运营数据，信任建立通常需要数周的低风险稳定验证；在持续使用三周内，约 40% 的活跃用户会绑定个人信用卡以授权自主采购。绑定敏感凭据的用户留存率高达 80%，因此“首次绑定信用卡/密码/敏感文件的时间”成为衡量产品认可度的关键内部指标。 为了防范大模型幻觉与越权风险，Instinct 在架构上将核心智能与安全执行层彻底解耦。输入端通过隔离防火墙拦截恶意提示词注入；在执行端，系统引入独立的外部“看门狗”（watchdog）监督机制。在智能体思考链转化为实际工具调用前，看门狗会独立审查参数异常（如捏造的专有名词或错误账号），并拥有暂停、拦截或拒绝交易的最高权限，确保用户始终保留随时撤销授权的数据主权。 算力供应链危机与低延迟解耦的推理经济学 (56:00) 在 10% 日复合增长的驱动下，Instinct 的算力需求每周翻倍，Noah 将 40% 的精力投入于算力规划。由于硬件采购存在 3 至 4 个月的交付周期，若依赖现货市场将面临 3 至 4 倍的成本溢价；而过度提前采购 10 倍容量，在高速复合增长下也会在约一个月内耗尽。此外，主动型智能体需要全天候自主扫描与环境感知，其长期 Token 消耗量远超传统的单次问答工具。 为降低单用户推理成本，Instinct 摆脱了统一调用昂贵前沿商业 API 的做法，利用定制化的批量推理与服务架构削减开支。由于智能体的大量后台工作是主动且非即时交付的（数分钟甚至数小时完成即可），系统将非紧急任务与毫秒级延迟解耦，在同等硬件底座上实现了相比公有云端点 3 至 8 倍的执行效率，从而支撑起面向普通用户的免费服务与规模化扩张。 Instinct 获得了红杉资本（Sequoia）、Benchmark 和 Coatue 约 10 亿美元的融资，估值达 100 亿美元。Noah 强调，巨额资本的作用在于支持极度密集的算力消耗并避开过早设立月费门槛的局部最优陷阱，最终通过重塑各行业的端到端交易生态来实现商业闭环。不过，随着系统对现实生活控制权的深入，算力供应链的持续扩张压力与多应用生态的潜在抵触仍是其推进全自主智能体长期愿景的关键考验。

## Background 本期节目来自商业投资类播客《Invest Like the Best》（第 493 期），由 Colossus 平台于 2026 年 9 月 28 日发布。访谈嘉宾为个人智能体公司 Instinct 的创始人 Noah Shinn，节目深入探讨了 Instinct 作为一款无独立界面的个人助理，如何通过短信与电话替用户自主执行线上操作、依托“智能体互联网络”协调日程，以及其背后的抽成商业模式与算力挑战。 ## TL;DR Noah Shinn 在对谈中详述了个人智能体 Instinct 的颠覆性范式：它无需下载独立 App，而是直接拥有专属虚拟手机号和电脑，依托短信与语音等日常通道替用户自主完成订票、改签、退订等复杂闭环。产品依托严格的邀请制实现日环比 10% 的自然爆发式增长，在极小用户基数下已承载近 10 亿美元年化交易流水（其中 50% 来自差旅）。Noah 强调坚决摒弃出卖用户注意力的广告模式，转而效仿 Apple Pay 通过交易抽成（Take Rate）变现，并分享了为防范越权与模型幻觉而设计的双重解耦安全防火墙与看门狗机制。对谈还深入剖析了前置算力采购的指数级风险、从任务响应迈向长周期目标追求的演进，以及整个数字互联网软件界面终将坍缩为单一极简入口的未来图景。 ## Atmosphere & style 整场对话洋溢着一种深邃、笃定且充满前瞻张力的智性交锋。主持人 Patrick 与创始人 Noah 之间的互动兼具严密的商业逻辑推演与前沿技术落地的务实探讨，没有空洞浮夸的行业黑话，而是直击算力困境、人机信任飞轮与企业注意力税瓦解的真实内核。听众仿佛置身于一场正在重构数十亿美元底层规则的脑暴现场，既能切身感受到指数级增长带来的供应链紧迫感，又能在关于交互克制与去人格化哲学的拆解中体会到一种罕见的人性温情与极简美学。 ## Bulletin - `17:23` 个人智能体初创公司 Instinct 在零市场营销投入与严格邀请制下实现日环比约 10% 的复合增长，年化交易总额已逼近 10 亿美元，其中差旅与旅游场景占比高达 50%。 - `23:25` 产品采用无独立界面的极简形态，赋予 Agent 专属虚拟手机号与电脑，用户仅需通过短信、电话或邮件交互，并在第三周达到 40% 的信用卡绑定率，授权后留存率高达 80%。 - `26:08` Instinct 坚决拒绝广告驱动模式，计划向商家收取交易分发抽成（Take Rate）；同时构建了与 Agent 核心彻底解耦的入站防火墙与外部看门狗审查机制，防止提示词注入与越权操作。 - `01:22:20` 为应对日增 10% 带来的算力每周翻倍困境，公司通过定制底层部署架构将异步任务转为批处理以提升数倍能效，并以约 100 亿美元估值完成由红杉、Benchmark 及 Coatue 领投的近 10 亿美元融资以补贴前置算力。 ## Key takeaways - `00:00` 个人智能体（Personal Agent）不仅是一个消费级应用，而是重新定义人机交互入口与数万亿美元商业范式的核心载体。 - `04:11` Instinct 刻意摒弃独立 App 形态，直接嵌入短信、电话与邮件等既有渠道，通过赋予 Agent 专属虚拟手机号和电脑实现全网通用自主执行。 - `07:15` 全新上线的“Instinct to Instinct”互联网络允许两个智能体直接在后台协商日程与动线，终结跨时区人工来回扯皮的沟通成本。 - `10:00` 构建不同于传统扁平好友图谱的“可信加权社交网络”，辅以细粒度权限控制与越权主动告警机制，用技术透明度强化现实人际信任。 - `12:15` 多智能体协同已自发涌现出自动化聚会策划与顺风拼车编排等高阶用例，将社交活动中的组织与后勤摩擦力降至零。 - `15:00` 无界面智能体将颠覆传统依赖聚合界面的 OTA 与先到先得的订座机制，通过精准传递用户上下文实现供需双方的最优撮合。 - `23:25` 用户信任呈现飞轮效应：使用第三周信用卡绑定率达 40%，且一旦开放核心敏感权限，用户留存率随即跃升至 80% 左右。 - `26:08` 安全防御采用与 Agent 决策核心解耦的独立双重机制，由独立的入站防火墙与出站看门狗系统实时监控并拦截恶意注入与不可逆操作。 - `28:30` 拒绝“羊毛出在猪身上”的广告变现逻辑，Instinct 坚持与用户根本利益绝对对齐，定位于高阶目标守护者而非单一任务执行器。 - `31:07` 商业模式避开底层清算管道 2% 的微薄基点竞争，瞄准高附加值分发决定权，效仿高端分发渠道向商户抽取交易佣金（Take Rate）。 - `34:52` 企业间利益冲突本质是“注意力溢价”对阵“实体履约价值”；Agent 抹平了依靠 App 内停留时长加征的商业税，但零摩擦将带来总交易量的爆发。 - `47:35` 交互设计放弃盲目堆砌 AI 能力指标，将“可理解性”置于首位，并根据人类 F 型视线扫描规律精心设计低认知负荷的短信版式与语气。 - `56:58` 算力采购面临指数级增长与数月交付周期的严峻错配，日增 10% 导致所需算力每周翻倍，现货应急采购则面临 3 到 4 倍的极端溢价。 - `01:00:00` 通过对非即时性任务进行异步批处理和定制底层推理架构，实现了数倍的吞吐能效提升，将推理单价压降至极具承受力的水平。 - `01:02:30` 原生主动型 Agent 具备自发“苏醒与休眠”机制，绝大多数 Token 消耗在幕后的全天候推演与状态监测中，算力消耗远超被动对话型 AI。 - `01:12:35` 所有软件界面未来将坍缩为极致轻量的常开语音或随身入口，并在需要处理复杂长流程时按需即时动态生成 Web 界面，终结 App 疲劳。 - `01:17:40` 拒绝电影《她》式的拟人情感依赖，Instinct 将自身定位于克制、知进退且执行力拉满的高情商幕僚，致力于长周期高阶目标的自主求解。 - `01:22:20` 以 100 亿美元估值吸纳近 10 亿美元风投资金，意在用资本杠杆承担前置算力成本并突破订阅制局部最优，追求数十亿普通人规模的交易终局。 ## Detailed walkthrough ### `00:00` 史诗级软件竞赛与 Instinct 的诞生 当前的人工智能生态正在经历一场前所未有的角逐。这不仅是少数创业团队与全球科技巨头的正面交锋，而且留给各方跑通模式的时间窗口被压缩到了“以月为单位”。在单年算力投入动辄数以亿计、创业公司天然缺乏传统分发渠道壁垒的背景下，一款依靠产品力实现爆发式病毒传播的个人智能体（Personal Agent）正脱颖而出。这场竞赛的终局并非只是做出一个受欢迎的消费级应用，而是重新定义数万亿美元规模的人机交互入口，甚至重塑数十亿人与整个数字技术世界连接的根本范式。 Noah Shinn 创办的 Instinct 正处于这场风暴的中心。回顾过去，当代码生成（Code Generation）在软件工程领域掀起风浪时，行业已初窥智能体的威力；而个人智能体的普及所带来的颠覆性，被认为远远超出代码生成的范畴。然而，在解释 Instinct 到底是什么时，Noah 刻意剥离了所有浮夸的行业术语与概念包装：它就是一个“个人助理”（Personal Assistant）。与许多要求用户“相信未来、几年后才会兑现”的传统科技愿景不同，大众从 2023 年接触 ChatGPT 开始，心中就已经对 AI 应该具备的形态达成共识——人们等待的是一个真正能“替自己行动”、融入日常生活解决繁杂事务的日常智能，而 Instinct 所做的，就是让这个构想“恰如其分地直接运转”（it just works）。 ```mermaid flowchart TD accTitle: Instinct 极简无界面交互架构 accDescr: 展示用户如何通过现有的电话、短信和邮件与 Instinct 交互，Instinct 自主操作电脑与外部服务，无需安装新应用。 subgraph User["用户日常通信渠道（无独立 App）"] A["SMS 短信"] B["电话呼叫 / 接听"] C["电子邮件"] end subgraph InstinctCore["Instinct 个人智能体"] D["具备独立手机号与运算环境 (电脑)"] E["社交情商与情境感知层"] D --- E end subgraph Actions["执行环境与真实世界操作"] F["操作互联网 / 网页端应用"] G["读取 / 确认验证邮件"] H["直接触达目标服务 (取消订阅、预订等)"] end User <--> InstinctCore InstinctCore --> Actions ``` --- ### `04:11` 极致简单的界面与自主执行的“疯狂”用例 Instinct 最激进的设计哲学在于它“根本没有独立 App”，它不是一款新工具，而是一种全新的生活体验。Instinct 拥有属于自己的虚拟电话号码和一台虚拟电脑，这意味着用户无需学习任何新界面，可以直接通过短信、拨打电话或发送邮件与它交互。它具备极高的“社交情商”（social intelligence）： - 它只在真正关键且紧急的时刻主动打电话给用户（例如某位早期深度用户在几个月里仅接到过三次电话）。 - 电话中的语气高度克制且精准：“我知道这很打扰，但这篇文档必须在下午 3:00 前完成签署，现在已经是 2:55 了，您能在收件箱里处理一下吗？如果不方便，我也可以重新发一封邮件把它顶到最前面。” 这种极简交互的背后对应着不受限制的通用能力。由于智能体拥有一整台“电脑”，凡是人类在互联网上能够完成的动作，它原则上都能自主执行。此前美网公开赛（US Open）期间便有轰动一时的案例：一对夫妻被球场大屏幕（Jumbotron）抓拍转播，智能体根据极简提示，自行通过网络搜索、锁定信号流并找回了那段比赛画面的现场录屏。而在日常生活中，用户的创意甚至催生了更前沿的用法： - **端到端衣橱扫描与穿搭购物**：用户拍下衣柜里所有的上衣、裤子、袜子、鞋子，并扫描自己的身材与面部比例。Instinct 并非生成单调的文字穿搭清单，而是直接生成用户本人“穿着整套服装”的高清预览图，为用户规划整周搭配；不仅如此，它还能在全网数千个电商品牌中搜寻风格匹配的新品，每天向用户推送三套由本人亲自“试穿”的整套穿搭方案，用户只需确认心仪的一款，智能体就会自主下单并邮寄到家。 - **全流程理财与订阅杀手**：许多用户将银行账户接入 Instinct 并设定财务目标（如本月必须省下特定金额）。智能体会遍历历史交易与订阅扣款，主动询问用户“这项扣款你真的还在用吗？”一旦用户表示否定，它不会像传统金融记账软件那样只丢给用户一个外部链接，而是直接登录目标网站、读取邮箱中的动态验证码、在后台完整取消续订，最终直接向用户汇报：“本月已成功为您省下 2,000 美元。”这种对繁琐流程的彻底接管，意味着任何依赖“消费者惰性与拖延症”而存活的商业模式都将面临灭顶之灾。 --- ### `07:15` “智能体对智能体”互联网络（Instinct to Instinct） 为了解决真实世界中最消耗精力的协调成本，团队在近期上线了“Instinct to Instinct”互联网络，探索两个自主智能体之间直接对话协作的机制。该功能的初衷并非为了增加噱头，而是针对高频开会的职场人士痛点量身定制。在传统日程安排中，跨时区、跨行程的会议往往需要经历漫长且低效的反复拉锯： > “你周三有空吗？” > “抱歉那天不行，但周四下午可以。” > “周四下午我正在赶飞机，而且有三个小时时差，周五一早行吗？” 当对话双方都拥有 Instinct 时，繁琐的沟通被彻底终结。用户不再需要手动翻阅日历并反复推敲措辞，只需对各自的智能体抛出一句意图指令：“我想在本周末前和某某见个面。”两个智能体便在后台直接对接双方的日历空闲、旅行动线和偏好窗口，在几秒钟内完成协商并锁定最佳时间，让“订会议”这一经典互联网交互彻底退入幕后。 - **上线时间**：该互联网络在对谈时仅推出了约 7 到 10 天，展现出极高的实用价值。 - **去中间化**：剥离人类在日程沟通中的博弈与客套，使时间匹配从多轮来回推拉降维为单次意图驱动的自主达成。 ### `10:00` “可信个人网络”与加权社交图谱 Noah Shinn 解释了 Instinct 最核心的突破之一——“可信个人网络”（Trusted Person Network）。这一机制允许两个人的 Instinct 智能体直接跨设备协商，比如在需要找时间碰头时，双方便自动比对日程空档并在日历上锁定时间。这一机制的基石在于严格的信任边界，用户只会连接自己完全信赖的人，杜绝恶意窥探日历或收件箱的行为。在此基础上，系统提供了极其细粒度的权限控制体系：夫妻之间通常彼此完全透明，授权“向对方共享任何信息”；而面对工作同事，则可能仅开放工作日历以及涉及特定项目文档的部分收件箱权限，其余一律屏蔽。 这种设计催生出一种完全不同于传统社交网络的网络效应。传统的“好友关系图谱”（Friend Graph）往往是扁平的二元状态（互为好友或非好友），而 Instinct 构建的是一个带有不同“权重边”的信任节点网络。每个连接都承载着具体的信任等级与配置。Noah 提到，这种网络正在塑造新的社会动态与微妙的人际张力：如果有人试图利用网络权限越界刺探数据——例如对方只被授予了日历查看权限，其智能体却尝试挖掘其他未授权领域的信息——Noah 的 Instinct 就会立即发来短信预警：“顺便提醒一下，Patrick 正在尝试检索某某方面的信息。”这种技术透明度反过来强化了现实中的社交契约与规范，迫使人们在享受自动化协同便利的同时，不敢轻易破坏人际信任。 ```mermaid flowchart TD accTitle: 可信个人网络架构与动态权限控制 accDescr: 展示用户间通过Instinct智能体建立差异化信任网络及越权警报机制。 subgraph UserSide["用户与核心智能体"] UA["用户 A (授权方)"] --> IA["Instinct A"] UB["用户 B (访问方)"] --> IB["Instinct B"] end subgraph TrustNetwork["可信个人网络 (加权连接)"] IA <-->|"受限数据交互 (如仅限日历)"| IB IA -.->|"越权探测告警: 'B正在检索未授权信息'"| UA end subgraph Levels["差异化权限分级"] L1["配偶: 全量数据透明共享"] L2["同事: 仅限工作日程与特定文档"] L3["临时访问: 单日位置共享"] end IA --- Levels ``` - **加权社交网络（Weighted Graph）**：不同于 Facebook 等平台非黑即白的单层好友关系，Instinct 上的每个节点通过精细的权限权重彼此连接。 - **现实社会契约的技术映射**：系统结合技术层面的访问限制与现实社会的道德声誉机制，若发生越权行为，智能体会主动向宿主告警，破坏关系信任的代价将被即时暴露。 --- ### `12:15` 终结聚会组织噩梦：从日程协同到自动拼车 Patrick 对这一前景大为感叹，认为这将彻底重塑人际交互秩序。在日常生活中，哪怕只是夫妻之间临时找人，只要授权“一天的临时位置权限”，问一句“他现在到哪儿了”，就能省去无数沟通成本。Noah 强调，可信网络是目前平台中最具创造力、最超出用户预期的板块。在其他功能上，用户大体能预判智能体的行为，但在这个多智能体协同网络里，不断涌现出令人惊喜的自发用例。 最典型的痛点就是朋友聚会策划。传统模式下这是一项极其疲倦的主动劳动：某人在周四甚至周六晚上开始在群里发问“大家想干嘛”，接着便陷入漫长的扯皮，总有人时间不合适或不感兴趣，最后提议往往无疾而终。而现在，一个 6 人好友圈做到了全流程托管：这 6 个人的 Instinct 彼此互联，由于系统绑定了所有人的 Spotify 听歌历史与日程表，智能体每周会自动进行头脑风暴，根据 6 个人的音乐偏好推荐正在举办的音乐会或演出，并协调彼此都空闲的窗口期。更令人惊叹的是出行环节：智能体直接由其中一人名义叫了一辆 Uber，并沿途精细计算最优接送路径，自动向另外五人发出通知——“10 分钟后在楼下接你”、“5 分钟后在转角接你”，用一辆车将所有人顺次接齐。繁琐的社交推诿与繁重的后勤负担被一扫而空，人们只需坐享聚会本身的乐趣。 - **群聊后勤损耗清零**：消除群聊中“提议—改期—有人退出—流产”的恶性循环，智能体根据全员空闲档期与共同兴趣自动化提案。 - **Spotify 偏好画像结合**：通过读取群成员共同的流媒体听歌记录，精准匹配线下演出或新颖活动。 - **动态拼车路径优化**：单一 Uber 订单打通 6 处经停点，智能体自动调度各人下楼时间，兼具经济效益与协同效率。 --- ### `15:00` 软件界面的终结与餐饮预订机制的重构 Patrick 指出，Instinct 的直观之处在于它就像“一个拥有自己手机、电脑、邮箱且能给你打电话的超人助理”。他探讨道，AI 代码生成的爆发虽然惊艳，但历史上程序员始终是极少数群体；而个人智能体面对的是全人类的普遍需求。面对 Patrick 提出智能体将如何重构互联网、商业与交互界面的宏大问题，Noah 明确表示：“在未来几年内，互联网将被彻底重写，绝大多数人类与软件交互的方式将发生翻天覆地的变化。” 为了避免空谈愿景，Noah 拆解了短期到中期内的一个典型落地场景：餐厅预订。在传统的 Web 时代，餐厅订座采用先到先得的机制（First-come, first-served），用户必须准点守在屏幕前不断刷新页面，一旦手慢便一无所获。当用户拥有智能体后，最原始的玩法是让智能体每隔 5 秒就在全球每个城市的所有订座网站上持续轮询刷新。但 Noah 透露，Instinct 正在构建更高维度的两端重塑，并即将公布重磅合作伙伴： - **需求端错配的终结**：传统订座无法区分重要性。无论是随便吃顿便饭的本地熟客，还是庆祝 30 岁整寿、结婚纪念日或外地挚友重聚的用户，在系统眼里都是相同的代码。 - **意图与供给的精准撮合**：Instinct 让智能体得以在预订过程中充当代理中介，向餐厅直接传达上下文——“这是宿主配偶的 30 岁生日，属于人生重大里程碑”；而餐厅本身也渴望招揽愿意高消费、庆祝特殊纪念日的高价值宾客，而非普通占位者。智能体消除了繁琐的人工沟通成本，让餐厅端的需求与用户端的诉求实现帕累托最优匹配。 --- ### `17:23` 10 亿美元交易规模与重塑在线旅游生态 从餐饮预订延伸开去，在线旅游行业（OTA）正在经历更为剧烈的冲击。Noah 披露了一个极其惊人的财务数据：尽管 Instinct 目前依然维持着严格的邀请制（Invite-only）、处于非常早期的阶段且用户基数相当小，但平台上的年化交易总额（Transaction Volume）已经逼近 **10 亿美元/年** 的体量。在这一庞大流水中，仅**差旅与旅游（Travel）相关的支出就独占了 50%**。 传统的在线旅游聚合平台（如各大 OTA 与比价引擎）其核心价值在于界面整合：将成百上千家航空公司、连锁酒店与租车服务集中在一个美观的 UI 中，让用户挑选、结账并在邮箱里收到确认单。然而 Noah 认为，这种多层点击的图形界面终将被取代。在个人智能体的全新范式下，用户无需再打开复杂的比价网页逐项筛选，只要发送一条语音备忘录或一条短信：“帮我订好去某地的行程”，智能体就能调用全部上下文，在后台以极高效率无缝完成从比价、选座、支付到日程录入的全链路闭环。 - **惊人的单位价值产出**：在用户规模极小、不花分文市场费用的邀请制阶段，Instinct 年化交易流水已逼近 10 亿美元大关。 - **差旅场景统治力**：旅游出行占平台整体消费总量的 50%，成为智能体代理支付的最核心落地点。 - **聚合 UI 的瓦解**：OTA 平台赖以生存的信息聚合与前端展示价值，在直接具备行动执行力（Action-oriented）的端到端智能体面前正在迅速失效。 ### `20:00` 极简交互背后的复杂旅行编排与偏好外推 Noah Shinn 深入拆解了使用 Instinct 处理高难度现实任务的极简体验。以一句极其简单、口语化的语音留言为例：“嘿，我今晚必须赶到纽约，十万火急。”对用户而言，交互在说出这句话的瞬间就已彻底结束。然而在幕后，Agent 必须迅速拆解并编排一整套极其庞大且充满细节的逻辑链条： ```mermaid flowchart TD accTitle: Instinct 自动化行程编排与偏好外推流程 accDescr: 展示从用户一条简短语音输入到全流程出差预订与日历对齐的完整逻辑流 subgraph Input ["输入层"] A["语音输入：'今晚必须赶到纽约'"] end subgraph Context ["上下文感知与记忆外推"] B["定位当前位置（如旧金山 / 洛杉矶）"] C["提取长期偏好：航司 / 舱位 / 靠窗或走道 / 餐食 / 信用卡"] D["偏好外推（推断并复用酒店品味）"] end subgraph Execution ["执行与联动"] E["端到端完成机票与酒店支付预订"] F["全流程时间线排期：去机场 Uber + 接机 Uber + 当地日程"] G["无缝写入个人日历"] end A --> B B --> C C --> D D --> E E --> F F --> G ``` 系统首先通过位置权限识别用户所处城市（例如旧金山或洛杉矶），随后调取过往沉淀的用户个性化档案：偏好乘坐哪家航空公司、习惯坐头等舱还是经济舱、选座偏好走道还是靠窗、机上特殊餐饮需求、默认使用哪张信用卡，随后直接自主完成出票。紧接着是酒店预订，如果用户过去在该城市已经连续 7 次入住同一家酒店，Agent 会直接选定；即便去往新城市，它也能将用户在特定场景展现出的审美与品味“外推”（extrapolate）到全新预订中。最终，Agent 将航程、前往机场的 Uber、降落后的接机网约车以及当地所有的会议活动，全闭环编排并对齐到个人日历上。 这一全自动流程带来的不仅是省心，更是一次对传统服务中介的降维打击。当传统依靠提供标准化预订界面的平台（如各大 OTA）面临这种全新的、更符合人性的交互方式时，其商业格局将不可避免地重构。Noah 明确表示，Instinct 的目标并非盲目颠覆这些老牌企业，因为它们历年积累的庞大供应链数据和网络生态依然极具价值，未来几年双方更有可能是深度协作的关系，共同重新定义行业体验。 ### `23:25` 代理信任的飞轮效应与 40% 信用卡绑定指标 将个人数字生活的钥匙交给一个自主 Agent，必然涉及隐私与效率之间的博弈。用户输入的上下文越多、绑定的敏感账号与密码越全面，Agent 的主动性与能力上限就越高；然而用户内心天然存在对隐私泄漏的防御机制。谈及如何打破这种“先有鸡还是先有蛋”的数据飞轮僵局，Noah 分享了 Instinct 在实际运营中观察到的信任建立规律。 Noah 强调，团队始终坚持一项铁律：**用户必须始终对自己的数据拥有绝对的控制权**。数据分享的节奏完全取决于用户的舒适度，且任何时候都可以全额撤回。平台内部监测发现，信任的建立通常需要数周的循序渐进，而团队非常重视一系列代表信任建立的关键代理指标（proxies for trust），如“首次录入信用卡的时间”（Time to First Credit Card）、“首次输入账号密码的时间”等。 - **三周绑定率**：在用户使用产品的第三周，有高达 40% 的用户主动将个人信用卡信息完全托付给 Instinct（该统计已包含在此期间自然流失的用户）。 - **留存质变点**：一旦用户向 Instinct 授权哪怕一件核心敏感信息（如信用卡、核心账户权限），其**用户留存率（Retention Rate）便会飙升至 80% 左右**，这在消费级技术产品中是一个极其惊人的粘性数据。 - **直觉化依赖**：经历最初三周的磨合后，用户对 Agent 的依赖会产生滚雪球效应，找 Instinct 解决一切琐事甚至变成了一种“本能”（instinctual）。 ### `26:08` 敏感数据隔离与双重解耦安全防火墙机制 面对普通用户对 AI Agent 掌控核心隐私与财产权限的天然顾虑，Noah 将系统的安全架构清晰地拆分为两层逻辑进行论证：底层是传统的数据存储安全，表层则是针对 Agent 全新自主攻击面设计的独立防护系统。 对于第一类静态数据存储与隔离，属于行业内已有成熟实践的可解问题（tractable problem），只需要工程团队以极高的标准、专注与细致度进行加密防护，保障用户完全掌控数据边界。更具挑战性的是第二类问题：当 Agent 拥有全自主支付权限、邮箱完全访问权和日历管理权后，它直接暴露在复杂的互联网环境中，必须防御全新的攻击向量（如恶意 Prompt 注入或钓鱼欺诈）。 为确保绝对安全，Instinct 并未将安全风控逻辑集成在 Agent 本身内部，而是构建了一套完全与 Agent 核心架构解耦（decoupled）的外部防御体系： ```mermaid flowchart LR accTitle: Instinct 双重解耦安全防火墙架构 accDescr: 展示外部多模态数据输入经入站防火墙过滤，以及 Agent 思考与行动经独立监控系统拦截审核的双重风控机制 subgraph Inbound ["入站层（Inbound）"] A["外部多模态内容<br/>（文本/邮件/网页/媒体）"] B["入站安全防火墙<br/>（拦截/过滤/阻断恶意提示注入）"] end subgraph Brain ["核心层（Core）"] C["Instinct Agent 核心<br/>（推理规划 / 上下文感知）"] end subgraph Outbound ["出站层（Outbound）"] D["独立监控与拦截系统<br/>（实时审查思考与预执行动作）"] E["执行真实世界动作<br/>（支付/发送邮件/日历变更）"] end A --> B B -->|安全放行| C C -->|输出意图| D D -->|核准确认| E ``` 1. **入站防火墙（Inbound Firewalls）**：任何即将被 Agent 摄入的外部数据（包括邮件文本、网页内容、各类多媒体信息），都必须先经过独立的过滤网关，以识别并拦截潜在的恶意注入攻击，防止外部内容利用提示词诱导 Agent 做出越权行为。 2. **出站解耦监控系统（Decoupled Action Monitor）**：Agent 在产生任何外部操作意图或内部规划思考时，均受到另一套独立外部系统的实时监控。该系统具备随时暂停、拦截、审批或否决 Agent 下一步动作的绝对权限，从而杜绝 Agent 在未经审查的情况下产生不可逆的财产或隐私损失。 ### `28:30` 商业模式与个人利益对齐（Alignment）的抉择 对谈随后引向 Agent 领域最为核心的伦理与商业命题——对齐（Alignment）。主持人指出，这里的对齐并非指宏大的人类生存级安全（humanity scale alignment），而是指**Agent 是否在微观层面与单个用户的切身利益保持百分之百的一致**。这类似于雇佣关系：用户付薪水给人类助理，助理受雇佣契约约束，便不会暗中引入第三方利益来操纵决策。 然而，在 AI 产品的商业模式选择上，行业通常会分化为两派：一派向用户直接收费，另一派则走“免费+广告变现”的流量路线。后者通过向广告商收费盈利，势必会扭曲 Agent 的决策中立性。Noah 对此给出了斩截的定调：Instinct 绝不会为了商业变现去引导或操纵用户的行为， Agent 的存在意义决不能与用户的根本利益发生冲突。 ### `30:00` 拒绝广告模式：AI 绝不能背刺用户 当个人智能体 Instinct 变得比人类更聪明——不仅是教科书式的智力更高，连社交情商与洞察力都远超人类时，危险也随之而来。Noah 强调，如果 Instinct 利用这种超级智能去操纵、劝说用户购买他们本不需要的商品或订阅不想要的服务，世界将陷入可怕的境地。当下绝大多数消费级互联网平台（如 Google、TikTok、Instagram、Snapchat 等）表面上免费，实则充斥着付费广告，平台靠着“让品牌买单、促成用户转化”盈利。这正是典型的“如果你不为产品付费，那你本身就是产品”。作为程序员和技术人员，Noah 坚决拒绝构建这种模式。 为了在根本上杜绝操纵用户的可能性，Instinct 在产品逻辑和前沿研究中确立了极高的标准：**它绝不做单一的“任务执行者（Task Accomplisher）”**。在传统的 AI 产品中，用户给一段 Prompt，AI 机械地执行并汇报结果；一旦遇到恶意或欠妥的指令，缺乏高阶目标的任务执行者只会盲从。相反，Instinct 遵循更高维度的目标体系：与用户建立真正的信任、让用户感到被保护、在关键事务脱落时主动补位兜底。完成具体任务只是建立安全感与信任的一种手段，正是这种高阶目标架构，让 Instinct 在面临复杂边缘案例时能更稳健地站在用户一边。 - **高阶目标与传统 AI 对比**：传统 AI 仅仅是一问一答的任务执行机器，而 Instinct 把“守护用户利益、构建信任”作为最顶层的决策锚点，是传统功能集的超集。 - **商业飞轮与增速**：平台目前每年承载的交易流水已突破 10 亿美元，且正以每天 10%（10% day-over-day）的速度复合增长（今天 10 亿、明天 1.1 亿增量递进）。 - **平台商业化定调**：既然坚决不碰广告，Instinct 计划效仿 Apple Pay 或 Amex（美国运通）的双边模式——对用户端完全免费，对后端商家按照带来的分发价值统一收取全平台的交易抽成（Take Rate）。 ```mermaid flowchart TD accTitle: Instinct 与传统广告平台的商业模式对比 accDescr: 对比传统消费级平台的广告变现路径与 Instinct 站在用户侧的抽成变现路径 subgraph AD_MODEL["传统平台模式（如 Google / TikTok）"] U1["用户"] -->|"免费使用注意力"| P1["广告驱动平台"] B1["品牌广告主"] -->|"支付广告费推动转化"| P1 P1 -.->|"诱导/操纵消费"| U1 end subgraph INSTINCT_MODEL["Instinct 模式（用户利益至上）"] U2["用户"] -->|"高阶信任与委托"| A1["Instinct Agent"] A1 -->|"分发纯粹消费需求"| M1["商户 / 底层服务商"] M1 -->|"支付平台分发抽成 (Take Rate)"| A1 M1 -->|"交付实际商品与服务"| U2 end ``` ### `31:07` 支付分层与价值捕获：不做 2.5% 里的零头，做高附加值分发 面对主持人的提问——在信用卡发卡行、Stripe、Visa、MasterCard 等成熟支付链路林立的当下，Instinct 究竟如何切入分润——Noah 剖析了整个支付链条的真实利益格局。一笔数字化支付背后可能养活了 40 个上下游参与者，但这 40 家机构瓜分的其实仅仅是底层那 2% 到 2.5% 的薄利。Instinct 根本无意去扮演底层清结算通道，也不打算去抢 Stripe 或 Amex 在管道层里赚取的零碎基点（如从 2.5% 里抠 30 个 bps）。 真正巨大的商业价值在于**分发渠道对最终交易的决定权**。纵观各大平台的抽成生态：底层 SaaS/工具型的 Shopify 抽成约在 2% 到 3% 之间；具有强消费心智与聚合能力的 Amazon 抽成可达 10% 以上；而拥有极强生态控制力的 Apple App Store 则能高达 30%。Instinct 目前的交易流水中，有整整 50% 都来自旅游行业（Travel）。在旅游行业，OTA（在线旅行社）针对机票的抽成或许在低个位数百分比，但大量精品度假酒店（Boutique Hotels）为了获取优质客源，极其愿意支付超过 30% 的佣金。Instinct 前期完全可以借力这些既有的商业成熟机制启动，未来再将这一分发驱动的高 Take Rate 模式平移推广到所有主流数字商业领域。 | 平台 / 机构类别 | 代表案例 | 预估抽成率 (Take Rate) | 价值属性与商业定位 | |:--- |:--- |:--- |:--- | | **底层支付基础设施** | Visa / MasterCard / Stripe 等 | 合计约 2% - 2.5% | 基础支付管道，多方瓜分极微薄的基点利润 | | **建站与工具平台** | Shopify | 约 2% - 3% | 提供基础商业工具与支撑服务 | | **聚合电商市场** | Amazon | 约 10%+ | 掌握核心商品分发流量与货架曝光 | | **生态封闭渠道** | Apple (App Store) | 约 30% | 强护城河平台，掌控用户与数字内容交付 | | **旅游分发 (OTA / 酒店)** | 传统 OTA 及精品酒店直连 | 视品类在个位数至 30%+ 不等 | 依靠分发订单赚取丰厚佣金，Instinct 早期切入点 | ### `34:52` 智能体战争：当“注意力税”被无界面交互抹平 主持人敏锐地指出，一场剧烈的“企业级智能体大战（Corporate Agent War）”正在逼近，此前 Amazon 与 Muse 的摩擦便可见端倪。以 Uber Eats 为例，传统用户点外卖需要经历繁琐的 App 翻找与点击流，如果用户直接在 WhatsApp 上向 Instinct 发一句“来份常吃的老样子”，Instinct 调用其独立虚拟电脑代为下单，原本依靠用户眼球停留、开屏广告和推荐位盈利的 Uber Eats 势必会产生强烈的敌意和反弹。既得利益巨头终将筑起高墙防御智能体。 Noah 承认这必然是一场激烈的巨变，数字经济正在被重塑而非单纯被颠覆。为了厘清这种利益冲突的本质，Noah 团队对各行各业的数字服务建立了一套分析坐标轴：**评估其营收中究竟有多大比例来自于“用户的注意力/应用内停留时长（即广告和注意力溢价）”，又有多少比例来自于“底层交付的实体商品或服务本身”**。那些高度依赖用户界面注意力、广告流量变现的业务（如部分外卖平台、OTA、媒体聚合应用），在面对完全绕过 UI 界面的智能体时，将遭受最猛烈的价值链震荡；而真正具有实体履约能力的供应商，则将成为智能体生态中最核心的底层伙伴。 - **“无界面（No Interface）”的威力**：智能体直接代劳所有操作，打破了原有的 App 打开率与停留时长逻辑，消解了巨头的界面控制权。 - **冲突与重塑的本质**：巨头与智能体对抗的根本原因，是智能体剥离了巨头依靠“注意力”加征的商业税，迫使所有服务重新回归底层履约效率的价值比拼。 ### `40:00` 摩擦力归零与主动式商业范式 在探讨平台商业模式时，许多人习惯把现有模式简单拆解为“70% 是广告收入，30% 是底层商品与交易本身”，并悲观地认为一旦 Agent 剥离了广告展示页面，这些企业的营收就会被直接削减七成。诺亚·辛（Noah Shinn）提出了截然相反的视角：以 DoorDash、Uber Eats 或 Uber 为例，哪怕只是缩短用户从浏览到结账的点击次数与时间，其交易总量（Transaction Volume）就会因为摩擦力的下降而显著攀升。而 Instinct 正在做的事情，是将线上操作的摩擦力压缩至接近于零——在“主动式行为”（Proactive Behavior）的场景下，这种摩擦力甚至“在物理意义上完全归零”。 诺亚描绘了极具画面感的主动式服务场景：假设 Instinct 接管了你的日程日历，知晓你几点必须线下出席哪场会议。用户再也不用在脑海里默默计算“我还有 5 分钟必须叫车、路上预计堵车 45 分钟、让我赶紧打开 App 看一眼路况”，Instinct 会直接在最恰当的时间自动叫好一辆车在楼下等候，利用实时路况计算确保你永远不会迟到。在外卖场景中，当你深夜刚下飞机坐进网约车，Instinct 已经感知到你还未进食，它会发来一条短信：“你饿了吗？需要我帮你点 5 天前吃过的那家外卖送到家吗？我知道你已经在 Uber 上了，预计二者会同时到家。”用户只需回一个字“好”，交易便已闭环。 ```mermaid flowchart TD accTitle: 传统 App 被动交互与 Instinct 主动式服务对比 accDescr: 对比用户在传统 App 中多步操作的高摩擦路径与 Agent 依据上下文全自动触发的零摩擦闭环路径。 subgraph Traditional ["传统 App 交互（高摩擦·消耗注意力）"] A["意识到需求/查看日历"] --> B["打开 App & 承受广告/算法推荐"] B --> C["反复比对路况/菜单与手动结账"] end subgraph Proactive ["Instinct 模式（零摩擦·主动闭环）"] D["感知日程/航班/位置上下文"] --> E["主动预判需求（提前叫车/推送确认）"] E --> F["用户单字确认即完成交易交付"] end ``` 这种交互范式颠覆了过去互联网平台赖以生存的逻辑：以往平台通过“折磨式地让用户在 App 内消磨时间”来贩卖注意力盈利；而未来，用户在 App 界面上花费的时间虽然锐减，但由于摩擦力彻底消失，用户与企业服务的交互频次和总体交易额反而会大幅飙升。面对这种变革，诺亚强调 Instinct 作为一家初创公司，绝不会以咄咄逼人的傲慢姿态去强行“颠覆”巨头，而是带着真实的用户交互数据去拜访各行业企业，探讨如何通过合作打造双赢模式，使企业的交易生态与端侧用户的体验同时得到提升。 - **摩擦力与交易量的反比关系**：点击步骤越少、结账路径越短，真实购买转化率就越高；当 Agent 将摩擦力降至零时，原本被繁琐流程压抑的长尾需求将完全释放。 - **上下文感知的主动闭环**：结合日历日程、航班动态、实时车程计算与历史偏好，Agent 能在用户产生认知负担之前预先完成商品与服务的调度。 - **合作而非强行颠覆**：Instinct 目前正与各行业先锋企业接洽，利用平台产生的新型交互数据，共同设计兼顾商业收益与用户体验的对接通道。 --- ### `43:40` 商业如何备战 Agent 时代：注意力收割 vs 真实服务价值 主持人帕特里克（Patrick O'Shaughnessy）指出，如果未来半数以上的普通人都拥有像 Instinct 这样全能的个人助理，这项技术实际上将过去极少数富豪才能享有的私人秘书能力彻底“平民化”（Democratizing）。面对这个 Agent 深度渗透的世界，企业究竟该如何自处与布局？ 诺亚认为，企业必须审视自己收入来源的底层构成：你的营收究竟来自用户在应用内消磨注意力，还是来自用户对底层实体商品和服务的消费？ - **服务型与履约型企业**：如果企业的核心价值在于促成交易（哪怕以减少用户在 App 停留时间为代价），那么这类企业将在 Agent 时代迎来巨大爆发，因为 Instinct 会让交易的完成变得轻松百倍。 - **注意力收割型企业**：如果几乎 100% 的收入都依赖于强行占有用户的屏幕时间（甚至往往违背用户真实意志，比如通过诱导沉迷的算法机制将用户困在无休止的“无限下拉刷新”信息流中，令用户在刷完后感到空虚和懊悔），那么这类商业模式将遭遇致命冲击。Instinct 的最高使命是“代表用户的最大利益行动”，正是要将人类从这种被算法绑架的注意力黑洞中解放出来。 ``` +--------------------------------------------------------------------+ | 企业商业模式在 Agent 时代的分野 | +------------------------------------+-------------------------------+ | 基于真实服务与交易转化 | 基于恶意 UI 与强制注意力留存 | | (如 Uber、外卖、电商核心交易) | (如无限流信息流广告、套路点击)| +------------------------------------+-------------------------------+ | 特征：以促成履约为目的 | 特征：千方百计延长屏幕停留时长| | 应对：积极拥抱 Agent 接口 | 困境：用户意志被 Agent 代理后 | | 结果：摩擦力清零，订单与 GMV 暴增 | 结果：无意义的留存泡沫被挤破 | +------------------------------------+-------------------------------+ ``` 对于传统企业如何着手准备，诺亚建议切忌一刀切地盲目豪赌，企业无须冒着“70% 营收归零”的风险盲目全盘开放接口，更科学的路径是运行可控规模的“灰度实验与 A/B 测试”。例如，先针对 1% 的用户群体开放 Agent 对接，以此观察用户真实的消费意愿、品牌美誉度以及交易量的变化。特别是在“商品发现”（Product Discovery）环节，大量用户往往有消费意向却受制于搜索匹配困难，Agent 的介入能够精准帮用户找到心仪之物。目前，Instinct 正在与一批具有前瞻思维的 CEO 和高管团队紧密合作，跑通这套基于可度量数据的渐进式应对方案。 --- ### `47:35` 超越能力论：为什么“可理解性”胜过 AI 功能堆砌 访谈转向 Instinct 背后的技术实现与算力开销。主持人提到，诺亚目前将高达 40% 的精力都花在思考算力（Compute）问题上，但在切入算力细节之前，诺亚特别希望阐明 Instinct 作为一款产品其底层的精神气质与质感追求。 诺亚坦言，探讨“产品的感知与质感”是他可以聊上一整天的核心议题，这也是当下 AI 产品经理急需建立的“全新心智肌肉”。回顾过去三年，整个科技行业充斥着各种喧嚣的产品发布会，所有团队都在疯狂标榜“AI 现在能做到 X”、“因为底层的某个技术小突破，AI 又解锁了 Y 能力”。在诺亚看来，这种无休止的“能力秀”让普通消费者感到极度疲劳（Fatigued），甚至根本不知道日常该如何使用，这完全本末倒置了。 Instinct 在创立之初便确立了一条极其关键的产品原则：“放弃盲目追逐能力（Capabilities），将全部精力聚焦在‘可理解性’（Understandability）上。”所谓可理解性，是指用户能清晰感知系统正在做什么，并能高度准确地预判当自己发出一条指令或产生某种互动时，Agent 会做出何种反应。正是这种清澈、确定且符合直觉的心理预期，带来了无与伦比的使用愉悦感，也是 Instinct 在严格实行邀请制、零市场营销投入的情况下，能够实现**每天约 10%（10% Day-over-Day）病毒式口碑增长**的根本秘密。用诺亚的话说，产品沟通的终极意义不仅在于文本字面所传达的逻辑，更在于互动瞬间让用户产生的笃定与信任感。 ### `50:00` 文本排版细节与降低认知负荷 Noah 谈到在设计 Instinct 的短信交互时，团队甚至会深究到单条短信的具体形状与版式，以及用户看到它时的真实心理感受。如果直接给用户推送一大段密密麻麻的文字，用户就必须手动滑动屏幕才能看清后续信息，体验极差。相反，更合理的做法是把关键信息前置（front-load），让用户在阅读前 30% 内容时就能抓住核心结论，其余部分则作为可选阅读。 ```mermaid flowchart TD accTitle: 文本排版与用户视线扫描模式 accDescr: 对比传统大段文本与为 F 型阅读优化的智能短信排版结构 subgraph Problem["传统排版问题"] A["长篇大段文字"] --> B["产生视力疲劳与滑动负担"] end subgraph EyeScan["人类视线扫描规律（F型/旗帜状）"] C["首行精读 80%"] --> D["次行快速扫过 50%"] D --> E["后续行视线急剧衰减"] end subgraph AgentCraft["Instinct 优化策略"] F["核心要点前置 (前 30%)"] --> G["精简行宽与行数"] G --> H["大幅降低认知负荷与阅读疲劳"] end EyeScan -.-> AgentCraft ``` Noah 指出，心理学与人机交互研究表明，绝大多数人在浏览大段文本时的视线轨迹像一面“旗帜”（或者经典的 F 型浏览模式）：第一行读 80%，第二行读大概 50%，随后视线迅速向右下方衰减。Instinct 在生成回复时能够主动考虑这种人类阅读习惯：既然用户绝大部分注意力只会停留在前两行以及每行的前半句，那么 AI 就必须据此组织措辞，以极低的信息疲劳度和认知负荷呈现给用户。Noah 强调，虽然团队在底层架构上投入了大量精力来保证系统运行又快又便宜，但这种对产品交互“体感”的极致考量，才是 Instinct 难以被取代的核心护城河。 ### `51:30` 品味优先与渐进式测试验证机制 主持人询问，这些细腻的交互体验究竟主要源自 Noah 及核心团队的个人“品味”（taste），还是来自严密的量化迭代过程（例如通过 A/B/C 测试 emoji 反应、短句与长句的留存数据）。Noah 回应称“二者皆有”。针对“用户使用两三周后是否真正对 Instinct 产生信任”这种偏软性、长周期的核心指标，单靠自动化评测（Evals）或模拟测试是极难准确度量的。 因此，团队采用了一种主观把控与渐进式灰度发布（staged rollouts）相结合的方法： 1. **创始人主观内测**：Noah 自己往往先编写一套稍有差异的交互体验，仅向自己一人发布并试用把玩，凭自己极度挑剔和主观的审美体验来校验感受。 2. **核心团队验证**：如果创始人自身觉得舒适顺畅，再推给公司内部全体团队成员测试反馈。 3. **种子用户试水**：随后推送给范围极小的早期尝鲜群（early access group），收集真实的使用心理感受并做针对性微调。 4. **全量公开发布**：确认一切就绪后，才面向全部用户正式推出。 Noah 承认，未来几个月或几年内，市场上肯定会出现其他看似能提供相同功能、同样强大的 Agent 产品；但 Instinct 这种不仅关注输出字面内容，更死磕“用户整体情绪感受与使用愉悦度”的执念，将是决定胜负的关键。 ### `53:30` 5 个邀请码引发的裂变狂潮与社会学现象 面对主持人关于“现存巨头凭借分发渠道迅速赶超，还是初创公司凭借产品品质快速获取分发”这一经典科技命题的提问，Noah 详述了 Instinct 惊人的自然增长数据。尽管平台至今保持着严苛的邀请制，但 Noah 澄清这绝非为了饥饿营销或刻意制造优越感，而是出于防止服务器因用户暴增而彻底崩盘的自我保护。Instinct 给予每位新用户的邀请名额仅有 5 个： - **冷启动阶段**：最初仅向约 200 名亲朋好友开放，次日自然微增至 205、210 人，每天仅有一两个人转赠邀请码。 - **复利加速阶段**：日增速逐渐从 1%~2% 上升到 3%~4%。当用户数突破数千人后，部分用户开始自发在社交网络上晒出惊艳的实用案例，增速一跃冲到 6%~9%。 - **当前爆炸式裂变**：在市场营销花费为 0 美元的情况下，Instinct 当前保持着每天 10% 到 11% 的环比复合增长（day over day）。 这种裂变引发了一系列奇特的社会现象：每天有近 10% 的现有用户自愿交出手中极其宝贵的 5 个名额之一来拉新人。许多人甚至不好意思直接开口向身边的朋友要邀请码，生怕自己“挤不进对方最核心的五人朋友圈”，转而羞怯地发邮件向 Noah 私下讨要；另外一些人则以“我还手里攥着 3 个名额没送出去”在社交圈炫耀。在二手电商平台 eBay 上，一枚 Instinct 的激活邀请码甚至被炒到了 300 美元的高价。 ### `56:58` 指数级计算资源的预测困境与采购杠杆 在狂飙突进的增长背后，Noah 透露自己现在有整整 40% 的时间都饱受底层算力采购与供应链问题的折磨。传统的消费级互联网产品（如早期的 Instagram 或 Facebook）在用户翻倍时，应对的主要是网络 I/O 与请求量的扩展；而作为一款免费的个人 Agent，Instinct 的算力消耗随用户规模直接挂钩，并且在 10% 的日复合增速下，所需算力差不多每周就要翻一倍。 ```mermaid flowchart LR accTitle: 采购周期与指数级增长的时间错配困境 accDescr: 展示算力采购的数月前置交付周期与每周翻倍的实际需求之间的错配矛盾 subgraph Demand["业务端：算力指数膨胀"] A["日增 10%"] --> B["算力需求每周翻倍 (2X)"] B --> C["三周内需求翻 5X 到 10X"] end subgraph Supply["供应链：硬硬件交付滞后"] D["采购订单下达"] --> E["漫长的供应链交付期 (3-4 个月)"] E --> F["算力机架真正上线"] end subgraph Risk["决策风险与杠杆"] G{"该提前订购多少倍？"} G -->|"采购过少"| H["算力瞬间见顶，80% 用户被宕机"] G -->|"现货应急补单"| I["面临现货市场 3 到 4 倍的暴利溢价"] G -->|"提前订购 100M 用户量"| J["承担千万级资金沉没与过载风险"] end Demand -.-> Risk Supply -.-> Risk ``` 这种采购决策的残酷之处在于算力供应链具有长达数个月的前置交付周期（lead time），根本无法做到“按需次日即买”： - **采购规模的两难**：如果仅提前按当前 2 倍算力采购，一周内就会被耗尽；如果按 5 倍采购，撑不过三周；若直接放大 10 倍杠杆去提前订购，资金压力和风险巨大，且依然会在几周内告罄。 - **现货市场的惩罚**：如果预测不足，临时在二级现货市场抢购算力，将被供应商加收 3 到 4 倍的极端溢价。 - **长期复利的惊人规模**：即便假设当前的日复合增长率放缓到 5% 至 8%，连续复利 3 到 4 个月（即下一批算力真正上线的交付等待期），平台也将滚雪球般膨胀到 1 亿（100 million）用户规模。 这就逼迫创始人不得不面临一个高危的抉择：“我们现在到底要不要冒着算错就要承受 3~4 倍惩罚的巨大风险，硬着头皮去提前采购支撑 1 亿用户的算力？”这成了 Noah 眼下每天都在推演的最沉重课题。 ### `01:00:00` 推理成本与底层架构优化 当谈及每位用户使用 Instinct 每天或每周在模型推理（inference）上消耗的具体成本时，Noah Shinn 指出，团队目前最核心的竞争优势之一，就在于彻底摸清了如何高效交付产品服务。无论是在严苛的 A/B 测试、内部基准评估，还是用户留存与活跃度追踪中，Instinct 都能交付媲美顶尖前沿模型（如即将到来的 Opus 5 级别能力）的推理表现，但其综合成本却被压缩到了极低、极具承受力的水平。目前 Instinct 面向所有受邀用户完全免费，尽管 Noah 暂未对外做出商业上的终身免费承诺，但他坦言自己的个人愿景就是实现“让每个人终身免费使用这款产品”。 这种成本奇迹依赖于底层基础设施架构的精细化重构。Noah 解释称，如果直接调用主流模型供应商的标准前沿 API，开发者就不得不为每一个请求支付统一且昂贵的“一刀切”溢价。然而在 Instinct 的实际运行中，海量的主动式（proactive）任务并不需要毫秒级的超低延迟响应——许多后台分析任务即便耗时数分钟甚至数小时完成也完全合理。通过针对具体工作负载定制部署架构，Instinct 将大量吞吐转为批处理工作模式。在相同的算力底座上，特定的负载形态能带来 3 倍、5 倍乃至 8 倍的能效提升。正是通过在各个微小环节拼凑出“这里优化 30%、那里提升 5 倍、别处再压降 10%”，层层复利最终转化为极其低廉的推理成本。 ```mermaid flowchart TD accTitle: Instinct 推理负载分流架构 accDescr: 展示即时响应与后台主动批处理的算力分流机制 subgraph UserInteraction["用户交互层"] Req["用户输入 / 紧急交互"] BgTask["后台上下文感知任务"] end subgraph RoutingLayer["负载调度与形态重塑"] RealtimeRoute["实时管道 (极低延迟)"] BatchRoute["异步批处理管道 (容忍数分钟至数小时)"] end subgraph ComputeEngine["算力部署层"] FrontierCompute["前沿算力即时响应"] OptimizedCompute["高能效定制部署<br/>(能效提升 3x - 8x)"] end Req --> RealtimeRoute --> FrontierCompute BgTask --> BatchRoute --> OptimizedCompute ``` - **推理成本控制**：通过精细化定制推理部署架构，摆脱了公有 API 昂贵的固定定价模式。 - **能效复利效应**：借助批处理和对非即时性任务的降级处理，在相同算力下达成数倍的吞吐效率。 - **产品定价目标**：目前维持全免费模式，远期目标是探索让全球大众终身免费获取 AI 助理服务的路径。 --- ### `01:02:30` 主动型 Agent 带来的算力需求爆发 关于 Instinct 未来若达到 10 亿用户体量将对全球算力产生多大冲击，Noah 明确表示，其规模将远超以往所有人的预估。此前以代码生成（CodeGen）为代表的爆款 AI 产品，本质上依然依赖“用户提问—模型生成—用户检查—再次提示”的交互飞轮；尽管这类产品也会在后台消耗大量 Token，但其底层逻辑依然是“被动响应型”的。相比之下，Instinct 是一套具备“自主苏醒与休眠（wake up and sleep）”机制的原生主动型架构。它全天候在后台推演用户的现实处境，例如监测到日程冲突或出行延误时，主动预判是否该提前叫一辆 Uber，或者在网约车抵达后而用户未出现时提供应急协调。 Noah 强调，真正的主动性意味着“算力需求呈数量级（orders of magnitude）膨胀”。在原生主动型架构中，用户显式发起的交互只是整个系统中极小的冰山一角，海量的算力都被消耗在无形的后台观察与推理中。例如，Instinct 可能会因为掌握了用户的作息习惯，在清晨 6 点自动唤醒，扫描用户的日程、邮件与航班动态以确保当天的一切井然有序；如果一切正常，它会选择静默沉睡，绝不在清晨打扰用户；而到了下午 4 点，如果察觉到某项对用户极具价值的事项即将发生，它又会主动苏醒去执行任务并适时通知。Noah 总结道，为了给用户提供这种细致入微的关怀与兜底，模型在暗处消耗的 Token 数量才刚刚揭开冰山一角。 - **行为范式跃迁**：从传统的“提示词驱动交互（Prompt-driven）”进化为“自发苏醒/休眠的主动型决策（Proactive Agent）”。 - **算力结构逆转**：显式对话只占总 Token 消耗的极小比例，绝大多数算力用于全天候上下文监测与后台批处理预演。 - **精细化服务代价**：越是追求对潜在意外的周密预判和对用户的零噪音打扰，背后模型推演所需的计算资源就越庞大。 --- ### `01:04:55` 对竞品 Muse 的看法与市场渗透真相 当被问及对同赛道产品 Muse 的看法时，Noah 表达了客观的赞赏，认为那是一款非常有意思且路径截然不同的产品。他分析称，即便两家在部分底层架构思路上有相似之处，但核心切入点有着本质差异。Instinct 将绝大部分精力倾注在极其细腻的“软性特质（soft qualities）”上——完全站在真实人类用户的立场去思考什么信息才是关键的、如何让交互结果一目了然且毫无理解负担，而不是强行向用户灌输一套全新的复杂独立软件或全新操作界面。 谈及竞争格局，Noah 直言自己“极少把时间花在琢磨竞争对手上”。他给出了一个非常具象的观察视角：“你现在走出办公室，去附近的咖啡馆坐坐，看看咖啡馆里有几个人在真正以他们想象或希望的方式使用 AI？我可以断言，几乎没有。”即使是在全球科技创新的前沿城市，AI 对普通人真实生活的日常渗透依然停留在极为初级的表层，更不用说其他国家和二三线城市。Noah 认为这依然是一片巨大、开放且充满生机的蛮荒绿洲，未来的竞争版图需要数月乃至数年才见分晓，当前唯一值得聚焦的就是全力打磨出终极的产品体验。 - **产品哲学分歧**：相较于推出全新界面或独立复杂 App，Instinct 更专注于降解认知负担与优化信息呈现的人性化体验。 - **现实渗透率反思**：即便在科技高地，大众日常深度使用 AI 的真实比例依然极低，市场远未进入存量竞争阶段。 - **研发专注度**：将注意力收拢于用户体验的纵深突破，而非追踪同行的一举一动。 --- ### `01:06:20` 早期失误与系统性主动防御架构 在探讨构建 Agent 过程中的失误与阴暗面时，Noah 强调，技术探索必然伴随着涌现属性（emergent properties）与错误的交织，但团队的核心工程哲学是“永远不要做被动应付者，而必须始终做主动防御者（never be reactive and to always be proactive）”。开发团队必须走在系统潜在风险的前面，预判每赋予 Agent 一项新能力时所带来的全新攻击面（surface areas）。 Noah 毫不避讳地承认，这正是 Instinct 从一开始就坚持采用严格邀请制的原因。在系统极其早期的版本中，由于尚未搭建起完善的防火墙机制与主动监控探针，底层基础设施确实存在防护漏洞与设计缺陷。但团队的做法绝非头痛医头式地“打补丁”，而是推倒重来，直接构建了一套全新的监控隔离架构，从系统层面根除这一类隐患。此外，Noah 再次重申了 Instinct 坚守的第一准则——“用户必须永远拥有数据的绝对控制权”。用户可以自主决定向 Agent 开放多少数据，且无论何时改变主意，都能瞬间撤销 Agent 对任何具体第三方服务的访问权限。 - **防御哲学**：拒绝事后修补的被动思维，采用预见攻击面演进的主动式系统防御。 - **架构重构而非补丁**：针对早期缺乏防火墙与监控探针的架构漏洞，建立系统级常态化阻断机制。 - **用户主权机制**：数据授权颗粒度完全由用户把控，提供单键即时撤销任何第三方 API 访问的最高权限。 --- ### `01:08:00` 数据特权与安全责任的未来哲学 面对未来可能掌握数以亿计用户超高精度个人上下文的现实，对安全和防黑客能力的责任要求被推向了极限。Noah 直言，安全保障是 Instinct 面对的“最重要的问题”，它必须从一开始就融入公司的底层价值观、团队基因以及产品架构的骨髓之中。 Noah 梳理了过去二十至三十年科技史的周期性规律：每当一项突破性的消费级技术雏形初现时，公众的第一反应往往伴随着巨大的社会阻力与安全疑虑，人们极易将“未知的新事物（different）”盲目等同于“不安全的事物（unsafe）”。回顾个人电脑、移动互联网乃至云计算的诞生，无一例外都经历过严苛的数据泄露审视与舆论反弹。当前 Agent 产业也正在经历这一必经的历史阵痛，但这也倒逼着核心团队必须以超越常规软件工程的警惕性，为即将到来的数据大交融筑起最坚固的护城河。 - **安全第一位性**：将用户上下文安全视作公司生死存亡维度的最高命题，而非外挂的功能特性。 - **历史认知规律**：技术革命初期，大众往往会将界面的“陌生与剧变”投射为对“系统安全性”的恐慌。 - **代际责任感**：伴随 Agent 洞察人类私密维度的不断加深，开发者必须主动承担起建立前瞻性防护壁垒的责任。 ### `01:10:00` 独立看门狗机制与对抗性安全防御 Noah Shinn 强调，Instinct 的核心原则之一是让用户始终完全掌控自己的数据，绝不能让用户感到失控。大语言模型（LLM）天生容易出现幻觉，但对于一个能够调用真实工具、操作真实世界的个人 Agent 来说，幻觉可能带来严重后果。因此，Instinct 没有依赖单一模型的“自省”，而是构建了一套系统性的解决方案：在任何操作执行、任何思考链（thinking trace）转化为工具调用（tool call）之前，都会经过一个独立于 Agent 本身奖励与激励系统的“看门狗”（Watchdog）系统进行审查与验证。 ```mermaid flowchart LR accTitle: 看门狗安全审查架构 accDescr: 展示从用户意图到工具执行过程中，看门狗系统独立于 Agent 进行前置拦截与校验的流程。 subgraph AgentSystem["Agent 核心系统"] A["用户指令/思考链"] --> B["生成工具调用候选"] end subgraph WatchdogSystem["独立看门狗审查"] B --> C{"独立看门狗过滤"} C -- "检测到幻觉/错误专有名词" --> D["拦截并修正"] C -- "安全验证通过" --> E["调用真实工具执行"] end ``` 这种机制能够精准识别因底层模型采样错误而凭空捏造的专有名词或危险行为。Noah 指出，捕获这类错误在事后看来很清晰，但关键在于事前硬化防御。Instinct 配备了世界顶尖的安全团队，每天都在积极进行红队测试，寻找越来越复杂的对抗性用例（adversarial cases）和边界情况。随着对抗训练的深入，底层模型面对此类攻击的鲁棒性正在持续快速提升，异常发生的概率也变得越来越低。 - **核心原则**：用户对数据拥有绝对控制权，杜绝失控感。 - **机制隔离**：审查与拦截逻辑与 Agent 的底层目标/激励机制彻底解耦，避免“自己批改自己试卷”的漏洞。 - **动态硬化**：安全团队持续输入对抗性样本，推动系统在边界场景下的防御能力自我进化。 --- ### `01:12:35` 界面坍缩：从屏幕打字到常开语音与即时生成 UI 在探讨 Instinct 与 iPhone 迭代路径的异同体验时，Noah 指出 Instinct 既不是传统的独立 App，也不是专属硬件设备，它目前直接寄生于 iMessage、WhatsApp 等现有通信通道之中。很多人认为未来会推出功能繁复的客户端，但 Noah 坚信界面的演进趋势不是变复杂，而是进一步走向极致的简单——最终所有软件界面都会坍缩成一个极其轻量、甚至无感知的交互入口。 如今已有大量核心用户超过 90% 的交互纯靠语音完成。Noah 分享了自己的日常使用习惯：他把 iPhone 侧边的操作按钮（Action Button）直接绑定到 Instinct，按住就能说话：“帮我跟 Patrick 打个招呼，两小时后发封邮件给他，你应该有他的邮箱地址。”说完松开即走，完全不需要解锁手机、点开信息软件、打字发送再等回复。 这种交互形态未来甚至不需要新硬件，一对普通的 AirPods 即可实现： - **常开监听与智能辨识**：系统能精准识别主人的声纹，知道主人何时是在对它说话、何时是在与现实他人交谈。 - **随身耳机伴侣**：在骑行、跑步或徒步时，Agent 会在耳边自然播报：“这份合同需要你审批”、“刚出来一条关键新闻，要点如下”、“某某发来求助，建议这样回复”。用户只需回答“排 15 分钟日历”、“这条不重要直接删掉”或“就按我说的答复他”，即可闭环处理事务。 - **动态按需生成 UI（Files 特性）**：当需要呈现复杂信息（如旅行路线规划、婚礼筹备方案）时，Instinct 不会强迫用户去用某种固定 App，而是当场动态生成完整的网页应用（Full Web Application）。过去 20 年软件行业开发新 App、收集反馈、打补丁的漫长生命周期，让消费者陷入了严重的“应用疲劳”（App Fatigue）；未来软件的形态将彻底瓦解，一切应用将归并为一个超级入口。 --- ### `01:15:00` 长期高阶目标的自主追求与原生企业运营 Noah 强调界面的极简绝不代表能力的退化，相反，Agent 的核心能力边界正在呈指数级扩张。最根本的质变在于：从“执行单一日常琐事”进化为“追求长周期、高阶目标（High-Level Objectives）”。 | 维度 | 传统日常任务模式 | 高阶目标导向模式 | |:--- |:--- |:--- | | **交互形式** | “帮我记录今天的跑步数据与卧推重量” | “接下来的 3-4 个月内，协助我减重/增肌 X 磅，并突破配速目标” | | **执行周期** | 单次、即时、响应式触发 | 持续数周至数月，具备长期时间跨度的自我校准与主动推进 | | **业务应用** | 员工在管理后台手动下达“订购 X 件商品” | 设定“确保此仓库库存不低于下限且不超过上限”，Agent 自主监控与补货 | | **自主权** | 机械执行给定的静态步骤 | 根据环境动态变化，自主选择并组合可用工具达成最终结果 | 甚至已经有小型企业完全“原生”运行在 Instinct 之上——整个公司的核心工作流与库存监控无需任何人工中介，由 Agent 在底层自主调用软硬件工具保持运转。用户只需定义目标边界，剩下的执行路径完全交由系统在无时间上限的周期内持续求解。 --- ### `01:17:40` 拒绝《她》的亲密依恋：做知进退的“高情商操盘手” 针对电影《她》（*Her*）中所描绘的人类与无所不能的 AI 产生浪漫羁绊的设定，Noah 明确表示 Instinct 坚决不走这条路线。团队完全不希望 Instinct 去刻意发展人际亲密关系或假装具备人类情感，更不追求那种油腔滑调（cheeky）的耍宝人设。 Noah 对 Instinct 的定位是一个“极度克制、安静且执行力拉满的高情商操盘手（socially aware operator）”： - **场景自适应力**：像一位精明练达的幕僚，无论走进哪间会议室、面对哪类对象，都能瞬间拿捏最得体的沟通分寸。 - **软性反馈学习**：它会敏锐捕捉到用户的负面反应。例如用户不喜欢某种措辞、或者某次因为甩了太长的一篇文本导致回复率暴跌，它会默默记下教训，未来自动精简汇报格式，绝不给用户造成认知负担。 - **命名“Instinct”的初心**：Noah 解释，“Instinct”（本能/直觉）这个名字代表着一种相互尊重与信任。它绝不是那种用户闲极无聊拿来捉弄、辱骂，或者只配用来倒垃圾干脏活的卑微小工具；它是一个极具创造力、专业过硬、知进退且值得托付后背的可靠行动者（competent actor），让用户在复杂世界中能够本能地依赖它。 ### `01:20:00` 品牌命名哲学与平台渠道策略：不做人格化，做无摩擦体验 在探讨产品的品牌形象时，Noah 解释了 Instinct 为何没有像许多虚拟助手那样取一个具象的人类名字。他认为，给 Agent 取人名是在刻意进行人格化包装，而 "Instinct"（本能/直觉）这个词甚至很难具象化为一个实体。这种命名给用户留出了一张“纯净的白纸”（fresh slate），用户在接触它之前不会带有任何预设的偏见或形象预期，而是纯粹在长期的实际交互与产品体验中，逐步建立起自己对 Instinct 这一品牌的感知与定义。 面对 Patrick 提出的平台依赖与“敌友关系”问题——比如 Meta 旗下的 WhatsApp 或苹果掌控的 iMessage 是否可能在某一天封杀 Instinct——Noah 表示自己并不将 Instinct 狭隘地定义为某一个特定生态的应用。从第一性原理出发，团队最关心的核心是：用户今天最信任、最习惯的日常交互界面究竟是什么？Instinct 的目标是在用户每天已经在使用的渠道中交付体验，而不是将自身死死绑定在 iMessage 或其他单一协议上。团队会根据用户最实际的习惯持续“变形”（shapeshift），始终保持极简、极易触达且“零摩擦”的交付状态。 - **去人格化命名**：避免设定特定人设，让用户没有先入为主的认知负担，完全由产品行为定义品牌。 - **渠道第一性原理**：不以“iMessage 应用”或“WhatsApp 应用”自居，而是无缝切入用户信任的既有通道。 - **形态适应性**：保持介质上的灵活性与动态转变能力，核心是交付无摩擦、即时可用的个人代理体验。 --- ### `01:22:20` 100 亿美元估值背后的资本逻辑：跳出订阅制的局部最优解 Patrick 提及 Instinct 最新一轮以约 100 亿美元估值筹集了近 10 亿美元资金，领投方涵盖了红杉资本（Sequoia）、Benchmark 以及 Coatue 等全球顶尖风投机构。谈及未来的资本需求以及选择投资人的考量，Noah 坦言自己非常幸运能与这些经历过数次类似重大技术范式转移的顶级伙伴同台。相比于数字本身，他更看重这笔资金背后反映出的市场对这种全新产品体验的极度渴望与真实需求。 ```mermaid flowchart TD accTitle: Instinct 商业模式与资本策略对比 accDescr: 对比传统 SaaS 订阅模式的局部最优与基于资本杠杆追求的交易抽成终局模式。 subgraph SubModel ["传统路线：局部最优解"] S1["设定高门槛订阅费<br/>(例如 $100/月)"] --> S2["锁定短期确定性现金流"] S2 --> S3["受限于高端付费人群<br/>难以实现数十亿人普及"] end subgraph ScaleModel ["Instinct 路线：风投杠杆与规模终局"] V1["引入顶级风投巨额资本<br/>(Sequoia / Benchmark / Coatue)"] --> V2["补贴先期高昂算力成本<br/>打造极低门槛体验"] V2 --> V3["跨越早期业务减速带<br/>验证跨行业交易量"] V3 --> V4["构建数十亿用户的交易网络<br/>逃逸出订阅制束缚"] end ``` Noah 直言 Instinct 是一项极度消耗资本（capital-intensive）的事业。为了将 Agent 推广给全球数十亿普通人并保持极低的使用门槛，团队必须在前期消化高昂的基础算力开销。他指出，团队本可以轻易选择一条更稳妥的捷径——例如设定每个月 100 美元的订阅费来换取短期的现金回报，但这会陷入传统软件的“局部最优解”（local optimum）。风险投资的真正价值，就在于提供承担计算后风险（calculated risk）的底气，用资本帮助公司跨越初期的“减速带”，去充分验证跨行业的真实交易规模与全新商业模式，最终逃离单纯依赖订阅费的狭窄路径。 - **顶级资本加持**：由 Sequoia、Benchmark、Coatue 共同支持，合伙人具备跨越多个技术周期的实战经验。 - **算力成本前置**：面向数十亿大众提供普惠服务意味着极高的算力投入，初期必须依赖资本做大规模前置投入。 - **跳出订阅陷阱**：放弃每月 100 美元的 SaaS 订阅现金流诱惑，利用风投资本去博取基于交易佣金与深度价值交换的全局最优解。 --- ### `01:24:10` 最善意之举与结语：在技术浪潮中沉淀的人际联结 在播客传统的闭幕问题中，Patrick 询问 Noah：“别人为你做过的最善意的一件事是什么？”（What's the kindest thing that anyone's ever done for you?） Noah 并没有列举具体的单一事件，而是从一个更高维度的视角进行了反思。他表示，能够在这个过程中与行业内外、工作及私人生活中的许多优秀伙伴建立起深厚的联系，并被这些充满善意和支持的人们所包围，这本身就是他所经历过的最大善意与幸运。在推动 Instinct 这场激进的技术探索时，这种来自周围人的包容、信任与周全考虑让他深怀感激，而他也希望能够将同等程度的善意与体贴回馈给身边的每一个人。Patrick 总结道，Noah 正在打造的不仅是一家迷人的公司，更可能是当今全球 AI 生态中最具想象力的一家企业。 ## Notable quotes > `04:11` 我知道这很打扰，但这篇文档必须在下午 3:00 前完成签署，现在已经是 2:55 了，您能在收件箱里处理一下吗？如果不方便，我也可以重新发一封邮件把它顶到最前面。 — Noah Shinn > `10:00` 顺便提醒一下，Patrick 正在尝试检索某某方面的信息。 — Noah Shinn > `15:00` 在未来几年内，互联网将被彻底重写，绝大多数人类与软件交互的方式将发生翻天覆地的变化。 — Noah Shinn > `20:00` 嘿，我今晚必须赶到纽约，十万火急。 — Noah Shinn > `23:25` 用户必须始终对自己的数据拥有绝对的控制权。 — Noah Shinn > `30:00` 如果你不为产品付费，那你本身就是产品。 — Noah Shinn > `40:00` 你饿了吗？需要我帮你点 5 天前吃过的那家外卖送到家吗？我知道你已经在 Uber 上了，预计二者会同时到家。 — Noah Shinn > `47:35` 放弃盲目追逐能力，将全部精力聚焦在‘可理解性’上。 — Noah Shinn > `56:58` 我们现在到底要不要冒着算错就要承受 3~4 倍惩罚的巨大风险，硬着头皮去提前采购支撑 1 亿用户的算力？ — Noah Shinn > `01:04:55` 你现在走出办公室，去附近的咖啡馆坐坐，看看咖啡馆里有几个人在真正以他们想象或希望的方式使用 AI？我可以断言，几乎没有。 — Noah Shinn > `01:12:35` 帮我跟 Patrick 打个招呼，两小时后发封邮件给他，你应该有他的邮箱地址。 — Noah Shinn ## Mentioned Noah Shinn, Instinct, Patrick O'Shaughnessy, ChatGPT, US Open, Spotify, Uber, Apple Pay, American Express, Stripe, Visa, MasterCard, Shopify, Amazon, Apple App Store, DoorDash, Uber Eats, Muse, Google, TikTok, Instagram, Snapchat, Opus 5, AirPods, iPhone, iMessage, WhatsApp, Sequoia Capital, Benchmark, Coatue, 电影《她》(Her) ## Video description My guest today is Noah Shinn. Noah is the founder of Instinct, a personal assistant that you text, call, or email the same way you would a person. It has its own phone and its own computer, and it can do almost anything on your behalf that you would do yourself online. Instinct is about a year old, still invite-only, and growing roughly 10% day over day without any marketing spend. We discuss the surprising ways people are already using Instinct, the trusted network that lets two people's agents coordinate directly, and what Noah has learned about how long it takes users to trust an agent with their credit card, inbox, and calendar. He explains why he refuses to build an ad model and why he believes a personal agent should never influence users against their own interests. We also cover the take rate model behind the more than $1 billion a year now flowing through the platform, how travel and food delivery businesses should prepare for a world of agents, why he spends 40% of his time thinking about compute, and why he believes all of software will eventually collapse into one simple interface. Please enjoy my conversation with Noah Shinn. For the full show notes, transcript, and links to mentioned content, check out the episode page ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠here⁠⁠⁠⁠⁠.&nbsp; ----- Become a Colossus member to get our quarterly print magazine and private audio experience, including exclusive profiles and early access to select episodes. Subscribe at ⁠colossus.com/subscribe⁠. ----- ⁠Ramp’s⁠ mission is to help companies manage their spend in a way that reduces expenses and frees up time for teams to work on more valuable projects. Go to⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ ⁠ramp.com/invest⁠⁠ to sign up for free and get a $250 welcome bonus. ----- Trusted by thousands of businesses, ⁠Vanta⁠ continuously monitors your security posture and streamlines audits so you can win enterprise deals and build customer trust without the traditional overhead. Invest Like the Best listeners get a special offer of $1,000 off Vanta when you go to ⁠vanta.com/invest⁠.&nbsp; ----- WorkOS⁠ is the infrastructure B2B and AI-native companies use to sell to enterprise. It covers everything enterprise security requires: SSO, SCIM, RBAC, Audit Logs, AI governance, and more. Trusted by 2,000+ fast-growing companies, including OpenAI, Anthropic, Cursor, and Vercel. ----- Rogo is the AI platform for finance. They're building agents for Wall Street that are trained to understand how bankers and investors actually do work: from diligence and modeling, to turning analysis into deliverables. To learn more, visit rogo.ai/invest. ----- ⁠Ridgeline⁠ has built a complete, real-time, modern operating system for investment managers. It handles trading, portfolio management, compliance, customer reporting, and much more through an all-in-one real-time cloud platform. Visit⁠ ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠ridgeline.ai⁠. ----- Timestamps: (00:00:00) The Most Exciting Software Race Ever (00:00:49) What Instinct Is (00:04:11) The Wildest Things Users Have Done With Instinct (00:07:15) The Instinct to Instinct Trusted Network (00:13:24) How Agents Will Reorder the Internet (00:14:44) Reinventing Restaurant Reservations (00:17:23) Travel and $1 Billion in Transaction Volume (00:21:00) How Users Learn to Trust an Agent (00:23:25) Keeping Sensitive Data Safe (00:26:08) Alignment and the Business Model (00:31:07) Take Rates and the Payments Stack (00:34:52) When Incumbents Push Back on Agents (00:40:46) How Businesses Should Prepare for Agents (00:44:43) Designing for Understandability (00:48:15) Taste Versus Data (00:50:03) Invite-Only Growth at 10% a Day (00:54:46) How Far Ahead to Buy Compute (00:56:58) The Cost to Serve Each User (00:59:24) How Much Compute Proactive Agents Will Need (01:02:30) Thoughts on Muse (01:03:59) Mistakes and Staying Proactive (01:05:34) The Future of Security (01:08:32) Where the Interface Goes Next (01:12:35) Agents That Pursue Goals (01:14:35) Personality and the Movie Her (01:16:27) Why It's Called Instinct (01:17:34) Allies, Enemies, and Distribution Channels (01:18:43) Raising Capital (01:20:34) The Kindest Thing [Source](https://traffic.megaphone.fm/CLS4620503535.mp3)

[4:11] Noah Shinn: 我知道这很打扰，但这篇文档必须在下午 3:00 前完成签署，现在已经是 2:55 了，您能在收件箱里处理一下吗？如果不方便，我也可以重新发一封邮件把它顶到最前面。 [10:00] Noah Shinn: 顺便提醒一下，Patrick 正在尝试检索某某方面的信息。 [15:00] Noah Shinn: 在未来几年内，互联网将被彻底重写，绝大多数人类与软件交互的方式将发生翻天覆地的变化。 [20:00] Noah Shinn: 嘿，我今晚必须赶到纽约，十万火急。 [23:25] Noah Shinn: 用户必须始终对自己的数据拥有绝对的控制权。 [30:00] Noah Shinn: 如果你不为产品付费，那你本身就是产品。 [40:00] Noah Shinn: 你饿了吗？需要我帮你点 5 天前吃过的那家外卖送到家吗？我知道你已经在 Uber 上了，预计二者会同时到家。 [47:35] Noah Shinn: 放弃盲目追逐能力，将全部精力聚焦在‘可理解性’上。 [56:58] Noah Shinn: 我们现在到底要不要冒着算错就要承受 3~4 倍惩罚的巨大风险，硬着头皮去提前采购支撑 1 亿用户的算力？ [64:55] Noah Shinn: 你现在走出办公室，去附近的咖啡馆坐坐，看看咖啡馆里有几个人在真正以他们想象或希望的方式使用 AI？我可以断言，几乎没有。 [72:35] Noah Shinn: 帮我跟 Patrick 打个招呼，两小时后发封邮件给他，你应该有他的邮箱地址。

[0:49] Instinct: I don't want to bother you too much, but you need to sign this document by 3 p.m. and it's 2:55 right now [10:00] Instinct: Hey, by the way, Patrick's looking for this type of information. [14:44] Noah Shinn: the internet's going to be rewritten. [20:00] Noah Shinn: Hey, I need to be in New York tonight. [36:28] Noah Shinn: Hey, I want my usual from this restaurant, [40:00] Noah Shinn: Are you hungry? Do you want the same thing you ordered five days ago? I can send it to your house so it arrives when you get there. [40:00] Please order that. [62:30] Noah Shinn: Instinct has the ability to wake up and to sleep at any moment in time during the day... Maybe Instinct wakes up at 6 a.m. because it knows that you wake up at 7 a.m. And then it goes and scans everything, makes sure that everything's ready for the day for you. And then it realizes, 'Hey, actually, now is not a great time,' like it should just go to sleep or do this one thing in the background and not notify the user. [65:00] Noah Shinn: You walk outside, you go to the nearby cafe, and you think about how many people within that cafe are actually using AI in the way that they imagine that they would or the way that they want to be using it, and I would say very little... You go over to other countries or other cities, and it's certainly lesser of the case. [71:15] Noah Shinn: Say hi to Patrick in two hours from now. I think you have his email address; you can go send him an email. [75:45] Make sure this inventory level never drops below X and never exceeds Y. [77:30] Noah Shinn: socially aware operator [83:45] Patrick O'Shaughnessy: What's the kindest thing that anyone's ever done for you?

[0:00] Ramp is the only platform built to make your finance team leaner, faster, and better, saving businesses 5% annually on average so you can stay focused on growth. Ramp just opened for business in the United Kingdom, so if you're running a business in the UK, you can now use Ramp's AI-powered finance platform to manage cards, expenses, bills, approvals, and accounting all in one place. I run my business on Ramp, and so should you. Learn more at ramp.com slash invest. purpose-built for financial institutions serving tens of thousands of bankers and investors at hundreds of leading firms worldwide. Rogo's AI agents autonomously execute large chunk of your firm's workflows. They can screen deals, draft sims, run buyer outreach, and do diligence on data rooms, all with full security and compliance handled. Every action feeds Rogo intelligence, the firm's context layer, turning years of deal experience into a secure, governed institutional more, visit rogo.com slash invest. OpenAI, Cursor, and Vercel all have something in common. They all use WorkOS. To achieve enterprise adoption at scale, you have to deliver on core capabilities like SSO, SCIM, RBAC, and audit logs. Instead of spending months building these mission-critical capabilities yourself, you can just use WorkOS APIs to gain all of them on day zero. Visit workos.com to get started. I can't help but wonder about the sort of grand game that is afoot here. When you really step back and think about what is happening right now, like what is the game, what is going on here right now? Like if you had to sum it up, the thing you're in the middle of. You're going up against the biggest players in the world. The window to do it is actually on the order of months. The compute required to it is over $0.00 just this year alone. And you have no distribution advantage, but you have this massively growing viral product. This is probably the most exciting software race ever. And the outcome is like trillions of dollars. Noah, I know this is the first time that you're talking about the business in this long form like this. I'm incredibly excited to ask you all about it. It seems like we're in a moment... with personal agents that is very similar and arguably probably bigger. than what we saw with code generation a year or so ago. Maybe frame this up for us to begin. Like, how are you thinking about the impact that of course your product, but these agents are going to have on the world. Why is this the area that you've chosen to dedicate yourself to completely? I think. Out of this will come a new way that most people on the planet interact with technology broadly. I run Instinct. It's a new company. that I started about a year ago. So it's still a new company. and we're building a personal assistant. I'm not going to spice it up because it's really just a personal assistant. And I think that the thing that has really enabled us to gain quite a bit of traction early and quite a bit of excitement early is it just works like it works in the way that I would say we all like we meaning I am also a user. We all have really been waiting for AI to to act for us. So, I would say one interesting thing is that This is not like any other... you know, transformational consumer app experience where it's the founder coming on and saying, you know, I have this new vision for the world where this is going to, you know, Trust me on this and in a few years from now, you guys will all get it. This is a very different moment of building product because you and I, we all already have this idea of what AI should act like. We've already had that idea since 2023 when we all started the brainstorm, when we started to use ChatGPT for the first time. You know, that agent that is able to help you with effectively anything in your everyday life, whether it's something new and creative that you want to do, whether it's something that you traditionally spend several hours on and it takes a long time to develop. It's just everyday intelligence. be with you, to act with you. It's meant to feel like a very simple interface. So when I mean simple interface, what I mean is, um, We actually don't even have an application. Like this is not a new app. This is not a new tool. It's a new experience. So, you know, it has a phone and a computer so you can text it. You can call it. It can actually call you too. So I don't know if you've experienced this yet. I have. Oh, you have. I think it's only called me about three times so far in the several months that I've been using it. It's funny if there's something genuinely pressing and has a deadline, it will call you and it'll call and it'll say, I don't want to bother you too much, but you need to sign this document by 3 p.m. and it's 2.55 right now. Can you play? [5:00] do that it's in your inbox I can I can even send you another email to push it up to the top. So it's very socially intelligent and very socially aware. You can email it, it has its own email address. I don't want to confuse the simple interface with limited ability because the whole nature of it is that there shouldn't be any new application needed to be able to interact with AI. It should have the social intelligence and awareness to be able to act with you, just like we do with other people. It's really meant to be the simplest possible experience. Yet the capabilities are, you know, it has a computer so it can do quite literally anything that you might want to do or that you do yourself on the Internet. Tell us the couple of craziest stories that come to mind for what just things that have happened because of instinct. I ask this question because you're not programming to do any one thing. You're programming with a phone and the computer to be able to do anything. The US Open example, I think, is one that everyone's familiar with this couple wanted the they showed up on the Jumbotron and they wanted like the video footage of it and somehow it went out and figured this out. and brought it back to them, which is kind of crazy. What are your favorite couple examples of just wild things that have happened because of Instinct? For those that like online shopping, there's actually a group of people that are sharing this and making this a more recurring use case, which is... scanning every item or every piece of clothing in their wardrobe so they go in the wardrobe and scan every single item you know every every top and bottom and sock and you know whatever um shoes and and then they would scan themselves to like their their face and their body and their proportions and And then they would go try on clothes and they would have it plan out their week in terms of what to what to wear. So every week with their existing wardrobe, you know, what shoes to wear, what what, you know, what top to wear with this certain thing and this certain version of it. And then it would it would not send it as here are the items you should wear like a bullet point list, but it would be them. wearing it. and showing like this is what today will look like. But not only that, it's also with shopping too. So they'll take those same capabilities, but then say, go shop on the internet. They come up with thousands of outfits, so many different places like this one. Here are the pieces from the top to bottom and you know and then and then they can just say order so it'll come in with a thousand different options then they'll scan you know keep in mind they're looking at themselves wearing wearing the clothing right and then they'll be able to point to like that one I really want can you send that to my house. And then not only that, but they'll say, every day, can you actually just come up with like three creative new outfits from head to toe? And then send those to me. And then they're able to, with the click of a button, actually just... you know get get that new outfit a lot of like goal oriented sort of use cases to like a lot of here's a certain personal finance goal that I have like I want to save this much money by this month and it's actively working with them to be able to serve to do that or users connect like their bank accounts to to instinct and and then it can scan over all the transactions and subscriptions that they've had and then surface that to to the user and say like do you really actually use this? And then they're like, oh, no, I didn't even know that. And then it can go in the email and unsubscribe from it. Yes, yes, it can go. And, you know, these popular consumer like fintech products where they'll just like surface it to you and then tell you, like, by the way, you should cancel these. This will go end to end, so they go into the site, sign in. If there's some sort of confirmation that goes to their email, or it has access to the email, so they use that, and they sign in, they go to the very end, they cancel the subscription, and then it just sends back... the number of what it saved you. Like, hey, by the way, you're saving $2,000 this month because of these things. Man, any product that's dependent on consumer laziness or inertia is toast, huh? Like toast. Yes. That's good. Good for the consumer. Can you tell us about the Instinct to Instinct Network? Oh, yes. The future of agents talking to each other. There are more practical scenarios where that is something new that I think will be transformed the way that certain people act in certain ways. I think it's only been around for a week or so, or I guess 10 days. Honestly, the purpose of this is not to think about like a new, new feature to build, but more about very busy working professionals. You know, they're taking meetings all day long and they're constantly meeting with like new people and existing people. You know, the process of getting a meeting on the and so laborious you know you text like hey you want to meet at this time and then they say no that doesn't work like but these times work and you say oh but i'm traveling so maybe this time works and i'll be in this time zone though and there's just so much back and forth there however if it was possible for you know the two users who were using instinct probably you know also like brainstorming these times that they're available with instinct anyways on the other on two ends are able to just communicate directly like hey the goal is just to find the time right we don't need to play this back and forth game and nobody's really playing games here we're just trying to A very canonical use case, actually. I think the way that you can find time on a calendar now is just so different, where you just say your intention. I want to meet with this person, ideally by the end of the week. [10:00] or during these periods of time. find me time, it will then coordinate with the other person's instinct to be able to find like when time is available or not, and then eventually put something on the calendar. I think the key unlock here, we called it a trusted person network. And The key is that you should only be connected with your trusted people. And there are some interesting social dynamics that are coming out into play that I've been learning about, where it's designed such that you only bring in people who are trusted people, who are not going to maliciously try to find out what your calendar is or what's in your email. But also not only that, we provide all the controls to be able to give different levels of access to different people. So... You know, a lot of spouses actually coordinate their instincts because they can generally just share everything. So that might just be, hey, this this is my spouse. Just share any anything with them. Right. Maybe with with certain colleagues. Right. It's hey, my work calendar is available. my certain parts of my inbox if they're asking for documentation or something like that are available. But everything else, let's just keep that off limits and it will be able to create create that. When you think about this, there's almost like these network effects that are built. through these like nodes in a network with these different edges with different weights too right it's not like a friend graph where you're just saying like i'm connected to you and that's it right it's it's i'm connect i'm connected to you i trust you i have also um you know configured it in this way such that there's varying levels of access so it's this interesting um like network effect almost almost being created and different social dynamics that are being um also explored and created as well there are some interesting cases i've learned about where if if somebody violates your um whatever your trust within the network there's an implication on the on that relationship itself right so so let's say if you and i connected right And then I said, hey, Patrick, I trust you. Let's connect on the platform so that it's easier for the next, maybe the next time to put time on the calendar. And then you use that and you then start digging for certain data in a certain area. First of all, maybe I only gave you calendar access, right? But you're digging for data in a certain area. And then my instinct texts me like, hey, by the way, Patrick's looking for this type of information. Now there's kind of almost like a trust broken between the two of us. All of these little social dynamics that come into play, we really thought about how to... Curate this network such that you do get the benefit, right, to schedule time and make plans with others and other things like that. But then also use some of the existing rules in society to be able to enforce both rules in terms of social, cultural norms and also of technical limitations in terms of what is visible and not visible. we're really going to see just like an entirely new order, aren't we? It's going to be fascinating. The number of things I can think of Let's just take, you know, with my wife or something. This would make life so much easier in so many different ways so quickly. Even little things like, you know, you were trying to find this house earlier, like, If I could just say to my instinct, hey, there's no one nearby. Like, where is he? And you give me a one-day permission or something like this. There's just so many ways to imagine it being useful. It's pretty wild. With the trusted network, we find a lot of very interesting cases, or example, use cases that are shared, because it is the one part about the platform that is, I would say, if every other area of the platform, the user generally has a good feeling of what it should act like, here is just something new and creative. For example, I heard the other day of, you know, when you're scheduling plans with friends or with like a spouse or something like that, it's always an active act, right? It's like, okay, every Thursday, maybe I'll like think about something on like the Saturday night about what I should do. And so it's always this active act. And then you like text the group chat, like, hey, do you want to do this? And then always somebody is like not willing to do it. And then eventually like falls apart, right? And that happens like every time. And then finally, when you get people together, that's when you have that, that, you know, you're able to actually, you know, attend that event or whatever it might be. There's like a And every week, just brainstorm something creative that is interesting, that we all like, and actually just... make it a different experience every week. But then it'll go through that every week, and not only coordinate when everybody's available, and what their interests are, and if they like certain things or not like certain things, or what shows are available, or what concerts, you know, based on their music taste, 'cause all of their Spotify history connected, and all, you know, that same friend group, they had, it was like a group of like six, They had one person order an Uber. But then that Uber to go around to all the six people and coordinate like, hey, we'll pick you up in 10 minutes. And then the other person will pick you up in five minutes from here. And I found the optimal route to go and pick up all the people. And now they're all just sharing like one Uber. So it's actually like very economically friendly. There's like all these random examples there of people discovering like new use cases. When you can finally remove the barriers of a lot of the rote social communication here and just make it like remove all of the logistical burden. how simple the explanation of this is, that it's easiest to literally just pretend you have a superhuman person. [15:00] person with a phone, a computer, an email address, they can call you. just like dealing with a person and that's it. And That makes me wonder how you think about everyone in the world having one of these having an instinct having an agent and how that will reorder things. Code has been incredibly exciting to watch, of course. Everyone listening has had their own version of a magical experience. with making something. But not that many people were software engineers before this. It's a relatively small sect of people that have made code historically. Maybe now it's going to be way more. But this just seems like a different new market, new paradigm. How do you think this will start to reorder now? The world, the internet, commerce... etc. What's the future of interfaces, I guess, is another way of saying it. I'll break this down in, I would say, like two parts. Just so that we don't sound like, you know, complete, you know, abstract visionaries and saying that we're going to reinvent the Internet. Although, I'll be clear, I do think that in the coming years... I think the internet's going to be rewritten. I think that the way that that most people on the planet interact with software is going to be very very different. To break it down though, I think in the short and medium term, famously, maybe reservations is one where, you know, like booking a reservation. Right. So historically, how is that done? Well, the human has to go onto the site and to go to whatever restaurant that they care about and then click through the site and then put their, you know, two people this time and tried to get it right and then if they don't get it they're just too late and it's first come first serve but then now when you have an agent you can theoretically just have the agent check every five seconds. You know, not just across that one restaurant that you really care about, that's your favorite restaurant you can't get in. Sure, that's one case, right? But what if it's doing that across every site in the world, for every restaurant in the world, in every major city? right? And now you can see how the agent is able to exploit just something simple like restaurant reservations. So if I go into that a little bit deeper, we have some exciting partnerships to be released in the future. But We are sort of reinventing what the reservation system is just in this one category because it's important to our users. Um, which is actually better for both sides. So if you take restaurant reservations from scratch, from the user's end, they want the reservation for the significant life event that they might have. It's a birthday, it's an anniversary, it's a friend coming in from out of town. And on the restaurant's end, they just want interesting people, they want special occasions, they want to cater to those rather than maybe the local who keeps taking up a table once a night and there's no sort of special event there. And traditionally, because of the way that the internet has worked, at least for reservations, it's been first come first serve. So anyone that comes in, no matter how important or how not important it might be, whoever gets in first gets the reservation. But what if you had the ability for the agent to be able to communicate on both sides, communicate, hey, this is important. This is actually the person's spouse's birthday, the 30th birthday, a very special event is coming up. And the restaurant, too, can say, let's actually prioritize birthdays or major decade birthdays or major anniversaries or holidays. these certain people, I don't know, whatever it might be. Right. So now with the ability, with all of the logistical burden of having to describe exactly what your case is and what's happening and why it's important. Now we have the ability to perfectly match what does the restaurant want? What does the user want? and that will result in more special events for the users, being able to actually have a spot in the restaurant. And then on the restaurant's end, for them to be able to have a much better audience of people who, you know, it's all these different special events or special occasions. So I don't mean to go so deep on restaurant reservations. I just mean... That's just one example of a traditional model that is changing. I think the same case with... a lot of the major travel agencies, honestly. 50% of transaction volume that goes through our platform is due to travel. Honestly, we're still running like an invite only program and it's quite early for us now, but we're approaching over a billion dollars a year in transaction volume through the platform. On a small user base. On a very small user base. On a very small user base, there's already over a billion dollars a year transacting. 50% of that is travel alone. Travel agencies that that we might all use today, or we meaning the people who are not on instinct yet use today to or other things like that. Well, what is the benefit of that interface, right? It's well, they properly unify from so many different services and so many different hotel chains and and airlines and just presented in a nice way to the user when now they can just click on an option, check out immediately through there and just see it in their email. It's a great experience. But I think there's an even better experience through through instinct or other similar interfaces where you can just say, hey, I need to be, you know, I'm using a voice recording again because I'm trying to. [20:00] Describe how easy it is. Hey, I need to be in New York tonight. This is immediate. And that's it. Right. Well, what does that entail? That means that, you know, first, where is the user right now, right? Maybe you're in San Francisco, maybe you're in LA, you're somewhere. Instinct can... you know, if you want, it can see your location. So it'll say, okay, you're in San Francisco, you need to be in New York. Let me first find all the options, right? But not only that, all of the nuance too. What's the person's airline, preferred airline that they'd like to fly at? What's the preferred seat type? In what class do you want to sit in? Do you want to be in an aisle seat or window seat or the middle seat? Etc. What food options do you want to be delivered to you as well? What credit card would you like to use? Okay, and then it books that. That's the flight. then it moves on to the hotel. Yeah, presumably it learns about you constantly, so your preferences are stored. That is the nice thing. If you tell it one thing, You know, I prefer this. This is my style. This is what I like. This is where I like to stay. One is that when you're staying anywhere, anywhere in the future, it will be able to use that memory and be able to, you know, now make it a much easier experience for you in the future to be able to book that according to exactly what you want. But not only that, but it'll be able to take that general taste, that general preference set and actually experience it. extrapolate that to anything else that you might want to book. So things just become very easy as you start to give it more preferences over time. So anyways, just to close out this example, you know, it'll find the ideal hotel. Maybe you've already stayed there, you know, the last seven times. And so it's very easy in that way. It'll book it end to end, line it up with your calendar. So it'll put the flight, you know, the flight in, the Uber ride that you need to take to get to the airport, the Uber ride to get back from the airport down to the, you know, to the hotel. And then all of the events that you might want to do there. Just take a step back, you know, because I talked so much about what this what is happening under the hood. The user really just recorded a voice recording saying. I need to be in New York tonight, right? And everything else is solved. So that's what I mean when I say that meet the users where they are. Where is the delightful experience? Well, this is a new delightful experience that I believe is going to transform even the travel agency alone. And so for these businesses that are traditionally make a lot of money from providing standardized interfaces, what happens when a new standardized interface comes into play? that is just that much easier, right? What does that mean? And I'm not saying that we're in the business of trying to disrupt those businesses. I think they do provide quite a bit of value from the data that they've collected over time and the networks that they have. It'll be a collaboration over the next... year, a couple of years to be able to redefine what that industry acts like. with $1,000 off at banta.com slash invest. Ridgeline is the first end-to-end system of record with embedded AI for investment management firms running portfolio accounting, reconciliation, reporting, trading, and compliance on one unified platform. Firms are moving off legacy technology and onto Ridgeline because of how far ahead Ridgeline's AI features are compared to anything else in investment management software. I've been hearing from a lot of investment managers about AI, where to even start, and others convinced they can build their own order management system over a weekend. The reality is that running an investment firm will always require governance, controls, and a single source of truth for your data. And no amount of AI enthusiasm changes that requirement. Ridgeline is built on exactly that foundation, which is why I believe that the firms that come out ahead in the AI era will be the ones running on Ridgeline's unified platform. If you're serious about your firm's AI strategy, Ridgeline should be part of that conversation, at ridgeline.ai. What has it been like getting people and learning what it takes to get people to trust their agent, their thing. I would love to go fairly deep here. We were talking about this this past week. And I love the examples that you've given about moments where people are showing trust and data you started to gather and lessons you're starting to learn about the importance of trusting this thing. I'd love you to talk about the tradeoff between. privacy and effectiveness of these agents. Obviously, like the more context, the more passwords, the more everything you give it, the more it can do. And I think people are constantly running that trade off in their head when they're interfacing with the thing. Talk about that. Talk about trust, what you've learned so far. There's an interesting data flywheel or sort of like chicken and egg problem here, which is, the more data that you give to it. The more proactive that it can be, the more sympathetic to your situation that it can be, the more that it has to just act and be more useful to you. Right. What we find in the data is actually that. It takes actually several weeks to build trust. And I actually don't have a problem. [25:00] at that because one of the core principles on our end is the user should always feel in control of their data. They should always be in control of their data. They should share data with instinct at the rate at which they feel comfortable with. And if they want to take it back, they can certainly take it back. And what we find is over several weeks, I believe I was looking at the numbers the other day, three weeks in, there's a 40% chance that the user has shared a personal credit card with Instinct. That's 40% of the user base that's including, you know, some proportion of that user base probably turned in that moment too, right? Or at least turned before that point. So 40% of the user base... The reason why I'm sharing this is because time to first credit card or time to first account password or time to first a sensitive piece of information, these are proxies for trust. And we actually highly value this and we take it very seriously. So 40% of the user base, three weeks in, are sharing a credit card. Well, there's something there, right? There's something there. No pun intended. It's like, it's the... It's like instinctual to go to it and, you know, with whatever you need. And so through that process over the first, I guess, three weeks, the user learns to build trust. And then from there, it's just a snowballing effect. We actually find that when users connect at least one piece of sensitive information to instinct and really trust instinct. there's like an 80% retention rate, 80% if you share one piece of entry. Wow, it's crazy for consumer technology. How should people out there that want to try this but have a natural reticence? to trust, not instinct in particular, but just anything. any AI agent. with all of its information. Like how... how should they think about the actual risk of doing this. And how would you reassure them that you've built your technology in such a way? that the odds of something bad happening, if they do trust you with very sensitive stuff, are really low or close to zero. There's like a way to break this problem down into like two main parts. One is just the storage of sensitive information. There are many other businesses, there are many other products that also deal with this problem of, you know, they have accesses, sensitive information and what are they doing proactively to make sure that that is you know isolated and and locked down and then there's a second area which is the new problems that we need to solve right the new surface areas or the new capabilities that we need to be aware of the first case is this is a tractable problem it's just very hard work and attention and care that you need to put into to making sure you know a sense of data uh shared is is you know is safe and and that the user is in full control over that data There's that second category, which I'll spend more time talking about, which is, you know, this is the first time that an agent has been able to have, you know, I said within three weeks, 40% of the user base. is giving instinct access to a credit card autonomously, right? To be able to purchase, you know, theoretically anywhere. I don't know the numbers on email and what proportion of users connect in email, but, you know, you can imagine... full access to an email inbox and into a calendar. So there's a lot of service area here. There are systems that we've put in place that are detached from the, from instinct, the agent architecture itself. that are put in place to be proactive about these things and to sort of like decouple the risk, if I were to say. So, for example, um, Any piece of content, any piece of text, anything, any form of media that comes in that might be consumed by instinct, goes through what we call these firewalls, which can intercept, that can reject, that can, you know, block malicious pieces of content coming in and, you know, hitting instinct and, you know, trying to convince, you know, instinct to do something. There is also for every action that instinct might take or for every thought that it might have, that is being actively monitored by a system that is decoupled from instinct itself, which is able to pause it, intercept it, to approve or disapprove of what might happen next before it takes the action. I mean, those are just two pieces put in place, but there's just so much more. There's so much more under the hood. that is put in place to enable the agent to be as capable as possible, but also safe and trustworthy. I have this question around... Um, I don't have a better word than alignment. And I know alignment is a very loaded word in AI. I don't mean... humanity scale alignment, I mean an agent aligned with me personally. If I think about other agents I've hired, like employees, I pay the money and therefore I trust them to have my interest at heart. It won't be perfect. But they don't have some other alternative incentive stream that guides their behavior. They're guided by their employment. How do you think about the business model vis-a-vis alignment. you could, well, you'll walk us through how you thought about it, but many approaches will be taken some will be paid some will be free the free ones will monetize in different ways can you walk us through this like decision tree of what the business model is or will be and, and how you arrived at that as the ideal conclusion for the agent. We don't want instinct to influence the user's behavior in a way that is not aligned with [30:00] what the user wants. That sounds very good on the surface level, but I want to call out how important that is because imagine a world in which instinct is generally smarter than the user. I'm talking more socially intelligent, more socially aware, like textbooks smarter as well, right? I think it would be a very dangerous world if instinct were influencing the user's behavior to purchase something that they don't want to purchase or to subscribe to something that they don't want, right? And using its intelligence to be able to convince that. Yeah, so I think that, you know, you look at most of the major consumer businesses today who are able to influence users' behavior against what they might want to do. You know, I'm talking the major platforms, whether that's Google or TikTok or Instagram or Snapchat or so many others, where you have this platform and users, you know, it's free for the users. But then there's so many instances in which, you know, paid ads are pushed to the user and they try to convince the user to then purchase. And, you know, by the very nature that they do purchase, it now becomes this game of these brands paying to you know convince users to purchase items that they may or may not want This is the idea that if you not paying you the product Exactly exactly As a programmer as a technologist I just don want to build that reality I think that's a very dangerous reality. And so we instinct should act on behalf of what the user wants. We take this very seriously when we're building product, when we're through the various research projects that we have, which are instinct is not unlike any other AI product. Instinct is not a task accomplisher. So what I mean by that is, you know, with most other AI products, you write a prompt and then it does the task and then it tells you what happened. Right. That seems good in theory. But what happens is that if the user is now asking for something that might not be well-intentioned, if you have a task accomplisher, it's just going to do that and listen to the user. Instinct follows higher-level objectives. So instinct will follow, you know, learn to build trust with the user. Learn to make the user genuinely feel safer with you. learn to watch over the user and have their back when things might be dropped or other things like that And when the user asks it to do something if it is well and well one way to communicate safety and trust is just to do the tasks. And so we end up just doing a superset of what most other AI products are able to do. But I think focusing on higher level objectives is very important here because it enables Instinct to be more robust to these edge cases. So we don't want to influence users' behavior in a way that is not aligned with what the user truly wants. Even if you look at the business model side of this too, there's over a billion dollars flowing through the platform now every year, and we're just getting started. And honestly, we're going at 10% day over day. So you imagine the transaction volume is also compounding at 10% day over day. So that's not just a billion dollars flowing through the platform. Now that's 1.1 tomorrow, and then that's like 1.2 something the next day, and 1.3 something the next day. It's still very early, but when I see transaction volume that is so high that flows through the platform, what I see is a very basic case. It's similar to an Apple Pay, or it's similar to an Amex, or any other platform that provides distribution to underlying services, and a great user experience for the user. Apple Pay is a great experience You can go anywhere and you can scan your card and the user doesn have to pay for it right The user is getting a free great experience and the merchants on the other end who are benefiting from the business are paying to be a part of that platform So I see a blanket transaction take rate being enforced across the platform, which is just us exchanging distribution for being able to serve products on behalf of merchants. Can you say a little bit more about how that runs into the existing world. So I can imagine the layers being you could be a card issuer. You could be something like Stripe. You could be Visa MasterCard. You know, there are sort of rails that have been built in the payments world that create convenience and reduce friction and take a vig as a percent of the transaction. And those are some amazing businesses, to be sure. How do you think about which of those are partners, which of those are potential things you would displace. How does that vision of a small take rate on the transaction volume on Instinct, because it's a free product, how does that slot into the existing world, do you think? Just to put that into context, those are very great businesses, and there's so many people along the side. You make one digital transaction, and there are 40 people along the line that make money on that. Just to take a step back, those 40 people making money at each line of the stack are really sharing, what, 2.5%, 2%? [35:00] It depends on where the transaction is coming from. So it's a very actually small piece of the pie and We're not primarily interested in whatever, like doing what Amex does best and have the network to do or even like what Stripe does on the Internet or some of the underlying payment infrastructure do because it's such a small piece of the pie. And I think that they're providing real value where I see the majority of the value just like thinking on the business side. is If you look at most of the other you know, major platforms in the world and what take rates that they're able to achieve by, you know, again, it's free for the user. It's a great free experience for the user and the merchant, you know, is now recognizing the distribution source. You have Shopify that I believe is like between two, two and a half and three percent. that are providing their services and taking some take rate from those businesses. You know, I'm not sure where Stripe is. I think it's maybe on the lower end of that. And then you have Amazon that has a great platform and taking upwards of 10 percent. Right. And then, of course, the premier, you have you have Apple where you make any purchase, which is 30 percent. Right. So I'm not saying that we're going to be 30 percent. I think it's very unrealistic for a lot of businesses. of it's a free experience for the user it's a free rate experience for the service that we're providing to all of the underlying providers I'm not focused on that on that to, you know, finding like 30 bits on the 2.5 with some partnership with some payment provider. I'm looking at the, you know, can we provide so much value that we're on the upper end scale of this? The reason why I think that this is possible is very practically, you know, 50 percent of the. transaction volume flowing through the platform is travel alone. And, you know, we all know the travel industry and some of the rates that we've kind of... Yeah, the OTA rates are high, yeah. It's very high for flights. You know, maybe it's on the lower scale, but, you know, how many single-digit percentages can you take off of a flight? Or for hotels, you know, some of these boutique hotels, you know, they're offering to pay up to over to 30% for every... you know, a transaction that you're able to deliver for them. And I'm not saying, again, we're going to be at 30%, but you can see the range. These are existing, you know, business models that, that, that we can boot shop off of in the early days, but then what would it be like to take that and extend it to it? Effectively every major industry. This is only this is only the case. We only have the the ability to do this if it is true that most digital behavior then moves to these new types of interfaces, which is, well, again, like no interface, right? But it is only the case if there's significant distribution power. So I think it's actually a quite intellectual or interesting intellectual question here about how this evolves over time. It seems to me like... there's going to be a serious... corporate. Agent War. That You've already seen this with Amazon and Muse. you've got all these established players with tremendous vested interests in relationship with customers. that this could both instinct on their agents, could really disrupt in a major way. What are you thinking about that? Like, how do you interface with the other... great services out there. I'll pick a random one. You know, I use Uber Eats a lot. Like I order from Uber Eats all the time. it's kind of a pain in the ass to click through the thing. I can imagine a lot better experience on a snappy WhatsApp connection with instinct or something saying, Hey, I want my usual from this restaurant. And that's it. Like there's nothing else. And that it's got its computer and it goes on to Uber Eats and it orders or whatever. At what point does Uber Eats not like that anymore? At what point do these great businesses that have been built up start to be adversarial against... agents, do you think? What do you think are the most likely sources of conflict and reconciliation? It's going to be really interesting to watch. So I'm serious when I say, I mean, you're alluding to this too, that I think most digital services or industries are going to be not disrupted, but just changed and transformed. And so I was pulling effective like every digital service or business or application from, you know, various different verticals and industries and sort of plotting it along the line of saying what proportion is like the user's attention, you know, experience on the product as, um, a proportion of revenue and then what what proportion of it is delivering the underlying service meaning if the user didn't use the app how what proportion of that transaction would be or what portion of their revenue is due to the underlying good being provided or service being provided. And so I think that your question is of that upper end of the scale where, let's say, the majority or some significant amount of revenue is due to attention on the application or advertisement or things like that. So yes, that's Uber. I actually don't know what Uber's number is. That's a lot of the restaurant or the food delivery services, That is the travel agencies that [40:00] That's even Amazon themselves with the upselling that they do on their platform. Okay, there's one simple way to look at this, which is that, okay, well, it's 70% ads, 30% good. And so therefore, you're going to slash the 70% and they're going to be a 30% business moving forward. There's another way to look at it, which is I would, you know, take any business. Let's take let's take Uber Eats or DoorDash, you know, as an example. They don't know the numbers. I don't know the numbers, but they don't know the numbers of as they reduce the number of clicks, you know, needed to check out to order food to your house or to order an Uber as they reduce that time to check out the. transaction volume increases because it's reduced friction to get the same underlying good. And, you know, we're taking something like instinct and making that friction to do anything almost zero. It is literally zero in the case that there is proactive behavior. For Uber, let's say for ride sharing, right? If Instinct has access to your calendar and is... you know, owns your calendar and is booking, you know, all of these events and knows where you need to be in person here and in person there. You know, what if instinct just always had a car lined up for you at, you know, you know every time where you need to be so that it's out of you know out of your mind to know oh like you know in five minutes I need to order this uber because I need to be in this place in 45 minutes and There might be traffic, so let me check the app to see how much traffic there is to see when I need to order the ride. Instead, what if it was just proactive behavior? A car will always be lined up, and you will never be late because it's going to calculate traffic. It's going to find these things. What would that do to Uber's business or any ride share business if now the default for existing riders that love Uber or Lyft or any other ride share business, What would it be like to have productivity now make the friction to experience that, or have access to that good or service to be nearly zero, I think that that will result in that 30% looking a lot, lot bigger. You know, you talk food delivery service, maybe you're coming home late. off of an airplane and you need to be at your house and you haven't eaten yet and it just texts you and it says you know are you hungry do you want the same thing that ordered yesterday or five days ago I can send it to your house if you really like it I know you're you know because it ordered the uber too I know you're in the uber and you'll get here at this time and then the That's great. Or thank you. Please order that. So with very little friction now, what will that do to the total transaction volume, or the amount of times that a user interacts with the business, I actually think it'll go up. So it's this interesting game. I think it's this interesting transition period between users spending a lot of time on apps, like painfully spending a lot of time on apps, and that being a monetizable surface because that's where the user's attention is. to... The user is actually interacting with the business even more, which is counterintuitive. Actually interacting with the business more because the friction to do so is actually just much less. So I think it'll be this interesting game. All these various industries are going to move in a different way. And we hope on our end, I think that the way to approach any big change like this is... is not to come in hot. You know, we're still, you know, we're a new company, we're just getting started. Not to come in hot and immediately just start disrupting, you know, certain businesses, but go to them and just say, Hey, this is what we think. you know, this is what we think your business looks like. You know, you guys certainly know what your business looks like. This is how users on instinct are interacting with your business now already. What can we do in collaboration to make that a better experience for both sides where it makes sense for your business? It makes sense for our business. And certainly it's just a much better experience for the end user. That is something that we're exploring and that we're learning across so many different industries right now. If you think about the... things that traditional companies could be doing now to prepare themselves for an agent-rich world. Let's just pretend, you know, half of Americans or something have an agent that's doing all the stuff you just described, which sounds incredible. magical and very democratizing. I think that's like maybe something to highlight is This is going to bring to everyone capabilities that have been rare or expensive. And I think that's just like a really cool feature of agents in general. We can come back to that. what kinds of businesses are going to thrive in that world? Like what should businesses think about doing? to prepare for that world, to be successful in it, do you think? I think it comes again back to that breakdown. That's why I was doing that exercise the other day, which is just, you know, where is your revenue coming from? What proportion of that is the user spending time in your application? What proportion of that is the user, you know, having access to the underlying service? And it's just very clear if you're If your business benefits from more transaction volume, not at the cost or with even at the cost of less time on the application. then you're going to be in a really great spot because Instinct is going to make it you know, a hundred times easier to do that And if you're in a spot where nearly 100% of your revenue is due to the user's attention, I think that [45:00] even in a lot of cases against the will of the user and what they want to do. And there's so many games that, or so many, I would say, malicious product building almost of trying to convince the user to, to you know, against their will to use the application more. You know, we're thinking about a lot of these, I would say, a lot of these social media companies that are, you know, the user doesn't want, they don't feel happy when they're on the app, right? They're, they're, unwillingly giving their time to the app and they can't get off and they keep scrolling, right? but it's because that underlying business is benefiting from the user's attention. It's almost like liberating for the user to be able to, you know, again, this is why it's so important for Instinct to act on, behalf of what's best for the user because it's able to deliver experiences like this, where you can actually liberate the user from being sucked into these infinite scrolling moments like that. I would say any blanket advice is probably just like not well thought out. I think it's a case by case basis and it's certainly different within different industries. I would say very practically, we're finding where, you know, there are early partners with with very innovative CEOs or other executives that are really thinking ahead. They're really thinking ahead and they're willing to be early partners. And what we're what we're discovering is is almost like a playbook for how every business, no matter where they exist along that risk curve, can kind of discover, honestly, what are the risks that they have a little bit of data to be able to work with? And then we can work together on finding ways where we can land in a happier spot for both sides. One of those things is you don't have to go all in, right? Like you don't have to say, let's just turn it on, you know. for... Let's just let's just go and now you see like 70% of your revenue going to zero and then now you're stuck in this odd place you can mitigate the risk, right? You can scale down the experiment, you can, you know, run A/B test to figure out, okay, if we enable this certain thing across, you know, 1% of users or something like that, and we find how they interact with the business, and honestly, Does the user like it more? Is it a more enjoyable experience for the brand side too? Does the transaction volume go up? Like does the willingness to buy the product or access the product itself? There's a lot of work in product discovery too, right? There's a lot of times the user doesn't know. They want to buy something, but they can't find it, right? And so does that actually increase, you know, the user being able to find exactly what they want. So I think doing scaled experiments here is actually a really great playbook to run because you can scale the risk accordingly. At the end of it, you know, you get the data, you know, it's proportional data, so you don't have to run it across your entire user base. And that's what we're finding is really working so far. Again, it's still very early, so we're still discovering this in real time. Before I ask more questions about the world reordering nature of personal agents, I'd love to take a little side quest in the conversation and talk about what it takes to do all this, to provide all this. I want to hear about... What's been hard about building this, the technology itself? I want to hear about compute. I think you said to me at some point that you spend a big chunk of your time just thinking about compute right now. And maybe that's different five years from now, but certainly in the moment when you're growing fast, this is a really important thing. I'd love to hear your thoughts on that. Talk us through what it's been like to build. instinct itself and the key key and hard things to do so maybe before we do that just because i'm remembering all of our conversations It would be helpful for you first to frame up what you want it to feel like. and why it's so important to you that it sort of has this, distinctive quality and performance before we talk about then how you deliver those things So maybe first just say a quick word on that. what you want Instinct to feel like as a product. I could talk all day about this because I think that is just so important. And honestly, as a product builder, it's, I think, a new muscle to flex and to build. But really thinking beyond capabilities is something I really want to push here, which is, I think, over the last three years, we've seen, you know, uh, all these different product launches and new products and saying, AI can now do this, or AI can now do that, and did you know that it can do this thing because of this small technical thing that happened under the hood? And I think the consumer is first fatigued by all of this. They don't know how to access it. And then second, I think we're missing the point. So early on, actually, one principle that... that we held was, let's not focus on capability, let's only focus on understandability. So how much does a user understand about what's happening? How much, what is their ability to predict what will happen when they ask this or when they do this or when they interact with it in this way? Honestly, I think that that is one of the major angles or factors that has led to you know, engagement numbers that are completely off the charts or the viral word of mouth growth that is happening at 10% a day. It's understandable. It just feels if I were to describe it for lack of a better word, it should just feel It should just feel good in some way, right? In the way that it communicates to you, the purpose of communicating or sending a text message is not just the meaning of the text itself, right? [50:00] it's down to underlying even what is the shape of the text message itself, right? And how will the user feel when they see that, right? If you see a big block of text, that requires the user to scroll a little bit to find what the next message is. versus Maybe you front load some of the information so that the user really gets it in the first 30% and then can optionally read the rest. You know, thinking about how the user might read or even, you know, like read patterns where they use. I don't know if you know about this. No, what's that? The way that most people scan, you know, big chunks of text, they might read. 80% of the first line and then maybe like 50% of the next line and then it tapers off so it looks like a flag. That's a consideration that should be made. If Instinct, it's very... It has the ability to think about things like this, right? Okay, the user is going to spend most of their time reading the first two lines. And then certainly the first part of each line too, right? So how does it craft its message to deliver in the lowest fatigue way or with the least amount of cognitive load placed onto the user? So there's a lot of consideration that's put into this in terms of product building. That is, there's quite a bit of work that we do on the infrastructure side to make it fast and make it affordable to serve and But this is another area that I think is just so important and it's going to be a differentiator. How much of that is your personal and the team's taste versus... them being the result of a quantitative type process, like iterative quantitative process of, okay, we did ABC testing of all, you know, Emoji reaction versus short versus long. And this is the thing that is optimal. Like how much of it is an optimization exercise that's data driven? versus your own sensibility and your team sensibility. Honestly, it's just all of the above, because, you know, when you're thinking about building evals or, you know, evaluation or other testing frameworks to be able to test for these like very soft qualities or these actually very long term qualities, right? Two or three weeks in, does a user trust instinct, right? How do you measure that? How do you run, you know, simulated evaluations to test? if the user is going to feel trust in two or three weeks from now, right? And actually, a lot of it is just staged rollouts over time. So I might come in and build a slightly different experience, and then I'll release it to myself, and then I'll play around with it for a little bit and see how I feel about it. And, you know, I'm just like very opinionated about these types of things. And then, you know, then if I feel comfortable with it, I'll send it out to the team. I'll say, hey, you guys should try this. We'll see how they feel about it. And then they'll send it out to, you know, our, you know, smaller early access group, and then they'll play around. with it and see how it feels. And then if we're confident there, if there's any tweaks that we need to make, we'll do that. and then we'll eventually roll it out to the general public. So I think this is very important because I think that Instinct is quite capable and is one of the most capable products out there. But I'm not saying that over the next couple months or years that others are going to come and deliver the same, seemingly same experience. But I think that this deep focus and priority on how the user feels and how to make it the most enjoyable experience beyond you know, the words that it's saying, just the way that it feels is just so important. one of the great things about the history of technology is this race between incumbents getting quality and innovation versus upstarts like you getting distribution. Obviously, you've built an incredible product, the feel of it. Like you said, it's the worst it'll ever be. How do you think about that challenge? Like your speed of scaling, what your ambition is for how to get big really, really quickly, really? Do you think this is a winner take most take all market? Like what will the market shape of agents be? I'm just really curious how you're thinking about "Okay, you've got this foothold, you're growing 10% day over day. You do that math, it gets really big really quickly. You need a lot of compute. It's a free product. Like, there's all these... It just... You're a very chill guy. It just seems like a stressful situation to be in, facing down how big this could get as quickly as it could get. Talk us through that many headed monster of a problem. Well, I think you just described... everything all in once in about 20 seconds there. And I think maybe if we think about the growth story so far, just to describe like where we're coming from, we are, I guess, like famously or infamously... serving an invite-only product, which is honestly not meant to be an exclusive thing. Although some users are treating it like that, but that's really not the intention. The goal here is to be able to, you know, I'm ambitious, I want to scale this thing as fast as possible, but I also want to do it in a responsible way that enables us to... not wake up one morning and have 10 times the number of users and then 80% of them actually can't talk to it because there's not enough compute. The interesting thing, both the benefit and also the detriment is, whoa, we first started out this program, we gave it to like 200 people. It was like, you know, close friends and family members, and we just said, like, go try it out. And then... Like the next day, like, say like five people came onto the platform because they just referred it to somebody. Oh, just to describe a little bit more. It's an invite-only platform. And every user will have [55:00] Five invites to be able to get. So just five invites. Um, And so we rolled it out, 200 people, next it was like 205, and the next it was like 210. So it's like, okay, cool, a couple people are sharing it to one person, right? But then that just started to accelerate. It was very odd. It was like, Okay, it was like not 1%, 2%, then it started to be like 3%, 4%. And then once we hit a couple thousand users, some people just started to share it online, too. Just like natively share, like, you know, some cool use cases that they had with it. And then that accelerated the growth. It actually started to turn into like 6%, 7%, 8%, 9%. And now I believe we're like 10% or 11% day over day. And just to call that out, it is, it's not... that we're doing some sort of creative marketing event like every you know we spent zero dollars on marketing so far it's not that we're doing something every single day to be able to support this this growth Every day, about 10% of the audience, or slightly less because you can refer multiple people, are making a decision to give up one of their five valuable invites to somebody else. but that is happening every single day at a 10% rate. So I think there's something very significant there. When I talk about strength of word of mouth, I think this is... It's honestly a little bit surprising. It's one of the strongest cases of word-of-mouth growth. of You know, I find all these stories of people saying, um, Hey, actually, like, they'll ask me and they'll actually be ashamed. They'll, like, email me and say, like... Hey, can I please get an invite? I think I have a friend that has it, but I don't know if I make it into his five friends. Right. Right. And then there are also people who are saying, who are bragging like, oh, I got these like, I got three invites left, right? And I'm just like holding on to it right now. So there's so much happening of people that want access to the product. But there's like these odd social games happening of people using their invites. I was seeing the other day, there were some invites that were selling on eBay too. I don't know if you saw this. It's like 300 bucks. People were buying these invites on eBay. And it's just to control the growth. So then, you know, the big question comes, you know, the thing that most people are asking right now is, Yeah, I mean, you have much bigger players that are able to distribute to, you know, a billion or two billion people on the planet, you know, immediately. Um, And maybe they might not be compounding naturally as fast, but they have such a great top of funnel distribution. And then you have you have us, you know, compounding at, you know, a very, very fast clip every day. But then... But, you know, we don't own a major service that has two or three billion people and able to distribute immediately that day. So it's an interesting question, like, where does the curve line up and where's the inflection point? There's another problem that comes with this too, which is that it's an interesting scaling problem. This is what is, I think, the core of the problem that I spent like 40% of my time just... I'm worrying about, which is it's not like the traditional other consumer products that grew very fast, where it's, you know, to double the number of users on, let's say, Instagram or Facebook or something, it would be, you know, this many number of other requests going through the platform. And like, yes, there's a scaling story there. And it's certainly hard infrastructure work. We also have that, to be fair. But what do you do when the underlying compute also needs to, you know, 10% day over day? We've been doing this for several, several weeks now. What does it mean when The amount of compute that you need access to is now doubling effectively every week. Do we buy compute? 2X... 2x of what we have right now. Well, we're going to consume that in a week, right? So then do you buy 5x? Well, we're going to consume that in less than three weeks, right? Dude, you buy 10x. So now it's like your 10x leverage, right? if you can even stomach what it's like to buy 10x ahead but then you're going to consume that and you know couple weeks from now, right? So that is the hard problem. It's thinking about how far ahead, how far ahead do you buy? It's going at a faster rate than cloud code or codex or some of those other applications where they also had, you know, reason about other similar exponential type of problems. The other subtlety here is that it's not like a certain business where when you double the number of users, you can buy like 2x more resources in order to power it. It's that the resource has a lead time of several months, right? So you can't just go out tomorrow and start, you know, buy a compute because honestly, you get taxed like 3 or 4x what it is. If you're wrong, you're wrong by 3 or 4x. Bye. But then now, so two or three months at 10, let's say it slows down, right? Let's say we don't actually do 10%, you know, for a very long sustainable period. Let's just say it's like 5% to 8%, right? But 5% to 8% compounding day over day for... three, four months, which is the lead time to get, you know, bring compute online. And that's even aggressive itself. That's 100 million users, right? So so. then do you buy compute for 100 million users? So those are the types of questions I'm wrestling with today. Which is just like, if you're wrong, you're very wrong. You get charged three or four X. - What about if you zoom in on the individual user and the cost to serve them? [60:00] On a day per day basis, like do you have a center per week? I don't know what the right metric is, like how much me using my instinct costs in inference. per day or something like this. Do you have a sense of like that scale? What is that scale? One thing that I think is to our advantage is we've figured out how to serve the product, which is when we run, no matter how you evaluate, whether it's A/B tests, whether it's internal evaluations, or it's, you know, tracking engagement across users that might be on one model or the other, we're able to deliver the same performance as, honestly, Opus 5, which is now, I guess, we're dating ourselves. um opus 5 is uh um you know like frontier level uh intelligence we're able to serve it's the same engagement rate the same um you know a b test performance it's the same internal evaluation performance but at a cost that is very, very low. It is actually very affordable. We're running this program. Every user has... It's Has has the product for free and our goal is really to deliver this product at an affordable at an I'm not going to commit to free for a lifetime for now, but that is my personal goal, to be able to deliver this product for free for everyone for a lifetime. It's just hard infrastructure work to, I don't know how deep we want to get on this, but just if I give one example, if you use Frontier APIs from some of the main providers, you are taking a blanket cost on a certain request. And for all of the requests that might be needed to power that product for that month. But there's a lot of different work that's happening through proactivity that is happening throughout the day that doesn't actually need to finish in hundreds of milliseconds. It needs to finish in, you know... minutes or even hours. There is a batch work that is consuming a lot of content that can be served with you know, deployment shapes that are 3x, 5x, 8x more efficient with the same underlying compute. So I think that when you customize these inference deployments to be able to, perfectly shape the data and the workloads that you're serving, you're able to find these 30% here, 5x there, 6x there, 10% here. And all of those compound to a rate where we were able to survey at a very low cost. How do you think about solving the bigger problem of how far ahead to buy? And if I think about this at true scale, like you get to scale of a billion users or something like this, like how much new compute demand do you think this will represent. I mean, it seems CodeGen's obviously been an enormous amount. but ground us in some sense of scale of like, whether it's per user or some way of understanding just like how much new compute this will require. I'm just saying it's going to be a lot because if you just think about, let's just ignore compute and just think about how many tokens are flowing through the platform. The products that are, you know, I would say breakout products of a couple of months ago, let's say in the code generation space were... where it requires the user to prompt it, and then it goes and runs for something, and then it comes back to the user, and then asks for something else, and then the user sends something again. And Sure, a lot of that is background work, and so there's quite a bit of tokens that are consumed there, but here it is. Instinct has the ability to wake up and to sleep at any moment in time during the day. That might sound a little odd, but its architecture is enabling it to do that so that it can really think about if you have a meeting that you're running late to and you need to order an Uber or if they ordered the Uber and the user's not showing up, I'll be able to help the user through those moments. So proactivity, I think, is going to continue to. to expand over time. There's this big build-out, this big compute build-out in the with the earlier breakout products in the AI space that are, just scratching the surface of productivity for our background work, here now we have something that is almost natively proactive. It is a smaller subset that is actually interactive, right? So I think that the amount of compute that is going to be needed is going to be, honestly, orders of magnitude more than what we thought that we needed. It's just with more productivity comes more care for the user, more time to think about certain things that could go wrong or not go wrong. There's just so much that's happening under the hood, you know, maybe instinct wakes up 6 a.m. because it knows that you wake up at, you know, 7 a.m. And then it goes and scans everything, make sure that, you know, everything's ready for the day for you. And then it realizes, hey, actually, now is not a great time. Like, it should just go to sleep or do this one thing in the background and not... not notify the user And then it realizes, oh, at 4:00 PM, there's something that's coming up. There's value that could be provided. Maybe the user doesn't even know how to interact with Instinct in that way. But Instinct thinks that it's well-meaning and it's something valuable to your user. So it might then just wake up at 4:00 PM and then do the task and contact the user and then the user will. [65:00] you know, could lean into it and do that task. You see how much work is going on in the background. That is, you know, with uh, you know, I would say these coding products, you don't really have that much background and productivity work happening. I don't mean to quote a number here. I'm just saying that the shape of... of the product and of the workload that's going to be run is, I think we're just scratching the surface in terms of how many tokens we need. What do you think of Muse? What do you think of the product? I think it's a great product. I was playing around with it for a little bit, and I think it's interesting. It's a different take, right? Because sure, you can say some of the underlying architecture might be similar, but I think it's fundamentally different. I think that the soft qualities that we were talking about earlier. of really thinking from the user standpoint about what's important and what's not important, and how to make this task easier and easier to read and other things like that, versus a new application and a new interface. And sure, maybe over time we have an application that also delivers on a different set of tasks, but I think it's a great product. I just spend very little of my time thinking about competition and other players in the space because I think... You walk outside, you go to the nearby cafe, and you think about how many people within that cafe are actually using AI in the way that they... imagine that they would or the way that they want to be using it and I would say very little right and that's and that's that's here right you know you go over to other countries or other cities and and um it's certainly lesser of the case so I think I think it's it's it's still an open space it's still an exciting time. I think it's an interesting game that's going to be played and rolled out over the next coming months and years. Um, so, uh, I'm just, I'm just focused on building the best product experience. What have been the blunder so far? Like what if what's gone wrong? What have you done about it? I'm sure more things will go. I mean, this is just going to be an explosion of emergent properties and mistakes, and the end is going to be really high. How do you think about things to guard against proactively, things to be reactive against. Like, yeah, tell us through the darker side of building this or the harder side of building this. I think it's important to never be reactive and to always be proactive, to look ahead for, you know, what new surface areas are being introduced and what new risks might come. And, um, I would say this is the reason why we're... Well, we ran this early access invite program from the start. which was, yes, an early version of the product did have, you know, it didn't have firewalls in place. It didn't have these active monitors in place and so many other pieces of infrastructure. that were meant to get ahead and be proactive about being able to provide that much control to the agent and existing system to be able to secure. So, yes, there's an early version of the product that had some of these qualities or some of these mistakes. And we addressed it. We went above and beyond and didn't just patch the problem. We built it. we built a different system to systematically solve these types of problems. So... I think the key here is, again, I said this earlier, the user should always be in control. The user should always be in control of their data. The user can share as much as they'd like or as little as they'd like. And if they ever change their mind at any point in time, they can always, you know, retract to access these certain services. It makes me wonder, what's the future of security? Even if you're the best in the world at this, which maybe you'll have to become. you're going to have so much information and context on so many people. And security is a big problem, I think, across the world. I mean, all these great hacking examples that we've studied now, it's wild what these things can do. The capabilities are going to get stronger and so on. Do you have like a general philosophy or I'm just curious for you to riff on the future of security and and. safety and guarding against, I mean, early in technology, revolutions of the past. There are these, always, there's these enormous hacks, there's enormous data breaches and so on. It's hard to imagine this won't happen in this technology revolution, too. How do you think about this and the responsibility of... providing safety. given how much you'll know about people? It's the most important problem. I think it comes to... building security and safety and and a mindset towards the, I don't know, some of the core values of the company and the people and the way that you build product. You know, this is topical because anytime, if you look in the past, the last, you know, 20, 30 years, when any new... sort of like breakout or new consumer experience has been revealed. There's always been, you know, immediate backlash of like, whoa, this is different and confusing different with being unsafe. And there's always this we track the last 20, 30 years, seeing some of that now. But also, I think just a step [70:00] core to the principles that you hold. You know, the user's The user is always in control of their data. They should never feel out of control. And then being proactive about the systems that you put in place to be able to get ahead of these types of things. If I call out one example, like a Who is Nation case that existed. that like First, language models hallucinate all the time, but with a product like this, you don't want a language model to be hallucinating. So we put in place a more systematic solution that will detect before an action is taken, before a thinking trace is executed as a tool call. before any action might be taken, it is validated and scrutinized by something that is decoupled from the same incentive system as the underlying agent. It's like a watchdog, yeah. Filter that's able to find, you know, hey, this proper noun was just generated out of thin air due to some sampling error in the underlying model. And so it's very easy in hindsight to be able to capture these. these types of mistakes. And these are hardened. So we have world-class security teams that are constantly working proactively to to find harder and harder adversarial cases, to try to find edge cases here and there. And it's becoming rarer and rarer over time. One subtle part about building a platform like this is that it's getting better over time. As we continue to do more adversarial testing, as we continue to be more creative about certain edge cases, these models are just getting better and better over time towards being robust to these types of attacks. On the other side of the ledger, everyone always shows that beautiful visual of each generation of the iPhone. better over time and more and more refined. What's that arc for you? It's so interesting because it's not an application. It's not a device. It's an interaction through existing communication channels, WhatsApp and iMessage, etc., What are the things that you're adding and envision adding over time that will make the platform and the product more powerful than it is today? You know, the product experience right now is very simple. Very, very simple. Simplicity is one key. um... that we focus on. But I actually think that it becomes even more simple over time, which is, we may, you know, roll out an application in the near future, but it's, I actually think that we're going to trend more towards a simpler interface, which is, do you even need to open the iMessage application and type in a certain, you know, a piece of content and send the message out and then look at what the response is after when it's done? We have a certain, there's a interact with instinct only through voice actually like more than 90 percent of the interaction of the messages that they send to instinct are primarily through voice where i even have this like action button you know the action button on your phone and it's paired to like i can click on the action button i can say like you know hey hey say say hi to you know patrick and in two hours from now i think you have his email address you can go send him an email and then I can just send and then it goes and sends. And I don't even now need to unlock my phone and then open the iMessage. application and type the message in you know you can even imagine with with a uh, real-time voice with voice recognition that knows what your voice sounds like. and has the discretion to know when you're addressing it and not addressing it where you could have let's say like an airpod and you know not a new type of airpod just like an airpod because it's good enough and um and it's just on and it's just on and you might be going on a hike or on a bike ride or on a walk on on a on a run or something like that and you're just like catching up like hey, this contract needs your review. Okay, I need to review that. Or hey, this news article just came in and here's the headline and here's the takeaway. This new project was released and you should take a look at this. And you're just like in your ear like, okay, got it, got it. Put that on my calendar. That's 15 minutes. Okay, yeah, that's not important. So go clear that part out. Oh, this person needs my help. I know the answer to that. So just tell them that it's this, right? It's going to be much, much easier over time. So I think we'll see. I think that there are long-term I think in the long term, I think that the interface has become, yeah, again, much, much simpler. But in the short and medium term, you know, maybe there are more expressive interfaces to be able to share. You know, we have like a files feature that enables Instinct to be able to send, you know, effectively entire sites like full web applications to, you know, show just much more, whether it's like a trip itinerary or... I don't know, like a wedding plan or something like that. I don't know, something that just requires much more and more creativity and more surface area. It can generate those on the fly and be able to show that to you. And I think that there's something interesting there where you look at the last 20 years of application building or product building, where anytime that you needed a new interface to be able to showcase some piece of information, you're going to be able to showcase some piece of information. you needed to well build an application for it and this is long software development lifecycle to be able to produce the application and ship it out to people and then and then you find like some feedback like hey I wish you look like [75:00] this and then you take it and you make the improvement and you ship out a new version right and this is on an order of like months Right. And then over time over the last 20 years, there have been all these applications built for so many different purposes, so many different things. And now the consumer is just fatigued, right? It's like, every time I needed something, like, is there an app that does that? Like, I don't need... I think that all of that collapses down in the future. I think that all of software is going to collapse down into, honestly, a single, very, very easy to use interface. Don't get that confused with capability being limited. I think capability is going to expand. What's the craziest thing you can imagine capabilities-wise? When we move away from more of these daily tasks, like, I want to do this, can you do it? And that friction is going to go down to zero. higher level objectives that are aligned with teaser where the user is able to describe not just hey, can you track this workout for me? I did this, and I performed here, and this is how I did. But more about Hey, in three or four months from now, can you, or over the next three or four months, can you work with me to make sure that I hit these certain goals? And actually, a lot of people are starting to experience that or are starting to discover that today of, you know, you want to, you know, gain this much, this many pounds. You want to lose this many pounds or you want to hit this certain mile time. Being able to specify objectives and goals and then be able to work with it to then achieve those goals and objectives. I see small businesses being run now that are native to instinct itself, meaning the entire business is now run on instinct. completely functioning on top of instinct. And even parts of it are fully autonomous too, where, you know, talk about high level objectives. It's in this certain area, make sure that this inventory level doesn't drop below this and doesn't go above this. And so now it's not, It's not saying, hey, please order this thing. It's not static, yeah. It's, you know, pursue this higher-level objective and use the tools and devices that you have available to be able to accomplish that. So I think that's the future. And I'm not, like... You know, there are so many cases of objectives that you could state and have it pursue for an unbound amount of time. that we could brainstorm here. But I think that's the meta level picture is it's the moving away from, you know, individual use cases over to higher level objectives is going to be where I think interaction will progress. it. to have a personality that's distinctive for each person. Like I'm thinking now of the movie Her, where there's this relationship that forms between this omniscient, omni-capable agent and the user. What do you think about that aspect? I think of instinct as this very like seamless thing. Almost quiet. extremely capable, reliable thing. but not as having a cheeky personality or something. How do you think about that component of... of the product. Well, you mentioned her, I guess, comically. It's certainly relationship building or anything in that area is not something that that we want instinct to do or to pursue or to develop with users. If I describe more about what it should act like and feel like and represent within the person's life, it's almost like that socially aware operator that knows no matter what room that they're in, what is the best for different people and what is the best interaction pattern for, and it learns that over time. So I think it's possibly one of the most customized, if you even call this an app, it's like one of the apps ever because it's able to not just You know, we're not just able to serve a different version of it. It's able to evolve over time. That's why we put so much care into like the software pieces, right? Pick up on when the user didn't like it being said in this certain way. Or if there's a lower response rate in this certain way, because it's just too much text or it's like, you know, they don't want to look into a huge file to be able to know what you're talking about. it's going to learn those over time and just become easier and more delightful for the user to use. So I think we're not trying to impose certain experiences onto users. I think we want to solve this more from a higher level. It's just ability to adapt to exactly what is the best communication style and task execution style. Why is it called instinct? Instinct I liked as a name for so many different reasons. I think the main thing is that Instinct should not be this, like, playful thing that you might like bully once in a while and that you like think you know downwards on and that it's like this this thing that only takes the dirty work off your plate and and I think it's this this new creative exciting you know competent actor that there's like mutual respect and trust and and they feel safe and and they trust that it has their back that it's intelligent that is competent that it's socially aware to know how to act in certain situations and it's it's a it's a bet but I like [80:00] that instinct is not you know named like I don't know someone's name or something like that to try to personify it into like a human it's more of I don't even know what an instinct is. I kind of develop my feel for what it is and for what the brand is through just using it. And again, towards really thinking about this feeling of what it should be like, I think that it's a benefit that every user is able to come into it with a fresh slate. There's no prior in their mind about what it looks like. or what its name is or anything like that. It's purely through the product experience itself. Who are your enemies and allies? Like I can imagine in WhatsApp or something, you could get shut off. That's a Facebook product and it's a key channel for you. iMessage is an Apple-controlled product like Who are your friends? Who are your enemies? How do you think about? There's some inevitable competitive realities here. I would focus less on like Instinct is an iMessage app or Instinct is a WhatsApp app or something like that. And more about just very grounded from first principles. What are the interfaces that users trust today that they're familiar with interacting with today? And how can we deliver the experience of Instinct through the channels that they already interact with every day? And just primarily thinking like we're not pinned to iMessage. iMessage itself. But certainly users that are comfortable with iMessage are very familiar with it. I think it's just more about just thinking about what is what is most practical for the user. And we're going to shapeshift as we see towards just delivering, again, that very simple... simple easily accessible, um... sort of like zero friction experience that it is today. I know your latest round is something like a billion dollars at roughly a $10 billion valuation. Some of the best investors in the world, Sequoia benchmark. KOTU as the leaders of the round. How do you think about the future capital needs of the business alongside this? And how did you pick the partners that you picked? I'm just very lucky to be able to work with some of the most supportive partners around the table who have been through, you know, different but similarly shaped, you know, technology transformations in the past. And I would focus less on the numbers and more about just on the demand in the space, the demand for a product like this or the demand for an experience like this. It's because it really is. of that product experience that I think that we've all really been waiting for. And so it's also capital intensive, right? We're a new company, right? We're trying to create something that's going to hopefully be distributed to billions of people on the planet every day and at a very affordable cost. When I say affordable, I mean, you know, compute costs are high and we're trying to deliver it in a very affordable way to our users. So it's just very capital intensive. There has to be some bootstrapping, right? There has to be some bootstrapping, especially if you think about the business model that we're chasing after. Because we could very easily just say, "Hey, it's going to be a certain subscription and everybody on the platform needs to pay $100 a month." And there are shorter term rewards that we can chase after. or we could raise a little bit of capital, use that towards, you know, what is venture capital meant for? It's to be able to, you know, take a certain calculated risk towards, you know, if there are small speed bumps that require a little bit of capital to get ahead of, to be able to prove, you know, transaction volume or to prove, you know, experience across certain industries. That's how you escape these local optimum of, you know, the subscription plan or other things like that. When I do this, I ask the same traditional closing question of everyone. What's the kindest thing that anyone's ever done for you? to surround myself and be able to build such great relationships with so many different people, whether it's in the industry or outside the industry and where I work or in my personal life that I'm very grateful to be around. And hopefully, certainly, I can also extend the same opportunity. same level of kindness and thoughtfulness towards, towards, um, towards them as well. So I would say, I know I'm not answering your question, it's more on a higher level. It's just the qualities of things that that I noticed that I really appreciate. Noah, you're building a fascinating company, maybe the most fascinating company of today's AI ecosystem in the world. Thanks so much for your time. Thank you so much. slash subscribe. Patrick O'Shaughnessy is the CEO of Positive Sum. All opinions expressed by Patrick and podcast guests are solely their own opinions and do not reflect the opinion of Positive Sum. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of Positive Sum may [85:00] positions in the securities discussed in this podcast. To learn more, visit PSUM.VC. Outro Music. You know how small advantages compound over time? That's true in investing and just as true in how you run your company. Your spending system is your capital allocation strategy. Ramp makes it smarter by default. Better data, better decisions, better economics over time. See how at ramp.com slash invest. As your business grows, Vanta scales with you, automating compliance and giving you a single source of truth for security and risk. Learn more at vanta.com slash invest. The best AI and software not in months. Visit workos.com to skip the unglamorous infrastructure work and focus on your product. Ridgeline is redefining asset management technology as a true partner, not just a software vendor. They've helped firms 5x in scale, enabling faster growth, smarter operations, and a competitive edge. Visit ridgelineapps.com to see what they can unlock for your firm. Every investment firm is unique and generic AI doesn't understand your process. Rogo does. It's an AI platform built specifically for Wall Street, connected to your real outputs. Check them out at rogo.ai slash invest.