Permission Inc. Unveils Agent to Monetize Consumer Data for AI

AI Needs Permission With Permission, Inc — Ai4

Ai4

Permission Inc. Chief Executive Officer Charlie Silver argues that mounting privacy regulations and polluted web scrapes will force AI developers to pay consumers directly for permissioned, high-quality data through automated agents.

AI-assisted summary of AI Needs Permission With Permission, Inc — original source. Check the original for full context.

Source published: 2025-08-27

Detailed notes

## Background Published on YouTube by the Ai4 channel on August 27, 2025, this conference presentation features Charlie Silver, Founder and CEO of Permission, Inc. According to the page description, the talk addresses how tightening regulations and mounting lawsuits are ending the era of unconsented, scraped AI data, making permissioned and auditable data an essential requirement for building trusted AI systems.

## TL;DR Charlie Silver demonstrates why training AI on unverified, scraped web data introduces severe risks from bot traffic, click fraud, and legal liability. He frames personal data as private property systematically exploited by Big Tech and introduces the Permission Agent alongside an Agent-to-Agent (A2A) architecture. By participating in opt-in value exchanges compensated with liquid ASK tokens, consumers can passively monetize their digital footprint while providing AI developers with high-integrity, verified, and legally compliant datasets.

## Atmosphere & style The presentation carries the pragmatic, self-assured energy of a veteran entrepreneur, blending historical dot-com anecdotes with sharp industry critiques and colorful pop-culture touchstones like Jerry Maguire and Michelin-starred chefs. Silver balances blunt skepticism toward Big Tech data practices and crypto speculation with genuine enthusiasm for consumer sovereignty and decentralized Web5 models. It feels like an engaging, no-nonsense executive pitch that seamlessly transitions from macro-level data dilemmas to immediate, practical technological solutions.

## Key takeaways - [00:04](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=4s) Explicit opt-in consent and permissioned data collection can deliver a 10x lift in engagement and response rates compared to standard marketing benchmarks. - [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) Training AI models on unverified web scrapes poses severe risks due to 30%–40%+ bot traffic and a $100 billion global click fraud industry. - [05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) High-grade, enterprise-ready AI relies on three foundational pillars: deep context, verified data provenance, and conscious, auditable consent. - [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) Big Tech monopolies have built multi-trillion-dollar valuations by extracting consumer behavioral data without paying royalties to the individuals who create it. - [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) Because complete digital privacy is an illusion in modern society, consumers must actively assert ownership over and monetize their digital property. - [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) Mounting regulatory frameworks like GDPR, CCPA, and numerous US state privacy laws are rapidly ending the era of unrestricted, unconsented data scraping. - [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) The Permission Agent is designed to act as a consumer's commercial advocate to negotiate data value and autonomously execute personalized tasks. - [14:11](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=851s) The digital economy is transitioning to an Agent-to-Agent (A2A) web where enterprise brand agents negotiate directly with consumer AI agents using opt-in value exchanges. - [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) Permission's native crypto token, ASK, converts traditionally closed, illiquid loyalty points into an open, liquid, tradable blockchain asset. - [21:01](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1261s) Web5 architecture and agentic browser extensions enable a decentralized internet where users passively earn rewards when brands query their permissioned data.

## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=4s) Twenty-Five Years in the Permission Business: From RealAge to Big Tech Data Extraction

Charlie Silver opens by introducing his quarter-century background in the permission data business, recalling the early dot-com venture he co-founded, RealAge, alongside Dr. Mehmet Oz. RealAge functioned as an interactive health risk assessment collecting hundreds of personalized biometric and lifestyle data points. After completing the evaluation, users were asked for permission to receive tailored recommendations and product information designed to lower their "real age." When users opted in—which roughly 60% chose to do—the resulting marketing response rates did not just improve marginally by three- to five-fold; they surged by 10x across the board, proving that explicit consent and shared data fundamentally transform user engagement.

Following the acquisition of RealAge by Hearst, Dr. Oz moved on to his television career on *The Oprah Winfrey Show* and later into politics, while Silver continued building startups. Over five years ago, Silver founded Permission, Inc. on a core thesis: consumer data holds immense financial value, yet Big Tech companies capture all of that upside without compensating the individuals who generate it. With the emergence of modern artificial intelligence, Silver argues that this vision of democratized, compensated data ownership is finally primed to operate at massive scale.

- RealAge collected hundreds of personal health data points per user, achieving a ~60% opt-in rate. - Permission-based marketing campaigns delivered a 10x lift in response rates compared to standard benchmarks. - RealAge was acquired by Hearst Corporation.

---

[02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) Garbage In, Garbage Out: Bot Traffic, Click Fraud, and the Michelin Chef Analogy

The classic computing adage "garbage in, garbage out" has never been more relevant than in the current race to train artificial intelligence models. While many development teams assume that scraping the open web yields rich datasets, Silver points out that scraped web data is overwhelmingly polluted. Between 30% and 40% of web traffic on the low end—and significantly more on the high end—is generated entirely by automated bots. Furthermore, organized internet click fraud has grown into a $100 billion global illicit enterprise, trailing only the illegal narcotics trade in revenue for organized crime. Teams building AI agents and large language models on unverified web scrapes are inadvertently basing their architectures on fabricated inputs.

To illustrate the danger of low-grade data inputs, Silver uses the analogy of a Michelin-starred superstar chef. Even if a culinary master possesses world-class technique, elite training, and flawless kitchen execution, poor-quality ingredients will inevitably ruin the final dish. Similarly, drawing on Walter Isaacson’s biographies of Steve Jobs and Elon Musk, Silver highlights how the world's most successful tech innovators obsess over internal components and unseen details. In artificial intelligence products, the foundational ingredient is data quality, and cutting corners on that ingredient degrades the entire output.

- Estimated bot traffic on the open web ranges from 30% to 40% on the low end to substantially higher on the high end. - Internet click fraud represents a $100 billion global illicit industry, ranking second only to narcotics trafficking in organized crime revenue.

```mermaid flowchart TD accTitle: Data Quality Comparison in AI Training

subgraph Scraped["Scraped Web Data Pipeline"] A["Scraped Web Traffic"] --> B["30-40%+ Bot Traffic & Click Fraud"] B --> C["Unverified AI Models & Fragile Agents"] end

subgraph Permissioned["Permissioned Data Pipeline"] D["Human-Consented Data"] --> E["Verified Provenance & Real Context"] E --> F["Auditable, High-Performance AI"] end ```

---

[05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) Context, Verification, and the Gold Standard of Consent

High-grade AI data relies on three foundational pillars: context, verification, and explicit consent. Contextual understanding remains a major hurdle for AI systems; for example, an algorithm must correctly differentiate whether the word "jaguar" refers to an apex predator in the jungle or a luxury automobile. Verification demands absolute certainty regarding data provenance—knowing exactly who generated the data, where it originated, and how it was recorded.

Consent represents the defining attribute of data integrity. Silver stresses that real consent cannot be buried deep inside convoluted terms of service agreements; it must be overt, conscious, and enthusiastic, likening it to accepting a marriage proposal. Without rigorous consent and auditable provenance, enterprise AI deployments face severe legal and operational liabilities: "If they don't know where the data is coming from, you're going to be this guy" facing regulatory scrutiny. Emulating the "gold standard" benchmarks used in healthcare and finance, AI datasets must combine human provenance, auditable records, and clear opt-in permissions.

---

[07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) Data as Property and the Illusion of Privacy

Silver frames personal data as private property that has been systematically expropriated by tech monopolies. Companies such as Google, Meta, Microsoft, and Amazon have achieved multi-trillion-dollar valuations primarily on the back of harvested consumer behavioral records. As a striking example of monetization speed, Amazon scaled an internal advertising business from zero to $80 billion in annual revenue in roughly six years, powered purely by proprietary consumer search, browse, and transaction histories.

``` +------------------------------------+--------------------------------------------------+ | Traditional Data Exploitation | Permission-Based Property Model | +------------------------------------+--------------------------------------------------+ | Data harvested without royalties | Individual retains direct ownership | | Monopolized by Big Tech platforms | Decentralized distribution and storage | | Concealed in opaque privacy policies| Explicit, conscious opt-in value exchange | | Commercial value captured by tech | Direct financial compensation to the user (ASK) | +------------------------------------+--------------------------------------------------+ ```

Addressing the popular phrase "data is the new oil," Silver argues that the current digital dynamic is equivalent to Exxon arriving at a private ranch in Texas, Oklahoma, or Montana, drilling for oil reserves, and extracting them without paying royalties to the landowner. Furthermore, true consumer privacy has become an illusion. Any individual who holds a credit card, pays a mortgage, subscribes to a magazine, votes, or carries a driver’s license is already having their data packaged and resold repeatedly. Because opting out of the modern economy is impossible without going completely off-grid without a phone or bank account, Silver argues that consumers should actively monetize their digital property rather than surrender it for free.

- Amazon grew its targeted advertising division into an $80 billion annual business in approximately six years. - The California Department of Motor Vehicles generates $80 million annually by selling driver's license and vehicle registration data to commercial marketers. - Major publication houses routinely generate their primary revenue streams by licensing subscriber lists.

---

[11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) Regulatory Pressures and Introducing the Permission Agent

The regulatory landscape is rapidly shifting against unconsented data scraping. While the European Union aggressively enforces GDPR across its member states and parts of Asia mandate strict privacy compliance, the United States is catching up. Despite legal challenges and delay maneuvers from companies like Meta regarding the California Consumer Privacy Act (CCPA), twelve additional US states have enacted comprehensive data privacy frameworks governing permissible consent and data processing.

Against this backdrop, Permission, Inc. issued a press release officially announcing the **Permission Agent**. Silver illustrates the agent's function using popular cultural archetypes: Hollywood power agent Ari Emanuel from *Entourage* and Jerry Maguire shouting "Show me the money!" The agent's core responsibility is to act on behalf of the consumer as a dedicated commercial advocate, extracting maximum monetary value for their data. Secondary functions mirror a digital "mini-me" or personal concierge—resembling Nia Vardalos's travel agent character in *My Big Fat Greek Wedding*—capable of executing personalized web tasks, booking hotels in New York, and shopping autonomously.

---

[14:11](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=851s) The Agent-to-Agent (A2A) Web and Opt-In Value Exchanges

The digital economy is migrating toward an agentic web governed by Agent-to-Agent ("A2A") communication protocols. Drawing on his participation in nationwide Google partner developer events, Silver explains that enterprise AI agents can now be configured and deployed within minutes. If a corporate brand like Nike launches an AI agent to identify and engage amateur tennis or golf players, that commercial agent must negotiate directly with consumer-side AI agents like the Permission Agent.

This automated interaction relies on the long-standing marketing concept of the "opt-in value exchange." Tracing the model back to S&H Green Stamps, Silver highlights successful corporate loyalty programs like Marriott Bonvoy, American Express, Southwest Airlines, United, Delta, and Starbucks. These enterprises excel because they respect the customer contract: in exchange for consumer engagement and zero-party data, the company returns tangible, predictable rewards and builds trust. Conversely, brands that treat loyalty programs inconsistently experience rapid consumer churn.

```mermaid flowchart LR accTitle: Agent-to-Agent (A2A) Value Exchange Architecture

subgraph Enterprise["Enterprise Layer"] EA["Brand AI Agent<br/>(e.g., Nike, Travel, Retail)"] end

subgraph Interaction["A2A Protocol & Value Exchange"] EA <-->|"Negotiates Query & Token Reward"| PA["Permission Agent<br/>(Consumer Mini-Me)"] end

subgraph Consumer["User Layer"] PA -->|"Passes Verified Data"| EA PA -->|"Deposits Liquid Tokens (ASK)"| UW["User Crypto Wallet"] end ```

---

[17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) ASK Token: Liquid Rewards and Crossing the Chasm into Mainstream Web5

Permission's economic engine is anchored by its native crypto token, **ASK**. Silver acknowledges that earlier efforts to scale the token faced regulatory hostility from previous US administrations, which made corporate advertisers hesitant to take on compliance risks. While Silver bluntly critiques the broader crypto market—stating that "90% of it is BS" and functions like speculative casino chips—he emphasizes that the authentic breakthrough of blockchain technology lies in Real World Assets (RWAs), as evidenced by Circle's public momentum with tokenized dollars.

``` Traditional Brand Loyalty Points vs. ASK Tokenized Rewards: • Traditional Points: Siloed within closed vendor ecosystems, illiquid, non-tradable, zero capital appreciation. • ASK Tokenized Rewards: Publicly liquid, transparent blockchain ledger, openly tradable across markets. ```

By tokenizing reward points through ASK, Permission converts closed-loop loyalty credits into an open, liquid, appreciating asset class. When integrated into the A2A economy, the Permission Agent enables average users to "cross the chasm"—moving emerging technology from early adopters to mainstream consumer adoption, as outlined in Geoffrey Moore's classic framework. Consumers configure the agent once, grant granular permissions around their geography, search intent, and browsing habits, and passively receive ASK tokens deposited directly into their digital wallets as brands bid for their data.

---

[21:01](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1261s) Web5 Decentralization and the Passive Earning Launch

Connecting Permission's roadmap to Jack Dorsey’s vision of "Web5," Silver describes the emergence of a truly decentralized web where data ownership returns to the individual rather than residing in centralized corporate silos across Meta, Google, Microsoft, and Amazon. The agentic web acts as the practical execution layer for Web5, giving consumers automated tooling to assert property rights over their digital footprints.

Silver concludes the presentation by reiterating the launch of the Permission Agent via browser extensions. The setup requires only seconds, operating silently in the background and alerting users whenever their anonymized, permissioned data has been purchased by brand agents. Silver invites attendees to test the software, visit Booth 640, and attend the post-session marketing networking reception in Room 203.

- Permission Agent operates as a lightweight browser-level tool executing background passive monetization. - Live demonstrations and inquiries hosted at conference Booth #640 and the Room 203 reception.

## Notable quotes > [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) garbage in, garbage out — Charlie Silver

> [05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) If they don't know where the data is coming from, you're going to be this guy — Charlie Silver

> [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) data is the new oil — Charlie Silver

> [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) Show me the money! — Charlie Silver

> [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) mini-me — Charlie Silver

> [14:11](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=851s) opt-in value exchange — Charlie Silver

> [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) 90% of it is BS — Charlie Silver

> [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) cross the chasm — Charlie Silver

> [21:01](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1261s) Web5 — Charlie Silver

## Mentioned Charlie Silver, Dr. Mehmet Oz, RealAge, Hearst, The Oprah Winfrey Show, Permission, Inc., Walter Isaacson, Steve Jobs, Elon Musk, Google, Meta, Microsoft, Amazon, Exxon, California Department of Motor Vehicles, GDPR, California Consumer Privacy Act (CCPA), Permission Agent, Ari Emanuel, Entourage, Jerry Maguire, Nia Vardalos, My Big Fat Greek Wedding, Nike, S&H Green Stamps, Marriott Bonvoy, American Express, Southwest Airlines, United Airlines, Delta Air Lines, Starbucks, ASK Token, Circle, Geoffrey Moore, Jack Dorsey, Web5

## Video description Charlie Silver, Founder/CEO, Permission, Inc.

Abstracts

AI systems are only as good as the data they rely on, and most of that data today is scraped, unverified, and unconsented. As lawsuits mount and regulations tighten, the era of unrestricted AI data is ending. For teams that require human data – whether for marketing, personalization, or agent-to-agent interactions – permissioned, auditable data is no longer optional.

This talk is a wake-up call for AI teams: permission isn’t just a competitive advantage; it’s imperative for trusted, compliant AI – and we'll explore what it takes to get it right.

[Source](https://www.youtube.com/watch?v=zdC7FGW5o7s)

中文简报

Permission Inc. 推出面向 AI 的消费者数据变现智能体 Permission Inc. 首席执行官 Charlie Silver 认为，日益严格的隐私监管和受污染的网络抓取数据，将迫使 AI 开发者通过自动化智能体直接向消费者付费，以获取经许可的高质量数据。 Charlie Silver 阐述了为何使用未经核实且抓取的网络数据训练 AI 会因机器人流量、点击欺诈和法律责任而带来严重风险。他将个人数据视为被科技巨头系统性剥削的私有财产，并推出了 Permission Agent 以及智能体对智能体（A2A）架构。通过参与以具备流动性的 ASK 代币作为报酬的主动选择性价值交换，消费者可以在为 AI 开发者提供高完整性、经核实验证且合法合规数据集的同时，被动实现自身数字足迹的变现。 Charlie Silver 指出，基于用户显式许可（opt-in）收集的数据能使营销响应率较行业基准实现 10 倍提升，表明明确授权能从根本上重构互动价值。 (0:04) 公开网络中有 30% 至 40% 以上的流量由自动化机器人生成，且全球网络点击欺诈规模已达 1000 亿美元，依赖未经核实的网络抓取训练 AI 模型将面临严重的输入污染风险。 (2:14) 随着 GDPR、CCPA 等隐私法规不断收紧，传统无限制抓取数据的时代正在终结；消费者的数据应被视为私有财产并主动进行商业化变现，而非任由科技巨头无偿提取。 (7:41) Permission, Inc. 推出 Permission Agent，旨在新兴的 Agent-to-Agent（A2A）网络中作为消费者的专属商业代表，协助用户自主授权数据并通过 ASK 代币获取流动性收益。 (11:14)

Reading summary

Charlie Silver, founder of Permission, Inc., argues that modern artificial intelligence faces a data-integrity crisis caused by scraping polluted web data, requiring an economic shift toward explicit consumer consent and direct compensation. Polluted Web Data and AI Integrity (2:14) Silver asserts that the computing principle of 'garbage in, garbage out' threatens artificial intelligence models trained on unverified web scraping. Automated bots generate between 30% and 40% of open web traffic on the low end, and organized click fraud has expanded into a $100 billion illicit industry. Using the analogy of a Michelin-starred chef whose culinary execution is ruined by substandard ingredients, Silver argues that cutting corners on raw data degrades AI output. To ensure enterprise-grade reliability, Silver outlines three prerequisites: contextual clarity, verified provenance, and overt consent. Much like rigorous data standards in healthcare and finance, AI inputs must be auditable and intentionally authorized by humans, rather than extracted via buried terms of service agreements. Personal Data as Expropriated Digital Property (7:41) Silver characterizes personal behavioral data as private property that has been systematically extracted by Big Tech monopolies without fair compensation. He compares this practice to an oil company drilling on private land without paying royalties to the owner, noting how platforms built multi-trillion-dollar valuations on harvested records—such as Amazon scaling its advertising unit to $80 billion in annual revenue in roughly six years. Simultaneously, Silver contends that true privacy in the modern economy is an illusion, citing examples such as the California Department of Motor Vehicles generating $80 million annually by selling driver records. Because individuals cannot realistically opt out of banking, driving, or commerce, Silver argues that users should actively monetize their digital footprints instead of surrendering them freely. Consent-Driven Marketing and Regulatory Shifts (0:04) Silver points to his experience co-founding RealAge, where obtaining a 60% opt-in rate from users across hundreds of personal health metrics produced a 10x surge in marketing response rates. Corporate loyalty models, from S&H Green Stamps to modern airline and hospitality programs, demonstrate that consumers willingly share zero-party data when offered clear, predictable benefits. This commercial dynamic is reinforced by accelerating legal pressures. Beyond the European Union's GDPR and Asian privacy mandates, twelve US states have enacted comprehensive privacy legislation following the California Consumer Privacy Act, narrowing the viability of unconsented data harvesting and forcing enterprises toward auditable opt-in models. Agent-to-Agent Interactions and Liquid Token Rewards (14:11) To operationalize consumer data rights, Permission, Inc. introduced the Permission Agent, an autonomous proxy designed to advocate commercially for individuals. In an emerging Agent-to-Agent (A2A) web, enterprise AI agents—such as brand tools deployed by retail or travel companies—must negotiate directly with consumer agents to access verified data, returning value back to the user. Rather than issuing siloed, illiquid brand loyalty points, Permission uses its native ASK token to deliver tradable crypto rewards into user wallets. While Silver criticizes roughly 90% of the cryptocurrency market as speculative, he maintains that tokenized Real World Assets provide the foundation for Jack Dorsey's 'Web5' vision, decentralizing data ownership away from corporate silos through background browser tools. Silver envisions a decentralized web where lightweight agent tools passively monetize personal data for everyday consumers. However, scaling this model depends on mainstream user adoption and enterprise willingness to integrate tokenized micro-incentives, particularly given past regulatory friction that made corporate advertisers cautious about cryptocurrency integration.

精炼阅读摘要

Permission, Inc. 创始人 Charlie Silver 认为，人工智能模型正面临公开网络抓取数据污染与法律合规风险，唯有建立基于明确授权的数据产权与补偿机制，才能保障高质量数据供给并打破科技巨头对消费者数据价值的垄断。网络抓取污染与高质量训练数据的三要素 (2:14) 公开网络抓取正严重污染人工智能训练集。Silver 指出，公开网络流量中至少有 30% 到 40% 来自自动化机器人，高位比例甚至更甚；同时，全球互联网点击欺诈已演变为规模达 1000 亿美元的非法产业，在有组织犯罪收入中仅次于非法毒品交易。依赖未经核实的数据源如同顶级厨师使用劣质食材烹饪，必然导致模型输出降级。 Silver 主张，高标准的人工智能数据必须依托语境（Context）、验证（Verification）和明确同意（Consent）三大支柱。算法需要准确理解多义词的真实语境，企业必须掌握可审计的数据来源以规避监管问责，而数据授权必须如同接受求婚般明确、自觉且积极，不能掩埋在冗长晦涩的用户服务条款中。数据资产确权与隐私幻觉的现实困境 (7:41) 个人数据本质上是私有财产，但长期被大型科技公司无偿占有。Google、Meta、Microsoft 和 Amazon 凭借用户行为数据累积了数万亿美元市值，例如 Amazon 仅用约六年时间，便将其基于搜索、浏览和交易历史的广告业务扩展至年营收 800 亿美元。Silver 将这种单向索取比作石油公司在私人牧场开采原油却不支付特许权使用费。与此同时，绝对的个人隐私在现代社会已不复存在。拥有信用卡、房贷、驾照或投票权的用户，其数据早已被反复打包转售，例如加州机动车辆管理局（DMV）每年通过向商业营销人员出售驾照及车辆登记数据获利 8000 万美元。Silver 认为，鉴于脱离现代经济网络并不现实，消费者更理性的策略是主张数字财产所有权并实现主动变现。监管趋严与代理对代理（A2A）价值交换 (11:14) 全球数据监管正在收紧。欧盟严格执行通用数据保护条例（GDPR），美国除加州消费者隐私法案（CCPA）外，已有 12 个州出台了规范数据许可与处理的综合隐私法律框架。在此背景下，商业逻辑正转向自动化的人工智能代理交互。企业端 AI 代理（如品牌营销系统）未来将直接与代表用户的消费者端 AI 代理对话，形成代理对代理（Agent-to-Agent, A2A）网络协议。这种交互依托于成熟的自愿价值交换模式，类似于万豪旅享家、美国运通或星巴克等企业的会员忠诚度计划。企业为获取精准的零方数据向用户提供可预期的实在回报，而 Permission Agent 则扮演消费者的商业经纪人与数字管家，在执行个性化网络任务的同时，代表用户与品牌代理谈判以争取最高数据补偿。 ASK 代币经济学与主流消费落地机制 (17:21) 为了让数据价值高效流转，Permission 引入了原生代币 ASK。Silver 坦言加密货币市场有 90% 充斥着类似投机筹码的虚假项目，且该代币早期推广曾遭遇美国前任政府监管阻力导致广告主态度谨慎；但区块链技术的真正突破在于现实世界资产（RWA）的代币化应用。与封闭、不可转让且缺乏升值空间的传统品牌积分相比，ASK 代币提供了流动性、链上透明度以及公开交易能力。在 Permission 设想的去中心化网络（Web5）架构下，普通消费者只需通过浏览器插件完成基础配置，授予地理位置或搜索偏好权限，系统便会在后台静默运行，在品牌方竞标采购其去标识化数据时，自动将代币报酬存入个人数字钱包，推动技术跨越早期采用鸿沟走向大众普及。 Silver 将浏览器端 Permission Agent 视为用户在去中心化网络中确立数据主权的实践工具。然而，该模式能否在大规模主流用户中推开，仍取决于品牌企业对代币化合规采购的实际接纳意愿，以及终端消费者对持续分享行为数据的信任程度。

Detailed notes — zh

## Background 该视频由 YouTube 频道 Ai4 于 2025 年 8 月 27 日发布，演讲者为 Permission, Inc. 创始人兼首席执行官 Charlie Silver。内容简介指出，当前 AI 系统高度依赖未经许可且未验证的爬取数据，随着监管收紧与诉讼激增，获取合规许可与可审计的数据已成为构建可信 AI 的必然要求。 ## TL;DR Charlie Silver 结合其创立 RealAge 的创业经历，指出当前网络爬取数据充斥着机器人流量与点击欺诈，呼吁行业建立包含上下文、可验证性、明示许可和可审计性的“数据黄金标准”。他主张数据是个人私有财产，大科技公司长期无偿榨取红利的模式不可持续。为此，Permission, Inc. 推出了 Permission Agent 与代币化奖励积分 ASK，借助 Agent-to-Agent（A2A）协议与 Web5 去中心化架构，让个人用户在智能体网络中自主掌控数据并实现合规变现。 ## Atmosphere & style 整场演讲洋溢着资深科技创业者的自信、犀利与务实，叙事节奏明快且充满极具穿透力的商业隐喻。从米其林厨师选食材、德州农田被偷采石油，到《甜心先生》的经纪人角色，Silver 将复杂的合规与加密技术讲得生动接地气。现场既有对科技巨头垄断与数据隐私幻觉的深刻剖析，也带着迎接 Agentic Web 新范式的振奋感，给观众带来兼具商业启迪与前瞻视野的沉浸式体验。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=4s) Charlie Silver 拥有 25 年许可业务经验，其早期创立的 RealAge 曾通过获取 60% 用户的健康数据许可选择加入，实现了广告转化率 10 倍的爆发式增长。 - [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) 大科技公司长期单方面无偿榨取网民数据红利，构建起数万亿美元的广告帝国（如亚马逊仅用 6 年打造出年入 800 亿美元的广告业务），而用户理应成为自身数据资产的直接受益者。 - [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) AI 面临“垃圾进，垃圾出”的根本困境：网络流量中至少 30% 到 40% 为机器人造假，全球点击欺诈产业链年规模高达 1000 亿美元，直接导致爬取数据严重失真。 - [05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) 构建下一代优质 AI 模型的“黄金标准”依赖四大支柱：语境上下文（Context）、真实性与出处验证（Verification）、明示知情许可（Consent）以及可审计合规性（Auditability）。 - [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) 现代社会中“隐私只是一种幻觉”，车牌数据、房贷记录与杂志订阅等信息一直在被公共机构和企业反复转卖变现，网民最务实的选择是直接通过自己的数据资产获取经济回报。 - [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) 随着欧洲 GDPR 严厉执行以及美国加州 CCPA 等 12 个州陆续立法收紧，基于爬取数据的无序 AI 时代正在终结，合规且经过明示许可的人类数据成为刚需。 - [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) Permission Agent 旨在成为用户的个人“数字经济经纪人”，在 Agentic Web 中既代表用户自主处理繁杂任务，也负责将授权数据向合规企业变现。 - [14:11](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=851s) 在 Google 推行的 A2A（Agent-to-Agent）协议生态下，企业商业 Agent 需与个人 Permission Agent 进行跨智能体握手，通过“选择性价值交换”获取真实人类数据。 - [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) ASK 代币将传统封闭且易贬值的会员忠诚度积分转变为区块链上的现实世界资产（RWA），为用户的注意力与数据授权提供具备全球流动性与透明度的代币化奖励。 - [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) AI 与 Crypto 的结合正迎来从早期采用者向大众主流市场“跨越鸿沟”的关键时刻，用户未来只需极简配置即可实现浏览器端的全自动被动收益。 - [20:59](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1259s) Permission Agent 协同去中心化 Web5 愿景，致力于打破中心化科技巨头对全网数据的垄断，建立用户真正掌控数据所有权的新型数据经济生态。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=4s) 个人经历与“许可”理念的源起：从 RealAge 到 Permission, Inc. Charlie Silver 首先向与会者致意，并介绍了自己长达 25 年专注于“许可业务”（Permission Business）的创业背景。他回顾了自己在互联网泡沫早期创立的健康风险评估公司 RealAge。当年与他共同创业的合伙人正是日后家喻户晓的 Mehmet Oz 博士（Dr. Oz）。RealAge 通过收集用户关于健康与生活状态的数百个数据点，主动征询用户是否愿意接收有助于改善生理年龄的相关产品与服务信息。当时有大约 60% 的用户主动选择加入（Opt-in）。这一举动带来了惊人的商业回报：其广告转化与用户响应率不仅不是轻微提升，也不是常规的 3 倍或 5 倍，而是“全方位实现了 10 倍（10x）的爆发式增长”。在成功将 RealAge 出售给赫斯特集团（Hearst）后，Oz 博士走向电视荧幕并进军政界，而 Silver 则继续投身创业。他在 5 年多前创立了 Permission, Inc.，其核心信念在于：数据本身蕴含巨大价值，大科技公司长期单方面榨取数据红利，而个人理应从自己的数据中获得应有的报酬。随着人工智能时代的全面爆发，这一愿景正迎来真正规模化落地的契机。 - 核心数据与事实：RealAge 收集数百项健康数据点，用户许可选择加入率达 60%，用户响应率实现全渠道 10 倍提升。 - 创业愿景：打破 Big Tech 垄断数据红利的局面，使个人成为自身数据价值的直接收益者。 --- ### [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) 垃圾进，垃圾出：网络爬取数据的真相与品质细节谈及当前 AI 行业的现状，Silver 强调了计算机科学的经典箴言——“垃圾进，垃圾出”（Garbage in, garbage out）。业界普遍存在一种盲目乐观的误区，认为随意在互联网上抓取（Scrape）数据即可用于训练优质模型。然而，现实中的网络流量充斥着虚假信息：保守估计有 30% 至 40% 的网络流量来自机器人（Bot），高估值甚至远超此比例；全球因网络点击欺诈（Click Fraud）产生的黑色产业链规模高达 1000 亿美元，成为仅次于毒品交易的全球第二大有组织犯罪。大量 AI 团队正是在这些充斥着机器人造假与欺诈的数据地基上构建智能体与大模型。 ```mermaid flowchart LR accTitle: 数据质量与 AI 结果关系图 subgraph BadPath["低质数据链路"] A["网络爬取数据\n(30%-40% 机器人/点击欺诈)"] --> B["AI 模型与 Agent"] --> C["错误输出 / 虚假幻觉"] end subgraph GoodPath["许可数据链路"] D["真实人类数据\n(知情许可 / 上下文验证)"] --> E["AI 模型与 Agent"] --> F["高精度决策与 10x 转化"] end ``` Silver 形象地打了个比方：“即便你是米其林超级大厨，拥有最顶级的烹饪技巧与刀工，如果食材是劣质的，做出来的菜品也绝不可能可口。”他进一步引用了沃尔特·艾萨克森（Walter Isaacson）为史蒂夫·乔布斯（Steve Jobs）和埃隆·马斯克（Elon Musk）撰写的传记。大众往往只看到 iPhone 优雅的外形、玻璃质感与交互界面，但乔布斯真正痴迷的是手机内部每一个零部件的极致细节；马斯克亦是如此。AI 产品的最终表现，完全取决于其核心原料——数据细节的品质。 - 行业痛点：网络流量中至少 30%–40% 为机器人流量，爬取数据严重失真。 - 欺诈规模：全球网络点击欺诈年规模达 1000 亿美元，是仅次于毒品走私的全球第二大非法产业。 - 核心比喻：米其林厨师离不开顶级食材；乔布斯对硬件内部细节的极致把控对应 AI 对底层训练数据的严苛要求。 --- ### [05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) 优质数据的四大支柱与“黄金标准” 在 Silver 看来，真正能驱动下一代 AI 的高质量数据必须具备四大核心要素： 1. **上下文（Context）**：AI 经常产生离谱的幻觉，根本原因在于缺乏语境。例如单词“Jaguar”，既可以指野生美洲豹，也可以指豪华汽车品牌，缺乏上下文的模型无法准确理解。 2. **可验证性（Verification）**：必须清晰掌握数据的来源出处（Provenance），确保数据是由真实人类产生而非伪造。 3. **明示许可（Consent）**：授权必须是清晰且自觉的，绝不能悄悄隐藏在冗长晦涩的用户服务协议（Terms of Service）中，而应像求婚一样，是双方明确知晓并满怀喜悦达成的共识。 4. **可审计性（Auditability）**：在各行各业都转型为 AI 驱动的时代，企业部署的模型与智能体若无法向监管和审计部门交代数据源头，必将面临巨大的合规灾难。 Silver 提出了“数据黄金标准”（Gold Standard of Data）的概念。正如医疗领域将最顶级的护理与诊疗称为黄金标准、金融领域曾以黄金为锚一样，融合了明确上下文、真实性验证、真人参与、主动许可与可审计性的数据，才是 AI 行业的黄金标准。 --- ### [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) 数据即私有财产与“隐私即幻觉” Silver 强调，数据本质上是个人的“私有财产”（Property）。然而在过去数十年里，Google、Meta、Microsoft、Amazon 等巨头通过货币化用户数据，创造了数万亿美元的惊人市值。以亚马逊为例，仅凭掌握用户的搜索、浏览与购买数据，就在短短 6 年内凭空打造出一个年营收高达 800 亿美元的广告帝国。 ```mermaid flowchart TD accTitle: 数据资产收益分配对比 subgraph Current["现状：大科技垄断数据资产"] direction TB User1["个人产生数据\n(搜索/购物/行踪)"] --> Tech["Big Tech 巨头\n(Google/Amazon 等)"] Tech --> Profit1["数万亿商业广告利润\n(Amazon 广告年入 800 亿美元)"] Tech -. "无偿开采，零补偿".-> User1 end subgraph PermissionModel["未来：许可经济与价值对等交换"] direction TB User2["个人数据资产"] --> Agent["Permission Agent\n(个人代理智能体)"] Agent --> Biz["企业营销 / AI 需求方"] Biz --> Reward["ASK 代币与通兑权益"] Reward --> User2 end ``` 针对流行的“数据是新时代的石油”这一比喻，Silver 补充了生动的画面：“想象一下你在得克萨斯、俄克拉荷马或蒙大拿拥有一片农场牧场，埃克森美孚（ExxonMobil）直接开着钻井机冲进你的土地抽走石油，却不给你支付一分钱补偿——这就是如今网民面临的真实处境。” 谈及数据隐私，Silver 直言：“隐私只是一种幻觉（Privacy is an illusion）。”在现代社会，只要你持有信用卡、申请房贷、订阅杂志、拥有驾照或参与投票，你的个人数据就会被反复打包、转卖并再次转卖： - 加利福尼亚州机动车辆管理局（California DMV）每年仅靠向营销人员出售车主驾驶执照数据（如丰田与宝马车主的年龄层与画像）就能赚取 8000 万美元。 - 赫斯特集团旗下的杂志业务，其首要盈利来源同样是向市场出售各类杂志的订阅者名单。 - 除非彻底脱离现代社会、不用手机、没有银行账户和信用卡，否则隐私根本无从谈起；因此，用户最务实的选择是通过自身数据直接获取经济回报。 --- ### [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) 合规浪潮与 Permission Agent 的角色定位全球数据监管正在全面收紧。欧洲对 GDPR 实施了严厉的执行惩罚；尽管美国起步较慢，且 Meta 等科技巨头通过诉讼和拖延策略阻挠监管，但加州的《加州消费者隐私法案》（CCPA）正在推进，且已有 12 个州相继出台了明确的数据隐私与知情许可法规。这意味着任何构建 AI 工具与营销模型的企业，都必须严格遵守真实许可的合规底线。在此背景下，Silver 正式发布了公司的核心产品——**Permission Agent**（许可智能体），并用三部经典影视作品中的标志性角色阐述其定位： - **《明星伙伴》（Entourage）中的 Ari Emanuel**：好莱坞顶级经纪人，唯一使命就是代表客户利益去争取最高报酬。 - **《甜心先生》（Jerry Maguire）**：践行“Show me the money”口号，千方百计为雇主争取真金白银。 - **《我的盛大希腊婚礼》（My Big Fat Greek Wedding）中的女主角**：作为无微不至的旅行社代理，扮演用户的“Mini-Me”，在网络世界代为处理订酒店、比价、购物等日常琐事。 Permission Agent 的核心使命就是成为用户的私人专属经济代表，一方面在后台代表用户将经过授权的数据进行变现盈利，另一方面在 Agentic Web（智能体网络）中代表用户自主执行各项复杂任务。 --- ### [14:11](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=851s) Agentic Web、Google A2A 协议与选择性价值交换 Silver 阐释了正在快速成型的智能体互联网（Agentic Web）。Google 目前正在全美各地大力推广其称为 **A2A（Agent-to-Agent，智能体对智能体）** 的底层协议，企业可以在数分钟内快速生成并部署商业智能体。例如耐克若想寻找打网球或高尔夫的目标客群，便可派出商业 Agent；但该 Agent 无法直接抓取用户，必须与代表个人的 Permission Agent 进行跨智能体握手。 ```mermaid flowchart LR accTitle: A2A 协议下的价值交换闭环 subgraph Enterprise["企业端 (Business Agents)"] Nike["企业商业 Agent\n(如: 耐克寻找网球/高尔夫用户)"] end subgraph Protocol["A2A 协议交互"] Nike <-->|"1. 提出数据需求与激励报价"| PA PA <-->|"2. 确认授权并回传上下文数据"| Nike end subgraph UserSide["用户端 (Individual Agents)"] PA["个人 Permission Agent\n(数字资产管家)"] User["终端用户"] <--> PA end ``` 这种两端 Agent 的交互必须依托成熟的商业机制——**“选择性价值交换”（Opt-in Value Exchange）**： - **历史渊源**：Silver 回忆起童年时与母亲一起收集并粘贴 S&H Green Stamps 绿色印花兑换商品的经历，指出积分奖励模式已存在上百年。 - **优秀标杆**：万豪旅享家（Marriott Bonvoy，Silver 自述曾长期保持白金会员）、美国运通（American Express）、西南航空（Southwest Airlines）、美联航、达美航空与星巴克。这些巨头深耕忠诚度积分，以真诚的奖励换取用户的主动授权与深度互动，进而建立起持久的信任与复购闭环。 - **反面教训**：绝大多数普通消费品牌虽然也设立奖励积分，却频繁随意更改规则甚至贬值权益，严重挫伤了用户的参与信任。 --- ### [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) ASK 代币化奖励积分与 AI + Crypto 的跨越鸿沟针对以往积分体系封闭、割裂的痛点，Permission 平台推出了加密代币 **ASK**。Silver 回顾了 ASK 早期的探索历程：此前由于美国上一届监管环境对加密货币的强硬打压，主流广告主不敢承担监管合规风险，导致项目未能大规模扩张；而如今整个政策与监管环境已发生彻底反转。 Silver 对加密行业进行了清醒的剖析：“作为在加密领域摸爬滚打多年的从业者，我必须承认 90% 的加密项目都是毫无价值的‘赌场筹码’（Casino Chips BS）。”但加密技术真正的突破点在于**现实世界资产（Real World Assets, RWA）的代币化**： - 如同 Circle 通过 IPO 受到资本热捧，将现实中的美元代币化，使其享受区块链的即时流动性与透明度；未来房地产、艺术品等资产也将全面代币化。 - **ASK 的定位正是“代币化的奖励积分”**。传统企业（万豪、星巴克等）的积分往往被禁锢在各自的封闭孤岛中，无法流通、缺乏公开透明度且不能自由交易；而 ASK 赋予了忠诚度积分真正的全球流动性、透明度与市场化增值空间。引用杰弗里·摩尔（Geoffrey Moore）在《跨越鸿沟》（*Crossing the Chasm*）中的经典理论，Silver 指出，AI 与 Crypto 的结合正处于从早期采用者（Early Adopters）迈向大众主流市场（Mainstream）的关键跃迁节点。普通用户只需花 2 秒钟下载配置好 Permission Agent，在日常上网过程中对地理位置、购买意向或社交偏好点击授权，代币收益便会自动被动流入加密钱包。 --- ### [20:59](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1259s) Web5 去中心化愿景与 Permission Agent 发布在演讲尾声，Silver 引用了科技巨头（如 Jack Dorsey）倡导的 **Web5** 构想：彻底终结 Google、Meta、Microsoft 和 Amazon 对全网数据的中心化垄断，建立一个用户完全自主掌控数据所有权、自由通过数据盈利的真正去中心化网络。而 Agentic Web 协同 Permission Agent，正是将 Web5 从抽象概念变为现实的核心引擎。 Silver 正式宣布 Permission Agent 即日起面向公众发布，并即将上线供用户下载使用。用户安装浏览器插件后即可开启全自动被动收益模式；当外部需求方合规调取数据时，浏览器会实时弹窗提醒收益已到账。除数据变现外，该智能体还将陆续推出各类个性化助理功能。最后，Silver 邀请现场观众参加在 203 会议室举办的营销招待酒会，并前往展区第 640 号展位体验交流。 ## Notable quotes > [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) 垃圾进，垃圾出。 — Charlie Silver > [02:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=134s) 即便你是米其林超级大厨，拥有最顶级的烹饪技巧与刀工，如果食材是劣质的，做出来的菜品也绝不可能可口。 — Charlie Silver > [05:00](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=300s) 数据黄金标准 — Charlie Silver > [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) 隐私只是一种幻觉。 — Charlie Silver > [07:41](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=461s) 想象一下你在得克萨斯、俄克拉荷马或蒙大拿拥有一片农场牧场，埃克森美孚直接开着钻井机冲进你的土地抽走石油，却不给你支付一分钱补偿——这就是如今网民面临的真实处境。 — Charlie Silver > [11:14](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=674s) Show me the money — Charlie Silver > [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) 作为在加密领域摸爬滚打多年的从业者，我必须承认 90% 的加密项目都是毫无价值的‘赌场筹码’。 — Charlie Silver > [17:21](https://www.youtube.com/watch?v=zdC7FGW5o7s&t=1041s) 代币化的奖励积分 — Charlie Silver ## Mentioned Charlie Silver, Permission, Inc., RealAge, Mehmet Oz (Dr. Oz), 赫斯特集团 (Hearst), Walter Isaacson, Steve Jobs, Elon Musk, iPhone, Google, Meta, Microsoft, Amazon, ExxonMobil, California DMV, GDPR, CCPA, Permission Agent, 《明星伙伴》 (Entourage), Ari Emanuel, 《甜心先生》 (Jerry Maguire), 《我的盛大希腊婚礼》 (My Big Fat Greek Wedding), Google A2A 协议, Nike, S&H Green Stamps, 万豪旅享家 (Marriott Bonvoy), 美国运通 (American Express), 西南航空 (Southwest Airlines), 星巴克 (Starbucks), ASK 代币, Circle, Geoffrey Moore, 《跨越鸿沟》 (Crossing the Chasm), Jack Dorsey, Web5 ## Video description Charlie Silver, Founder/CEO, Permission, Inc. Abstracts AI systems are only as good as the data they rely on, and most of that data today is scraped, unverified, and unconsented. As lawsuits mount and regulations tighten, the era of unrestricted AI data is ending. For teams that require human data – whether for marketing, personalization, or agent-to-agent interactions – permissioned, auditable data is no longer optional. This talk is a wake-up call for AI teams: permission isn’t just a competitive advantage; it’s imperative for trusted, compliant AI – and we'll explore what it takes to get it right. [Source](https://www.youtube.com/watch?v=zdC7FGW5o7s)

Source quotes

[2:14] Charlie Silver: 垃圾进，垃圾出。 [2:14] Charlie Silver: 即便你是米其林超级大厨，拥有最顶级的烹饪技巧与刀工，如果食材是劣质的，做出来的菜品也绝不可能可口。 [5:00] Charlie Silver: 数据黄金标准 [7:41] Charlie Silver: 隐私只是一种幻觉。 [7:41] Charlie Silver: 想象一下你在得克萨斯、俄克拉荷马或蒙大拿拥有一片农场牧场，埃克森美孚直接开着钻井机冲进你的土地抽走石油，却不给你支付一分钱补偿——这就是如今网民面临的真实处境。 [11:14] Charlie Silver: Show me the money [17:21] Charlie Silver: 作为在加密领域摸爬滚打多年的从业者，我必须承认 90% 的加密项目都是毫无价值的‘赌场筹码’。 [17:21] Charlie Silver: 代币化的奖励积分

Source quotes

[2:14] Charlie Silver: garbage in, garbage out [5:00] Charlie Silver: If they don't know where the data is coming from, you're going to be this guy [7:41] Charlie Silver: data is the new oil [11:14] Charlie Silver: Show me the money! [11:14] Charlie Silver: mini-me [14:11] Charlie Silver: opt-in value exchange [17:21] Charlie Silver: 90% of it is BS [17:21] Charlie Silver: cross the chasm [21:01] Charlie Silver: Web5

Original transcript

[0:04] Welcome everybody. Thank you for [0:06] attending. Uh we'll give it another [0:09] minute before a few of you stragglers [0:12] are [0:14] coming in. Anyway, let me introduce [0:16] myself. Uh I actually have 25 years in [0:22] the permission business. Uh, I do see [0:26] some faces out there that could remember [0:28] a company that I built in the early dot [0:31] period. It was called Real Age. Anybody [0:35] heard of it? It was a health risk [0:39] assessment and actually you I'm sure you [0:42] all know my partner Dr. Memed Oz. That's [0:45] where he got his media start. Real age [0:49] was a health risk assessment that [0:51] collected hundreds of data points about [0:54] someone's health and well-being and then [0:56] would ask permission if someone would [0:59] like to receive information about [1:01] products and services that can make the [1:04] real age younger. And about 60% of the [1:07] people opted in. And when they did, the [1:10] response rate wasn't a little better. It [1:14] wasn't three or 5x better, but it was [1:17] 10x better across the board. Once [1:20] somebody shares data and gives [1:23] permission, we know I know what the [1:26] results would be. I sold the company to [1:28] Hurst very successfully and that's when [1:31] Dr. Oz took over the Oprah show and now [1:33] he's in the Trump administration and [1:36] here I am building startups again. I've [1:40] been doing it for 30 years and I'm more [1:42] excited than ever about permission. [1:46] And by the way, permission was launched [1:50] several years ago, over five years ago, [1:52] with the one idea that data has value [1:56] and that individuals should be paid for [1:58] it. That big tech extracts all the value [2:01] from data. But it was time for [2:04] individuals to receive that value. And [2:07] now with AI, that vision is going to [2:10] scale and become real. [2:14] So everybody knows the old adage, [2:19] garbage in, garbage out. It couldn't be [2:22] more true when it comes to AI. [2:25] People think in this industry, oh, you [2:28] could scrape the web and that's great [2:29] data. Oh, you can get, [2:33] you know, data off of anywhere. You can [2:36] scrape it. Well, [2:40] I'll make the point that that's garbage [2:42] data. What do you think the web traffic, [2:45] how much is bots? Anybody who's been in [2:47] the internet business, do you know, you [2:50] have any idea how much is bots based? [2:54] Anybody have a guess? [2:57] >> That's on the high end. Absolutely. on [3:00] the low end 30 40%. So that means the [3:04] data that you're scraping is just all [3:07] made up. Do you know that organized [3:09] crime across the globe that internet [3:12] fraud is the second biggest business [3:15] next to narcotics? It's a hundred [3:18] billion dollar business. Click fraud. [3:21] And that's the data that people are [3:23] building their AI agents and models on. [3:28] Think about that one. So, we all know [3:31] garbage in, garbage out. [3:34] Imagine you're a Michelin chef, a [3:37] superstar chef. You have the best [3:39] techniques, the best skills, [3:42] but if you don't have the best [3:44] ingredients, [3:47] your food's not going to be very good. [3:49] So, keep that image in mind. [3:54] Quality is something [3:57] that is so important. Data quality. Who [4:00] read Walt any of Walter Isis books on [4:04] Steve Jobs or Elon Musk? Anybody read [4:07] those? [4:09] A few of you. Okay. Well, when you read [4:12] those biographies, [4:14] Steve Jobs of particular, what stood out [4:17] was that he was so fascinated with the [4:20] details. every detail that went inside [4:23] the phone. Everybody thinks about the [4:25] elegance and the interface and the shape [4:28] and the glass, but Jobs was focused on [4:31] what was inside. Every detail mattered. [4:37] Musk is the same way. These are the two [4:40] greatest entrepreneurs of this of of our [4:43] of this information age and it's all [4:46] about the details. And again, think [4:50] about AI products. It's the details of [4:54] the ingredients, which is the data is [4:57] going to determine the outcome. [5:00] Let's talk about what makes good data. [5:03] Context. [5:05] I mean, we've all used [5:07] various AI tools and every and now and [5:10] then you get this like weird results. [5:13] Well, context is important. I mean, this [5:16] is a perfect example. What's a jaguar? [5:18] Well, could be a wild animal or it could [5:22] be a car. AI has trouble with that [5:28] verification. [5:30] Do you know the provenence of the data? [5:33] Do you know where it's coming from? [5:36] How certain are you? So, verification is [5:40] absolutely essential for understanding [5:43] data quality. [5:46] And this is my favorite slide because [5:49] consent means permission. But in this [5:52] con context, somebody's thrilled to give [5:56] consent. Just like in a marriage [5:58] proposal, [6:00] consent is something that has to be [6:02] conscious [6:04] and clear. Can't be just buried in a [6:07] terms of service. It has to be overt. [6:13] And guess what? You don't want to be [6:15] this guy. [6:18] The data has to be auditable. [6:22] Imagine at a big enterprise you're [6:24] building every company now is an AI [6:26] company. [6:29] That's what's so exciting about this [6:30] industry. There is no, you know, every [6:34] business in the world is now an AI [6:36] company. [6:38] and the tools that they're building and [6:40] the models and the agents. If they don't [6:43] know where the data is coming from, [6:46] you're going to be this guy. [6:51] So, I like this term. I'm not sure we're [6:53] going to go with it. Who likes this term [6:56] gold standard of data? I actually like [6:58] it, but I'm not sure we're we're going [7:01] to run with it. Who likes it? [7:06] Okay. Well, that's about 50%. I'm of the [7:09] age and I'm of the mindset that the gold [7:12] standard actually means something. In [7:14] healthcare, gold standard is it's always [7:17] the highest quality of care. That's a [7:19] commonly used term. Obviously in [7:21] finance, although our fiat system is [7:24] kind of forgotten about the gold [7:26] standard, but [7:28] anyway, all of these things, context, [7:32] verification, real human beings, real [7:34] permission and consent, all mean the [7:37] gold standard of data. [7:41] Now, when I founded permission, it was [7:44] all based on the idea that data is [7:46] property. It is our individual property. [7:52] We own it. [7:55] But what's happened over the last [7:58] decades is big tech, Google, Meta, [8:02] Microsoft, Amazon, not that they're [8:06] evil, but they have monetized data [8:09] beyond anyone's compre, you know, [8:12] thoughts that could ever be done. [8:15] Data is real property. It obviously has [8:18] enormous value. These companies are [8:20] worth trillions. [8:23] Trillions. And the only asset is data. [8:27] Amazon went from zero to 80 billion. [8:31] They're going to do $80 billion in [8:33] advertising this year. In six years, [8:36] they've built an $80 billion advertising [8:39] business. Why? because they have all of [8:42] our data about what we buy, what we [8:45] search, what we, you know, purchase and [8:48] what we look for. [8:51] That data has incredible value. And, you [8:54] know, same with Google, what we search [8:56] for, Facebook, who we are friends with. [9:00] So, this property [9:04] is real and everyone knows the metaphor [9:07] data is the new oil. But I like this [9:10] picture because imagine you had a ranch [9:15] in Oklahoma, Texas, Montana, and Exxon [9:18] just came in and started just drill [9:20] taking your oil and not compensating you [9:23] for it. [9:25] That is the situation we have today [9:28] online. Your data is being exploited. [9:32] Your property is being taken without [9:34] compensation. [9:40] a lot of talk about privacy [9:44] and my view and I think it's factual [9:48] that privacy is an illusion. It's just [9:51] an illusion. Everybody wants to talk [9:53] about being private, but if you have a [9:56] credit card, if you have a mortgage, if [9:59] you have a magazine subscription, if you [10:02] have a driver's license, if you vote, [10:06] your data is sold and sold and [10:10] repackaged and sold again. [10:14] California Department of Motor Vehicles. [10:16] They make $80 million a year by selling [10:20] data, driver's license data. Who drives [10:23] Toyotas? Who drives BMWs? Marketers buy [10:26] that data. How old are people buying [10:29] this car and that car? And all the kind [10:33] of data that they collect, it is for [10:35] sale. [10:37] Every magazine company I worked at Hurst [10:40] after I sold Real Age, their number one [10:42] revenue is selling who subscribes to [10:45] what magazine. [10:48] So this concept of privacy is you got to [10:52] be that guy off the grid without a bank [10:56] account, without a credit card, without [10:58] a telephone [11:00] if you want your data to be private. [11:04] So our point is you might as well make [11:06] money from it because you ain't going to [11:09] it's not private. [11:14] Obviously compliance is a huge issue. [11:20] We all know about GDPR [11:23] and the European Union has very strictly [11:26] enforced GDPR. [11:30] parts of Asia. The US has been slow to [11:34] enforce data regulations. California has [11:38] the California Consumer Privacy Act, but [11:41] it has slowed down. Meta filed all kinds [11:45] of delays and lawsuits and stuff like [11:48] that. So, the California Privacy Act is [11:51] slowly rolling out. But there are 12 [11:53] other states that have private you know [11:56] data regulations about how data can use [11:59] what is real permission what is real [12:01] consent. [12:04] So marketers and anybody building AI [12:08] tools have to be con cognizant of these [12:12] regulations. [12:16] So, we announced today, we put out a [12:19] press release. The permission agent, who [12:23] knows who that guy is? [12:26] Okay. He's a Hollywood agent from [12:28] Entourage, Ari Emanuel. What was his [12:31] job? His job was to represent actors and [12:35] get them the most money they could get. [12:40] It's a great metaphor for the permission [12:44] agent. Who knows who that is? [12:48] Anybody seen this movie? I'm not going [12:50] to do the dance, the Cuba good and uh uh [12:53] show me the money dance. But here's [12:56] another agent whose job was to make [13:00] money for their clients. [13:04] That's what the permission agent is all [13:06] about. The permission agent is all about [13:09] taking your data [13:12] and making you money. But it also does [13:16] all kinds of other tasks. [13:20] Who knows who that is? [13:23] Who knows the reference? Well, it was [13:25] actually a very famous movie that cost [13:28] $2 million to make and one is one of the [13:30] biggest Hollywood blockbusters of all [13:32] time, My Big Fat Greek Wedding. That uh [13:35] she was the star and the creator of it. [13:38] Well, here she is as a travel agent. [13:42] So the permission agent is not only [13:44] there to make you money, but it [13:46] represents it's a mini me. It can do any [13:49] of your tasks online. Book you a hotel [13:52] in New York, find shopping, do anything [13:56] that you think about in and how the [13:59] agentic web is going to roll out, [14:03] but its primary job is to be Jerry [14:05] Maguire and area manual. It's to make [14:08] you money. [14:11] So [14:13] there's this agentic web we're all [14:15] hearing about [14:18] which is what Google calls the Google [14:21] protocol is literally a to a in that [14:25] acronym letter A numeral 2 A agent to [14:31] agent it is coming out fast just we're a [14:34] Google partner um [14:37] uh and I've been at several events [14:40] events that Google's putting on all [14:43] around the country, teaching business [14:46] that they could deploy business agents [14:48] to do tasks in minutes. [14:52] Literally, if you want to find you're [14:55] Nike and you want to find people who [14:56] play tennis or golf, you can create an [14:59] agent and set it out, but it's got to [15:02] connect to another agent. It's got to [15:04] connect to individual agents like the [15:07] permission agent. And when that happens, [15:12] this has to happen, which is this is a [15:16] term that's been around for decades in [15:18] marketing, opt-in value exchange. I'm [15:22] old enough as a kid to remember [15:24] something called SNH green stamps that I [15:26] used to paste in a book with my mother. [15:29] Okay. Well, thank you. But it was real. [15:31] It was a hundred-year-old business. And [15:33] they were adopters of this idea of [15:36] opt-in value exchange. I have Marriott [15:39] up there because that's my favorite [15:41] rewards program. Marriott Bonvoy. I love [15:44] my Bonvoy points. [15:47] I was platinum for quite a while. I I [15:50] lost that status recently. [15:53] Um but uh uh that the whole rewards [15:58] world which almost every consumer [16:00] business has some kind of rewards [16:03] program. That is the idea. We reward [16:06] you. We get your data. We want your [16:09] engagement. And the best consumer [16:12] companies in the world are committed to [16:15] it. American Express has built their [16:17] whole business on their rewards program. [16:21] Southwest Airlines, that's another one [16:23] of my good ones that I like. They are [16:26] excellent at their rewards programs, but [16:28] so is United and Delta. [16:31] And Starbucks is also excellent. These [16:34] are companies that are committed to this [16:37] concept of you opt in, you engage with [16:41] us, we want to build trust, and we'll [16:44] continue to reward you. And I love my [16:47] points, and that's why I stay coming and [16:50] doing business with these companies. [16:52] Now, the vast majority of consumer [16:55] companies aren't committed to their [16:58] rewards program. they have it, they [17:00] change it, they uh they change it again, [17:02] they do it again. And and obviously [17:05] that's reflected in consumer adoption. [17:09] So these companies that I just listed [17:11] are just fantastic in this idea of [17:14] rewarding users in exchange um for trust [17:18] and engagement. [17:21] So part of our platform [17:24] is something called ASK. [17:27] We started it many years ago as a crypto [17:30] comp you know with a crypto as a crypto [17:33] company with the idea in digital [17:35] advertising you needed to reward people [17:38] for their engagement. Well all kinds of [17:41] reasons I could tell you why we didn't [17:42] scale. The biggest one was the the [17:46] administration previous was, [17:49] you know, battled crypto and it made any [17:53] major advertiser not willing to take [17:55] regulatory risk. That's all changed. [17:59] That's all changed. [18:01] Who's here involved with crypto? Anybody [18:03] involved with crypto? [18:05] A few of you. Well, that's good because [18:08] let me tell you something about someone [18:10] who's been in crypto for a long time. [18:12] 90% of it is BS. It's BS. It's just [18:17] casino chips. But the really cool stuff [18:21] about crypto is what's called real world [18:24] assets. You just saw Circle go to the [18:27] moon with their IPO. That's a tokenized [18:31] dollar, something that exists in the [18:33] real world that benefits from [18:35] tokenization and being traded on a [18:38] blockchain. And you're going to see all [18:42] kinds of assets on real estate and art [18:45] become tokenized. That's where crypto is [18:48] really going to become mainstream is the [18:51] tokenization of real world assets. And [18:54] our crypto ASK [18:57] is a tokenized rewards point. And it [19:01] benefits a rewards point benefits from [19:04] tokenization because of liquidity, [19:07] transparency, [19:09] its ability to appreciate and have a [19:11] real market. All those companies that I [19:13] listed before have great rewards [19:15] programs, but their tokens are siloed or [19:18] their rewards points are siloed. They're [19:21] not liquid. They're not transparent. You [19:24] can't trade them. And that's what ASK is [19:27] all about. [19:30] So AI and crypto comes together [19:34] in this Ato world. [19:38] You see that guy? That's a crypto bro [19:40] with his Lamborghini. [19:42] The permission agent is going to make [19:46] this happen when everybody [19:51] simply and easily can benefit by just [19:54] giving permission to share their data, [19:58] their geography, their intent, their [20:00] social media, you know, everything that [20:03] happens online. Hey, I'm going to share [20:06] it and I'm going to earn from it [20:09] and that's what goes mainstream. [20:11] Okay? every you know if you know this [20:13] concept crossing the chasm [20:16] you know we're still in the early [20:18] adopter phase but I think very very soon [20:22] AI and crypto come together and go [20:26] mainstream that thing that white column [20:30] is the chasm [20:32] I don't know if you know the book [20:34] written by Jeffrey Moore in the 80s but [20:36] this is a concept for every new [20:38] technology how does it go from early [20:41] adopters [20:42] to mainstream. [20:44] And we're so excited that the permission [20:46] agent, you download it, you set it up, [20:49] takes two seconds, and all of a sudden, [20:51] you just give permission and you earn [20:53] passively. It just shows up in your [20:55] wallet. [20:59] So, [21:01] I think you probably all know who that [21:02] dude is. You know, he's been a tech [21:05] titan. He's out there calling about [21:07] what's web five. There's something [21:10] called web five which is truly the [21:13] decentralized web where instead of [21:16] Google, Microsoft, Amazon, Meta having [21:19] all the data, you control your data. You [21:23] own your data. You make money from your [21:26] data and the data is truly [21:29] decentralized. And AI and agents, the [21:33] agentic web and the permission agent is [21:36] going to make this real. [21:39] So today we put out a press release. We [21:43] announced the permission agent. We're [21:45] super excited. Uh it's going to be [21:48] available very soon. People can download [21:51] it and just start earning and passively. [21:55] They don't have to do anything. Their [21:57] brow, you know, in their browser it [21:59] says, "Oh, you just earned. Someone just [22:01] bought your data." Blah blah blah blah [22:03] blah blah. And we're super excited about [22:05] the permission agent. It's going to have [22:07] all kinds of other functionality that's [22:09] going to delight users and we really [22:12] think it's going to scale in a big big [22:15] way. [22:17] Thank you very much. [22:20] We have a reception in room 203 [22:23] right after this and uh as the marketing [22:26] reception. So please come have a drink [22:29] and whatever they're serving. So and [22:32] come to our booth number 640. [22:36] Thank you.