X Games CEO Spins Out Owl AI After Raising $11 Million for Automated Judging

X Games CEO融资1100万美元推进自动化评分，分拆成立Owl AI Jeremy Bloom正部署基于实况转播画面训练的计算机视觉技术，以消除全球打分与裁判制体育项目中的人为失误、偏见和误判。 前奥运选手、NFL球员兼科技创业者Jeremy Bloom探讨了主观打分和裁判失误如何损害职业体育的公正性。通过与Google Cloud及Sergey Brin合作，Bloom协助开发了Owl AI。这是一个构建于Google Cloud Platform上的计算机视觉平台，可利用标准实况转播画面对极限运动进行客观的实时打分。除了精准裁决与动作评估外，该平台还能生成实时多语言解说，根据专业认知水平为观众提供个性化观赛选项，并旨在为青少年运动员提供AI驱动的教练指导。 前奥运滑雪选手兼 X Games CEO Jeremy Bloom 指出，随着体育博彩市场规模从 2018 年的 4 亿美元激增至 120 亿美元，执裁失误造成的经济与竞争代价空前巨大，促使他致力于消除裁判的人为偏见与错误。 (0:04) Bloom 联手 Google 联合创始人 Sergey Brin 及 Google Cloud 团队打造了 Owl AI 评分系统，无需额外硬件即可直接从电视直播信号中分析动作，将主观技巧风格解构为腾空高度、旋转轴心对齐和抓板执行等客观指标。 (8:01) Owl AI 在阿斯彭冬季 X Games 亮相时精准匹配了人类裁判评分并预测出前三名选手，同时在盐湖城赛事中为日语、西班牙语、普通话和葡萄牙语转播生成了 100% 纯 AI 实时解说。 (13:08) Owl.ai 已完成 1100 万美元融资独立运营，未来不仅将支持观众按专业程度定制个性化解说，还计划将技术普及为青少年运动员的掌上 AI 教练，终极目标是彻底消除竞技体育中的执裁失误。 (17:39)

X Games CEO Jeremy Bloom has launched Owl.ai, an artificial intelligence platform developed to eliminate subjective human error in sports officiating and generate real-time broadcast commentary. Backed by $11 million in venture capital, the venture aims to introduce automated precision to high-stakes judged and refereed sports while democratizing youth coaching. The Officiating Crisis and Rising Economic Stakes (0:04) Drawing from his background as an Olympic freestyle mogul skier, Bloom highlights that athletes spend lifetimes preparing for runs as short as 22 seconds, where medal outcomes can hinge on fractional movements evaluated by tired, distracted, or biased human panels. Past governance and officiating controversies illustrate these vulnerabilities across sports, including the 2002 Winter Olympics figure skating scandal, an MLB umpire fired for gambling, the missed pass interference in the 2018 NFC Championship Game, and controversial officiating decisions in Formula 1. Bloom emphasizes that resolving subjective error has become critical as the economic stakes of sports expand. In the United States alone, the sports betting market expanded from a $400 million industry in 2018 to a $12 billion industry in the past year, magnifying the financial consequences of questionable calls for leagues, athletes, and fans. Computer Vision Scoring from Live Broadcast Feeds (8:01) To build an objective judging system, Bloom partnered with Google co-founder Sergey Brin and Google Cloud Platform engineer Josh. In a five-week development sprint leading up to the Aspen Winter X Games, the team built a model trained on Google Cloud Platform that evaluates men's snowboarding directly from live television broadcast feeds without requiring specialized hardware or proprietary camera arrays on site. The system operationalizes subjective judging criteria into quantitative computer vision metrics, measuring amplitude, on-axis rotation precision (such as 720s and 1080s), airborne economy of motion, and specific trick grabs and tweaks. During live tests on ABC's broadcast, the model mirrored human judging scores with only a single rank variance between fifth and sixth place. Furthermore, by analyzing practice sessions, the system correctly predicted the first-, second-, and third-place podium finishers, outperforming Las Vegas betting odds. Generative Commentary and Layered Audience Personalization (13:08) Beyond judging, Owl AI generates real-time, automated commentary in authentic action sports vernacular. At an event in Salt Lake City, the platform produced fully automated, 100% AI-generated audio tracks in Japanese, Spanish, Mandarin, and Portuguese to supply localized broadcasts for key global markets such as Japan and Brazil. The system dynamically integrates historical context and athlete statistics—such as spontaneously surfacing past competition milestones—without production prompting. Bloom projects that future digital distributions on platforms like Netflix will enable viewers to choose personalized commentary tiers tailored to their familiarity with a sport, ranging from introductory educational commentary for beginners to novice coverage and highly technical breakdowns for endemic fans. Commercial Spinout, Multi-Sport Expansion, and Coaching (17:39) Owl.ai has spun out as an independent company separate from Google and the X Games, raising $11 million in venture capital with Josh serving as CEO and Bloom as founder and executive chairman. The organization is actively exploring applications across basketball, combat sports, and judged Olympic sports such as gymnastics and figure skating, pursuing a primary goal of completely eliminating human officiating error. In addition to professional officiating and white-label league integrations, Owl.ai is targeting grassroots athletic development. With 60 million participants in U.S. youth sports, Bloom outlines a consumer product vision that enables young athletes to upload practice video directly to the platform for instant, automated evaluation and high-level coaching feedback. While Owl.ai demonstrated technical parity with human judges in controlled extreme sports broadcasts, scaling its computer vision architecture across diverse professional sports and grassroots coaching remains an ongoing initiative. The platform's broader adoption will depend on whether established sports federations and commercial leagues are willing to integrate automated officiating into traditionally human-governed competitions.

前奥运选手兼 X Games 首席执行官 Jeremy Bloom 指出，传统体育打分机制长期受制于人类裁判的偏见、疲劳与误判，而在体育博彩业激增的背景下，执裁失误的代价正被急剧放大。为此他创立了基于计算机视觉的平台 Owl AI，旨在利用现有电视直播信号实现客观打分与全自动多语种解说，并期望最终消除竞技裁判失误并推动大众青训普及。 人类裁判局限与博彩业经济压力 (0:04) Bloom 结合自身作为世界自由式滑雪冠军的经历指出，该项目运动员需要经历漫长训练，但决定胜负的滑行往往仅持续22秒；金牌与银牌的差距常取决于极为微小的动作差异。然而，现行评分权受制于7名人类裁判，这些人极易受疲劳、分心、主观偏见乃至舞弊干扰。为佐证传统执裁机制的普遍漏洞，讨论列举了多项标志性误判事件，包括2002年冬奥会花样滑冰打分丑闻、职业棒球大联盟（MLB）裁判因涉赌遭解雇、2018年NFL国联决赛对公羊队的关键传球干扰漏判导致圣徒队出局，以及一级方程式赛车（F1）阿布扎比站赛事总监的争议裁量直接改变年度冠军归属。 执裁争议的负面影响因博彩市场的膨胀而进一步升级。讨论提供的数据显示，美国体育博彩市场规模已从2018年的4亿美元急剧增长至去年的120亿美元。在庞大的资金流转与商业利益冲击下，判决不公已不仅仅关乎纯粹的体育伦理，更对赛事的公信力构成了直接威胁，这促使 Bloom 寻求通过底层技术手段重构执裁公正性。 视频流驱动的 Owl AI 视觉打分 (8:01) 在2021年出售其企业软件公司后，Bloom 于同年12月出任 X Games 首席执行官。面对极限运动难度成倍跃升（例如当年 Tony Hawk 首创的900度转体，如今10岁日本选手可单轮连续完成三次），Bloom 联络了 Google 联合创始人 Sergey Brin。Brin 随后编写了最初代码，并引荐了常驻博尔德的 Google Cloud Platform（GCP）工程师 Josh。团队在1月23日阿斯彭冬季 X Games 前开展了为期五周的敏捷冲刺，在 GCP 上训练出了针对男子单板滑雪的评分模型。该系统无需额外布置传感器或定制摄像头，完全依赖电视台的直播信号源（live TV feed）进行分析。 针对传统裁判主观评估的“风格（style）”，Owl AI 利用计算机视觉将其解构为四项客观工程参数：动作经济性（滞空时动作的利落程度）、轴向对齐度（转体与姿态精准度）、腾空幅度（起跳高度）以及具体抓板动作与身体微调的识别（如区分 stalefish 与 mute grab）。在阿斯彭赛事实测中，该系统在 ABC 的实况转播中与人类裁判组给出的最终名次高度吻合，仅在第5名与第6名间出现了一处微调；同时，系统通过分析练习赛画面生成的前三名领奖台预测，准确命中了冠亚季军，表现超过了拉斯维加斯的博彩赔率。 生成式解说与个性化观赛分级 (13:08) 在量化打分之外，Owl AI 还将能力延伸至自动化赛事解说。在盐湖城赛事中，系统针对滑板等在国际市场具有庞大受众的项目，实时生成了日语、西班牙语、普通话和葡萄牙语的四轨完整赛事解说，且全轨道音频均100%由 AI 自主合成。系统并非仅进行逐字直译，而是能即时调用历史战绩与统计背景，例如转播中系统曾自主补充“Scotty James 于2019年在铜山赛事夺冠崭露头角”的背景掌故，其叙事深度超出了现场转播制作团队的预期。 面向未来一至两年与 Netflix 等流媒体平台的合作，讨论预测赛事解说将演变为由观众自主定义的多层次定制流，以消除小众运动的观赛门槛。用户届时可按需选择三类解说声道：面向大众的“初学者级”，AI 扮演科普导师实时拆解技巧术语与基本规则；“进阶级”，保持实况叙事与基本技术说明的平衡；以及面向资深受众的“专家级”，专注于复杂的动作技巧与行业黑话，使不同认知背景的观众均能获得量身定制的视听体验。 独立商业化拓展与大众青训普及 (17:39) 目前 Owl.ai 已从 X Games 和 Google 独立拆分，作为独立实体运营并完成了1100万美元的风险投资。Jeremy Bloom 担任创始人兼执行董事长，Josh 担任首席执行官，团队成员包括 Tim 等人。公司目前正与职业格斗、篮球以及花样滑冰和体操等传统打分项目展开深入接洽。Bloom 将公司的终极指引目标（North Star）明确定义为彻底消除所有打分与执裁运动中的人为失误，确保判罚达到100%的准确率；尽管公司愿意向合作联盟提供白标技术方案，但强调核心驱动力在于维护体育诚信与精准，而非单纯追求商业回报。 在职业赛事之外，Owl.ai 规划向消费级与大众青少年体育拓展。针对全美约6000万参与青少年体育的儿童群体，公司计划推出“口袋里的顶级教练”应用。年轻运动员和家长只需向平台上传运动视频，即可获得即时的专业级打分评定与执教改进建议，以此将顶尖视觉评估工具下沉至草根群体，促进基层青训的公平性与技能发展。 总体而言，Owl AI 验证了仅凭电视转播信号对高难度运动进行客观评分与自动多语种解说的可行性，展现出重塑赛事公正与观赛体验的潜力。然而，讨论尚未触及该技术在规则高度依赖主观情境裁量（如团队接触对抗中的犯规判定）时的泛化难度，也未评估传统体育管理机构在让渡裁决权时可能出现的制度阻力，以及模型在复杂机位或非标准光照环境下的容错边界。

## Background 本视频由 Ai4 于 2025 年 8 月 27 日发布于 YouTube 平台。页面简介显示，内容为《体育商业杂志》记者克里斯·史密斯（Chris Smith）与 X Games 首席执行官杰里米·布鲁姆（Jeremy Bloom）的炉边对话，探讨人工智能如何通过 Owl.ai 等技术消除裁决偏见、提升体育赛事的公正性、透明度与球迷互动体验。 ## TL;DR 两届奥运选手、前 NFL 球员及科技创业者杰里米·布鲁姆分享了其从竞技体育跨界科技领域的履历，并重点探讨了主观打分项目与职业联赛中因人类裁判误判带来的巨大经济与职业灾难。为解决这一难题，布鲁姆与谷歌联合创始人谢尔盖·布林合作推出了 AI 裁决系统 Owl.ai，该系统不仅能通过实时视觉分析提供“博士级”公正打分，还能自动生成多语言极客风解说。如今 Owl.ai 已完成 1100 万美元融资，其终极目标是消除所有裁决类运动的人为失误，并为青少年运动员提供大众化的顶尖教练服务。 ## Atmosphere & style 这场对谈兼具深厚的体育情怀与极客创新碰撞的激情，氛围专业而富有感染力。布鲁姆以亲历者的视角讲述运动员因短短数秒误判而改写命运的残酷现实，引出对体育诚信的深刻思考；而在分享与科技巨头合作冲刺开发 Owl.ai 的过程时，现场又充满了科技改变未来的务实与兴奋。整场对话既有对传统体育困局的沉重反思，也洋溢着对技术赋能公平竞技的坚定信念。 ## Key takeaways - [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) Jeremy Bloom 结合其奥运选手、NFL 球员及科技创业者的跨界背景，出任 X Games CEO 并推动体育与 AI 的融合。 - [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) 打分项目的极短比赛时长（如滑雪仅 22 秒）与人类裁判的疲劳、主观偏见及行业腐败，暴露了主观裁决的巨大不确定性。 - [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) 误判会对体育产业造成毁灭性后果，如 2018 年 NFC 决赛的关键漏判和 F1 阿布扎比大奖赛的调度争议。 - [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) 美国体育博彩市场规模激增（从 2018 年的 4 亿美元飙升至 120 亿美元），利益的增加使得比赛裁决的公正性与诚信度面临更高要求。 - [08:28](https://www.youtube.com/watch?v=blkGMZn_LBA&t=508s) Google 联合创始人 Sergey Brin 亲自敲下第一行代码，促成了全球首个 AI 体育裁判系统 Owl.ai 的诞生。 - [08:28](https://www.youtube.com/watch?v=blkGMZn_LBA&t=508s) Owl.ai 在短短 5 周内基于 Google Cloud Platform (GCP) 完成开发与冲刺，并成功应用于冬季 X Games 的全球直播。 - [08:28](https://www.youtube.com/watch?v=blkGMZn_LBA&t=508s) Owl.ai 将复杂的“风格”拆解为空中动作效率、起跳高度、轴线控制及抓板细节等量化指标，消除了人类裁判的主观偏见。 - [08:28](https://www.youtube.com/watch?v=blkGMZn_LBA&t=508s) 在冬季 X Games 比赛实测中，Owl.ai 展现出博士级别的打分精准度，并凭借观察训练赛成功预测出比赛前三名。 - [16:24](https://www.youtube.com/watch?v=blkGMZn_LBA&t=984s) Owl.ai 能够自动生成兼具极客风格与专业深度的语音解说，精准辨识复杂动作并结合运动员历史数据呈现背景信息。 - [08:28](https://www.youtube.com/watch?v=blkGMZn_LBA&t=508s) 基于 AI 技术，Summer X Games 实现了日语、西班牙语、普通话和葡萄牙语的四语同步原声道直播。 - [16:24](https://www.youtube.com/watch?v=blkGMZn_LBA&t=984s) 未来观赛体验将实现个性化定制，观众可根据认知水平选择“初学者教学型”或“专业级”解说模式。 - [17:39](https://www.youtube.com/watch?v=blkGMZn_LBA&t=1059s) Owl.ai 已独立为初创公司并完成 1,100 万美元融资，计划推广至格斗、篮球、花样滑冰和体操等更多体育项目。 - [17:39](https://www.youtube.com/watch?v=blkGMZn_LBA&t=1059s) Owl.ai 的北极星目标是彻底消除裁判的人为失误，同时作为普及型“AI 教练”为数千万青少年提供专业训练指导。 ## Detailed walkthrough ### [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) 体育竞技中的误判困局与 AI 辅助裁决的雏形 在这场由 Sports Business Journal（《体育商业杂志》）记者克里斯·史密斯（Chris Smith）主持的对谈中，嘉宾杰里米·布鲁姆（Jeremy Bloom）展现了他极其多元的传奇履历——他既是两届奥运会自由式滑雪大跳台/雪上技巧（Mogul）名将、世界冠军，也是 NFL（美国职业橄榄球大联盟）第五轮选秀入选球员，同时还是一位成功的科技创业者（曾创立企业级软件公司并成功出售），如今则担任全球知名极限运动赛事 **X Games（极限运动会）** 的首席执行官。 布鲁姆分享了他从小建立的胜负欲与极致的自律精神。作为家中三胞胎/兄弟姐妹里最小的一个，他在前 10 年里几乎输掉了所有的比赛。为了战胜哥哥姐姐，他养成了“早到晚走”的拼搏习惯。12 岁时，他在父亲每年生日的例行采访中坚定地表示，自己未来既要参加奥运会滑雪，又要进入 NFL 效力。面对这两个截然不同的顶级竞技体育项目，他倾尽了所有青春与汗水——放弃了高中毕业舞会，甚至在半夜攀登雪山训练，只为在竞争对手沉睡时积累优势。 布鲁姆特别谈到了自由式滑雪这种“主观打分项目”的残酷性：运动员付出了整整一生的努力，最终决定命运的却只有仅仅 **22 秒** 的滑行。在这极短的时间里，运动员容不得半点失误（金牌与银牌的分水岭可能仅仅是空中分腿的角度差异）；但更具不确定性的是，现场的 7 位人类裁判同样面临巨大的挑战——裁判可能会疲劳、走神，甚至受到主观偏见或行业腐败的影响（如 2002 年盐湖城冬奥会花样滑冰裁判丑闻，以及最近 MLB 裁判因涉赌被解雇等事件）。布鲁姆指出，在当今科技高度发达的时代，整个体育产业（包括运动员、教练、赞助商和球迷）竟然依然在“默认并接受”裁判误判能够直接改变体育历史走向，这本身是非常不可思议的。 为了说明误判对体育产业产生的巨大灾难性后果，布鲁姆举了两个震撼人心的真实案例： - **2018 年 NFL 国家联赛季后赛（NFC 决赛）**：新奥尔良圣徒队（Saints）在即将锁定胜局进入超级碗的关键时刻，遭遇了一次极度明显的防守干扰犯规（Pass Interference），然而裁判却没有任何吹罚。这次致命误判直接导致圣徒队无缘超级碗，也彻底改变了无数训练了一辈子的球员的命运以及城市数以亿美元计的经济收益。 - **一级方程式赛车（F1）阿布扎比大奖赛**：赛事干事在收官战做出违规的调度决定，直接导致整个 25 站赛季积累下来的总冠军奖杯颁发给了错误的人。 此外，随着体育博彩业的合法化与爆发式增长，体育赛事的公正性与诚信度面临着前所未有的考验： - **体育博彩市场规模激增**：美国体育博彩的市场规模从 2018 年的 **4 亿美元** 狂飙到了去年的 **120 亿美元**。“想要找到腐败的根源，就去追踪资金的流向（Follow the money）。”布鲁姆强调，高额的利益沉淀使得哪怕是人类裁判最无意的失误，都会引发公众对赛事诚信度的剧烈质疑。 --- ### [00:08](https://www.youtube.com/watch?v=blkGMZn_LBA&t=8s) Owl.ai 的诞生：从与谢尔盖·布林的对话到顶级 AI 裁判 在 2023 年 12 月出任 X Games 的 CEO 后，具有 10 年企业级软件创业经验的布鲁姆敏锐地意识到，必须利用最前沿的技术来解决困扰体育界多年的客观裁判难题。他在旧金山生活期间，与谷歌联合创始人**谢尔盖·布林（Sergey Brin）**建立了深厚的友谊。布林本身也是一位极度热爱运动的极客，家中后院甚至搭建了高空飞人（Trapeze）设施，每周三还会邀请布姆及斯坦福教练一起切磋各项运动。 当布鲁姆将“重返体育界并打造全球首个 AI 体育裁判”的想法告诉布林时，布林极为赞赏，并在当周末便亲自敲下了该项目的第一行代码，开启了这项颠覆性创新——**Owl.ai**。随后，布林将布鲁姆介绍给了现居科罗拉多州博尔德的 AI 专家乔什（Josh）。 当时距离 2024 年 1 月 23 日举办的阿斯彭（Aspen）冬季 X Games 仅剩短短 **5 周** 空间。布鲁姆与 Josh 的团队基于 **Google Cloud Platform (GCP)** 展开了昼夜不停的开发极客冲刺（Sprint）。他们不仅要完成模型的训练与测试，还要将 AI 裁判的实时打分与分析直接呈现在 ABC 电视台的全球直播画面中。 在项目初期，外界普遍质疑：“AI 或许能精准判定球是否出界这种二元（Binary）结果，但它怎么可能理解主观的‘风格’（Style）？”对此，布鲁姆带领团队与赛事首席裁判及顶尖运动员深度合作，将抽象的“动作风格”拆解为了可量化的物理与视觉特征： - **空中动作效率（Economy of Motion）**：运动员在空中的姿态控制与身体平衡。 - **轴线控制（On Access）与起跳高度（Amplitude）**：转体是否精准、飞跃高度与滞空时间。 - **抓板细节（Grab & Tweak）**：是否在抓板的同时做出了身体的变形扭曲（Tweak）以展示极致风格。 经过短短 5 周的数据训练，Owl.ai 展现出了“博士级别”的精准打分能力。更重要的是，AI 裁判彻底消除了人类裁判不可避免的**意识与潜意识偏见（Bias）**：它不在乎面前比赛的是超级巨星肖恩·怀特（Shaun White），还是一位毫无名气的新人。这对于维护体育品牌的公信力（Credibility）至关重要。 除了公正打分之外，Owl.ai 在研发过程中还展现出了意料之外的惊人进化： - **自动生成专家级解说**：Owl.ai 开始能够像专业的 X Games 解说员一样，实时辨识出复杂的动作（例如精准区分是 Stale Fish 抓板还是 Mute 抓板，是 720 度转体还是 1080 度转体），并以极其自然且具有“极客风/原生语调（Endemic Voice）”的人声进行讲解。 - **多语言无缝实时同步直播**：Owl.ai 可以无缝将解说内容转化为多种语言。在盐湖城举办的 Summer X Games 上，赛事历史上首次通过 AI 技术实现了**日语、西班牙语、普通话（Mandarin）和葡萄牙语**的全球四语同步原声道直播（这对于滑板运动风靡的巴西和日本市场具有里程碑意义）。 ``` +-------------------------------------------------------------------+ | Owl.ai 工作流程架构 | +-------------------------------------------------------------------+ | v +-----------------------------------------------------+ | 1. 信号输入: 电视广播高清直播信号 (无需额外硬件/摄像头) | +-----------------------------------------------------+ | v +-----------------------------------------------------+ | 2. Google Cloud (GCP) 实时视觉与动作分析模型 | | - 评估空中效率 / 起跳高度 / 转体轴线 / 抓板细节 | +-----------------------------------------------------+ | +---------+---------+ | | v v +------------------+ +------------------+ | 3A. 客观打分与预测 | | 3B. 智能多语言解说| | - 消除主观偏见 | | - 原生极客语调 | | - PhD 级打分 | | - 多语种同步直播| +------------------+ +------------------+ ``` - **阿斯彭实测成果**：在冬季 X Games 男子单板滑雪比赛中，Owl.ai 的打分结果与人类资深裁判团的高度一致（仅有一处第 5 名与第 6 名的微小顺位差异）。 - **赛前预测黑科技**：Owl.ai 通过观察选手训练赛的表现，预测出了前三名的最终归属，其预测结果甚至与拉斯维加斯（Vegas）博彩盘口不一致，但最终比赛结果证明 AI 成功预测中了前三名的准确名单。 以下是根据您提供的逐字稿段落整理的详细内容记录： ### [16:24](https://www.youtube.com/watch?v=blkGMZn_LBA&t=984s) AI生成解说与个性化观赛体验 AI 在体育赛事解说中的应用不仅限于逐字逐句的简单翻译，更在于它能够以独特的方式理解并讲述运动细节。AI 能够深入掌握运动员的历史背景与比赛细微差别，提供富有深度和专业度的背景资讯。例如，AI 解说员在播报时可能会脱口而出：“斯科蒂·詹姆斯（Scotty James）在 2019 年的铜山（Copper Mountain）站比赛中凭借一场大胜破茧而出。”这种对历史细节的精准把握，甚至会让现场工作人员都感到吃惊与佩服。 此外，对于极端运动（Action Sports）这类具有门槛的专业项目，AI 能够为广大普通观众带来前所未有的个性化观赛体验。在不久的将来，观众在 Netflix 等平台观看极限运动会（X Games）时，可以根据自身的认知水平自主设定解说模式：初学者可以指定“教学型”解说以普及基础知识，而资深爱好者则可以选择“专业级”解说。这种个性化的解说定制服务，将大幅降低大众理解复杂体育项目的门槛。 ### [17:39](https://www.youtube.com/watch?v=blkGMZn_LBA&t=1059s) Owl.ai 的独立运作与消除裁判人头误差的终极愿景 Owl.ai 目前已从 X Games 和 Google 体系中独立出来，成为一家独立的初创公司。公司由 CEO 乔什（Josh）带领团队发展，并已顺利完成 1,100 万美元的风险投资融资。创始人兼执行主席杰里米·布鲁姆（Jeremy Bloom）指出，Owl.ai 目前正积极与各类体育组织洽谈合作，涵盖领域从职业格斗、篮球，到花样滑冰、体操等奥运评判类项目。 Owl.ai 的核心北极星目标（North Star）非常明确：彻底消除所有依靠裁判评分与执法的体育运动中的人为失误（Human Error）。除了解说辅助和数据分析外，该技术的终极愿景是成为每个人口袋里的“世界顶级教练”。仅在美国就有 6,000 万青少年参与体育运动，未来无论孩子还是家长，只需上传训练视频，就能获得媲美世界级教练的专业指导。 对于合作模式，杰里米强调公司非常愿意采用白标（White Label）形式与其他体育IP合作。他动情地表示，创立 Owl.ai 的初衷并非为了盲目追求高额的回报倍数（Big X），而是为了确保粉丝和运动员在 100% 的时间里都能获得公正客观的比赛结果。研讨会最终在对这一体育科技未来的期待与致谢声中圆满结束。 - ** Owl.ai 融资规模**：已成功募集 1,100 万美元风险资金。 - **目标覆盖项目**：职业格斗（Pro-fighting）、篮球、花样滑冰、体操等所有评分及裁决类体育项目。 - **受众规模数据**：仅在美国就有大约 6,000 万青少年（Youth Sports）参与各类体育锻炼。 - **公司核心宗旨**：通过技术实现 100% 的判罚准确率与公正性，保障运动员与观众的权益。 ## Notable quotes > [00:04](https://www.youtube.com/watch?v=blkGMZn_LBA&t=4s) 想要找到腐败的根源，就去追踪资金的流向（Follow the money）。 — Jeremy Bloom > [16:24](https://www.youtube.com/watch?v=blkGMZn_LBA&t=984s) 斯科蒂·詹姆斯（Scotty James）在 2019 年的铜山（Copper Mountain）站比赛中凭借一场大胜破茧而出。 — Owl.ai ## Mentioned Jeremy Bloom, Chris Smith, Sergey Brin, Josh, Scotty James, Shaun White, Sports Business Journal, X Games, NFL, MLB, F1, New Orleans Saints, ABC, Netflix, Google, Google Cloud Platform (GCP), Owl.ai ## Video description Christopher Smith, Staff Writer, Sports Business Journal Jeremy Bloom, CEO, X Games Abstracts The stakes in sports are higher than ever and sports leagues need to embrace tech or risk falling behind. One missed call can be the difference between going to the Super Bowl and going to the golf course. And, in the greater ecosystem, millions and millions of dollars are at stake - along with the integrity of the leagues. There are broader implications for AI on athlete development, media coverage and fan engagement. Olympic skier, NFL athlete, and seasoned tech entrepreneur Jeremy Bloom understands the stakes more than most and he presents a vision for how AI can elevate - not replace - human judgment and storytelling in sports. The X Games CEO is using AI to remove bias from judged sports and enhance fairness through the introduction of Owl.ai—an AI platform trained to score action sports like a superhuman judge as it brings unprecedented fairness, transparency and immersion while enhancing fan engagement and education to the global sports ecosystem. [Source](https://www.youtube.com/watch?v=blkGMZn_LBA)

[0:04] Jeremy Bloom: 想要找到腐败的根源，就去追踪资金的流向（Follow the money）。 [16:24] Owl.ai: 斯科蒂·詹姆斯（Scotty James）在 2019 年的铜山（Copper Mountain）站比赛中凭借一场大胜破茧而出。

[0:04] Jeremy Bloom: to win, he had to show up early, stay late, and outwork everyone else. [0:04] Jeremy Bloom: healthy disregard for the impossible [0:04] Jeremy Bloom: the difference between gold and silver is a single split of the legs. [0:04] Jeremy Bloom: tired, distracted, or prone to errors, corruption, and bias. [8:01] Jeremy Bloom: today a 10-year-old Japanese athlete can land three consecutive 900s in a single run. [16:24] Scotty James burst onto the scene in 2019 with a huge win in Copper [16:24] Beginner level: The AI acts as an educational guide, breaking down complex terms and teaching the rules and mechanics of the sport in real time. [16:24] Endemic / Expert level: Delivers high-level technical commentary geared toward hardcore fans who understand advanced tricks and terminology. [17:39] completely eliminate human error from every judged and refereed sport, guaranteeing that athletes and fans receive the correct outcome 100% of the time. [17:39] the world's best coach in your pocket.

[0:04] My name is Chris Smith. I'm a reporter [0:05] with Sports Business Journal and today I [0:07] have the honor and pleasure of being [0:08] joined by Jeremy Bloom, one of those [0:10] guys who really doesn't need an [0:11] introduction, but I'll give him one [0:12] anyway. Uh, and one of the true [0:13] multihyphenets, which I feel like is an [0:15] overused term these days. Uh, let's see [0:18] if I can nail them all. world champion [0:19] mogul skier, two-time Olympian, uh fifth [0:22] round NFL draft pick, then tech [0:24] entrepreneur, uh soldier company, now [0:26] the CEO of the XG games, and we'll be [0:28] talking today about Owl AI, new uh new [0:31] AI business you, uh have spun up this [0:33] year. Um so, thank you so much for [0:35] joining us. Before we dig into the AI [0:36] stuff, I think maybe it's important, you [0:38] know, help set the table here, right? [0:39] Because what you're doing with Owl is [0:41] really largely focused on solving some [0:43] problems in sports for athletes, for [0:45] fans, problems that you encountered as [0:47] an athlete. Uh so maybe walk us through [0:49] a little bit, you know, as a skier, as a [0:51] football player, you know, some of the [0:52] challenges you encountered that led you [0:54] to, you know, think through, you know, [0:55] there's an opportunity here if you apply [0:56] some tech, you know, can kind of create [0:58] some solutions with some of the new uh [0:59] possibilities that we have. [1:00] >> Sounds good. Good to see you and thank [1:01] you for for being here this morning. Um [1:04] raise your hand if you played sports, [1:06] elementary school, middle school. So a [1:08] lot of athlete, we're amongst athletes [1:10] in the room. I think we would all agree. [1:12] Um, you know, I grew up in uh in a [1:14] family I was the youngest of three and I [1:17] spent the first 10 years of my life [1:19] losing to my brother and sister at [1:20] everything. So, like it was really the [1:22] that the the first 10 years of my life [1:24] where I'm like, "Okay, I got to do [1:26] things differently. This losing thing [1:28] sucks." like my brother and sister would [1:30] beat me at the 5k race and at skiing and [1:32] at hiking and whatever the card games [1:36] and and so I you know I sort of created [1:39] the framework where it's like okay well [1:40] if I ever want to beat my brother and [1:42] sister or anything I I better show up [1:43] early and I better stay late. Nothing no [1:47] big secret no no magical supplement uh [1:50] but just really knew that I had to work [1:52] a lot harder than than them. And um I [1:56] always loved football and I always loved [1:58] to ski growing up in Colorado. And my [2:00] dad used to interview us on our [2:02] birthdays every year. And he would [2:04] always ask, "What do you want to be when [2:05] you grow up?" And when I was 12, I said, [2:08] "I want to ski in the Olympics and and [2:10] play in the NFL." And I think both my [2:12] parents have a healthy disregard for the [2:14] impossible. And so they said, "You can [2:16] do that, you know, if you put your mind [2:17] to it and you attack it." And so that's [2:19] what I did. And so at a very young age, [2:21] I really applied everything that I [2:23] possibly could, every second of the [2:25] waking hour to try to become an Olympic [2:28] skier and try to make it to the National [2:30] Football League. Two very different [2:31] sports. And that's what athletes do. I [2:35] mean, when when when you make that [2:37] decision that you want to be a [2:39] professional, that you want to be the [2:40] best in the world, you sacrifice [2:42] everything else. So, you don't go to [2:45] proms and homecomings cuz you're [2:46] training. uh you wake up in the middle [2:49] of the night and you go climb a mountain [2:50] because you know your competitors are [2:51] sleeping and you dedicate everything [2:54] that you have at a very young age to [2:57] this dream and that's what what what I [2:59] what I did and what every anyone who [3:01] reaches the Olympics or the NFL or any [3:04] major league that's the type of [3:05] dedication that you have to apply and in [3:09] skiing the sport of freestyle skiing [3:11] which is bumps jumps flips and all these [3:13] crazy things it's a really crazy sport [3:15] because you train your entire life for [3:18] 22 seconds. [3:20] That's it. 22 seconds is the time when [3:23] they say go until you cross the finish [3:25] line. And so whether it was the Olympics [3:28] or World Championships or World Cups, [3:30] you had 22 seconds to to do that. Now, [3:33] inside those 22 seconds, you obviously [3:36] have to be absolutely perfect to win. [3:38] You cannot make a single mistake. The [3:40] difference between a gold medal and a [3:42] silver medal is that it's one split of [3:44] the of the legs. And not only did I have [3:47] to be perfect, the seven human judges [3:49] that were judging that competition had [3:51] to be dialed in. They couldn't miss a [3:54] moment. And what you would find [3:57] oftentimes in subjective sports like [3:59] skiing is that human judges aren't [4:02] always dialed in. And sometimes they're [4:04] set up for for failure. Maybe they're [4:06] tired. Maybe they had an itch [4:10] or maybe there was corruption at work. [4:12] We saw, you know, for those of us old [4:14] enough to live through the 2002 [4:15] Olympics, we saw the Russian figure [4:17] skating scandal. Uh just last month, [4:19] Major League Baseball fired an ump for [4:21] sports betting. Um and so there's all [4:24] these like all this human error built [4:26] inside of sports. But these moments are [4:29] too important for athletes, for fans uh [4:34] to miss. And we still to this day with [4:38] all the technology around us, we live in [4:40] a world where it is okay, we make it [4:43] okay for referees and for judges to make [4:47] mistakes that that inaccurately [4:50] determine the outcome of these massive [4:52] moments. [4:54] We make it okay. It's still okay. [4:58] That's crazy to me. That is absolutely [5:01] insane that an athlete, a team, a coach, [5:04] a fan base, an ownership group, all [5:06] these sponsors would be okay with a [5:08] referee who clearly made a wrong call [5:12] and changed the course of athletic [5:14] history. And I'll give you a couple [5:16] examples and some some of which you may [5:17] remember. The 2018 NFC Championship [5:22] game, a game every child dreams about [5:26] playing. You win that game, you go to [5:28] the Super Bowl. [5:30] the Saints inside the 20 about to win [5:33] the game. Kick a field goal, win the [5:35] game, run out the clock. There was a a [5:39] no call. It was a pass interference [5:40] call. The guy literally hit the receiver [5:44] before the ball even came [5:46] near him. No call would have ended the [5:50] game. The Saints would have gone to the [5:51] Super Bowl. [5:53] The Saints didn't go to the Super Bowl. [5:55] They lost that game through a bad call. [5:57] And the NFL admitted it. the referees [5:59] admit. I mean, it was the it was the [6:00] most obvious non- call you've ever seen. [6:03] We should never allow that to happen. [6:07] Are there any sports betterers in the [6:08] room? [6:10] Okay. Well, there's a couple here. [6:12] There's a lot more outside of this room. [6:14] So, in 2018, sports betting was a $400 [6:17] million industry in this country. What [6:20] do you think last year, how much money [6:23] was spent on sports betting? 400 million [6:26] in 2018. Any guesses? Anyone want to [6:28] take two billion, [6:31] four, what did you say, five? 12. [6:37] So 400 million to 12 billion in a very [6:41] short period of time. And it's like [6:43] we've always like, you know, you want to [6:45] know where the corruption is? Well, [6:46] what? Follow the money, right? So even [6:49] at the best the best case possible, that [6:52] ref just made a bad call. That happens, [6:54] right? We're humans. We all make [6:56] mistakes. [6:58] But think about being one of those [7:00] players who trained your entire life, [7:04] your entire life to get to the Super [7:06] Bowl and now that player's retired or [7:08] out of the NFL and they should have [7:10] played in the Super Bowl and they [7:11] didn't. Or imagine being a fan of the [7:12] New Orleans Saints [7:15] and you missed out on that or or the [7:16] ownership. I mean, you're talking [7:18] hundreds of millions of dollars or [7:21] tourism for your city. Like it the the [7:23] stakes are way too high. And it's not [7:25] just the NFL. [7:27] Any F1 fans? [7:29] Okay, so you probably remember the I [7:31] think it was 2019 or around there, the [7:32] the the last race in Monaco, F1, where [7:35] the race director made the wrong call. [7:36] They stopped the race, allowed a driver [7:39] to advance, which was not per the rule [7:41] books, and literally at the end of the [7:43] season, the wrong driver hoisted the [7:45] trophy. [7:46] Are you kidding me? Like at the end of a [7:48] 25 race season, the wrong driver hoisted [7:51] the turf because the race director had a [7:53] bad day. [7:55] We can't live with these moments [7:57] anymore. We cannot live. These moments [7:59] are just too important. [8:01] >> So let's refocus this then on what [8:02] you're working on. Uh so December you [8:04] take over as CEO of the X Games. You are [8:06] overhauling this business in a lot of [8:08] different ways. Yeah. So maybe walk us [8:09] through a little bit of of the [8:10] opportunity you see with that property, [8:12] but at what point along the way do you [8:14] start to realize that, you know, there [8:15] is the possibility of using AI, using a [8:17] tech solution to start trying to figure [8:18] out, you know, some of these problems [8:19] that uh clearly have been keeping you up [8:21] uh late at night? [8:22] >> Yes, they have. Can you tell? Anybody [8:24] heard of the X Games? [8:26] >> Cool. That's awesome. I mean, you know, [8:28] the X Games been around for 30 years. [8:30] You might remember Travis Pastana's [8:32] double backflip, first ever in Staple [8:34] Center. You might remember Levi Lavali's [8:36] first ever backflip on a snowmobile of [8:39] all things. Some of you might remember [8:42] Tony Hawk, first ever 900. What if I [8:45] told you there's a 10-year-old [8:47] 10-year-old from Japan in X Games Today [8:49] who does three straight 900s on the vert [8:52] ramp? He's 10. That's how far our sports [8:54] have come. Um but ever you know so when [8:57] I left sports as you mentioned I got [8:58] into tech and started an enterprise [9:00] software company and sold it to private [9:02] equity in 2021 and was recruited back to [9:05] sports which I'm delighted about and of [9:07] course as I came back having spent 10 [9:09] years on the enterprise software side of [9:11] things and being really curious about [9:14] this whole AI stuff um I started [9:17] thinking about this problem and I lived [9:19] in San Francisco for four years and I [9:22] became close friends with one of the [9:23] co-founders of Google Sergey guy who [9:26] happens to be a big athlete himself and [9:28] he's got a trapeze in his backyard and [9:30] he every Wednesday we would go to [9:32] Stanford and we'd play a different sport [9:33] and then the Stanford coach would coach [9:35] us with a small group like we he's a so [9:37] I I I called Sergey and I said hey I'm [9:39] going back to sports I'm really excited [9:41] about it and he said well that's awesome [9:44] X Games is amazing like what should we [9:45] do together I said funny you should ask [9:48] let's go build the world's first AI [9:50] judge and let's use whatever technology [9:53] you know about cuz clearly I wouldn't [9:55] know about it. Like what do you what [9:57] what do you know that Google is working [9:59] on that could solve this problem that we [10:01] we just talked about it? Could we solve [10:03] it? Like could could that change the [10:06] course of history and make sure that [10:08] every sport regardless of if it's judged [10:11] or refereed that the right outcome was [10:14] always driven towards and he's like yeah [10:17] I think we could. So he started coding [10:19] it that weekend which was pretty cool. [10:20] The first lines of code was Sergey, sort [10:22] of iconic. And then Sergey introduced me [10:25] to a guy named Josh. [10:28] And Josh happened to live in Boulder, [10:29] Colorado, where I live. And he's sitting [10:32] right here in the front row. Uh, but you [10:34] know, Josh and I met, and we had five [10:36] weeks before the Aspen Winter X Games, [10:39] which was January 23rd, and I had just [10:42] taken the job. We had five weeks not [10:44] just to beta test the technology. We [10:46] were going to put it live on air on ABC. [10:50] And so it was this really fun fiveweek [10:52] sprint of with Josh and Josh was running [10:55] AI inside of GCP Google Cloud and so we [10:58] had all these really credible AI minds [11:01] working on this problem and we focused [11:03] on men's snowboarding [11:06] and we started training the model and [11:09] one of the biggest questions that you [11:10] might be asking yourself right now how [11:13] can AI understand style right of course [11:16] AI can understand binary outcomes like [11:18] was it a ball was it a strike [11:20] Was the tennis ball in or was it out? [11:23] But a lot of people struggle like how [11:26] could an AI possibly understand style [11:28] and all of our sports at X Games are [11:29] stylebased and that in lies the fun [11:32] challenge because AI turns out on the [11:35] binary stuff super simple like with the [11:37] cutting edge technology like super [11:39] simple but we as we unpacked it we [11:42] brought in our head judge we brought in [11:44] a lot of the athletes I have a [11:46] background as well and and we started [11:47] unpacking what is good style and it [11:49] comes down to just what it's it's good [11:51] economy emotion in the air what has good [11:53] econom economy of motion in the air. Are [11:56] they on access? What's the amplitude of [11:58] the jump? If there's a grab, is there [12:01] also a tweak which shows good style? And [12:04] so it turns out you actually can unpack [12:06] what good style is. And over that [12:08] fiveweek period, the AI judge became a [12:12] PhD level judge incredibly. And it [12:15] didn't care if it was Shawn White or if [12:17] it was someone you've never heard of, [12:19] which is an also another important point [12:20] of conscious and unconscious bias. If [12:24] you're football fans, you remember the [12:25] prevailing narrative of this past [12:27] playoff season was, "Well, the refs are [12:29] in the pockets of Patrick Mahomes, [12:31] right?" How many times did we hear that? [12:33] That's really bad for a brand. That's [12:36] really bad for the NFL because you [12:38] credibility needs to be at the core of [12:40] every brand. And if you're not credible, [12:42] you're no longer a real sport brand. And [12:45] and so it we we started training it and [12:47] it became a PhD level co uh judge and it [12:50] would score the athletes incredibly [12:52] well. Then it started doing things that [12:55] well we didn't even ask it to. It [12:58] started commentating the events like an [13:00] exame commentator. And our sports are [13:03] really hard to understand. Was it a [13:05] stalefish grab or a mute grab? Most [13:08] people don't know. Was it a 720 or a [13:10] 1080? And it started doing it in endemic [13:13] voices. Sounded like a human, like an [13:16] XGames human. And then you could say, [13:19] well, now do it in Japanese. Now do it [13:22] in French. Now do it in Spanish. And [13:24] that's really relevant to XG Games. We [13:26] have athletes who compete in our [13:27] competitions from over 25 different [13:29] countries. And so for the first time [13:32] ever in Salt Lake City, we using this [13:34] technology, we we simultaneously [13:36] broadcasted our event in Japanese, [13:39] Spanish, Mandarin, and Portuguese [13:41] because SummerX Games skateboarding is [13:44] so big in Brazil and so big in Japan. [13:47] And so now as we've sort of launched and [13:49] developed this technology, not only does [13:51] it show incredibly promising signs to [13:54] bring objectivity to subjective sports [13:56] like X games, but but like the NFL and [13:58] like soccer, but it can also bring [14:01] global fans closer to these games by [14:04] making these sports available in their [14:06] local language, which at X Games had [14:08] never been been done before. [14:09] >> So I want to unpack a few things you [14:11] just want, you know, you just dumped a [14:12] lot on us there. So [14:14] first take us through how did you [14:16] implement that judge at X games Aspen [14:18] right five weeks you have this you know [14:20] automated judge how do you work that [14:22] into the broadcast how do you work that [14:23] into the judging such that it's actually [14:25] having an impact on on the games so in [14:27] Aspen because it was only five weeks [14:29] first thing to note you don't need any [14:30] spec special hardware you don't need any [14:33] special cameras just took the live feed [14:35] from the from the TV broadcast and Josh [14:38] and his team were there and Google was [14:40] there because this was a big Google [14:41] project at the time and we we it judged [14:46] every single competitor and we would put [14:48] it on air. Now, we didn't give out [14:49] medals in Aspen based off the judge from [14:52] Owl. It was a test, but we wanted it to [14:55] be on air. But what we found is it was [14:58] remarkably uh consistent with the human [15:00] judges. I think there was only one [15:01] example where a fifth went to sixth and [15:04] a sixth went to fifth. And then we also [15:07] asked it to do some predictions. So we [15:10] asked Owl to watch practice of these [15:13] snowboarders and then we asked it to [15:15] predict first, second and third and um [15:19] we put that on air so the commentators [15:21] threw owl and it's got a human voice [15:23] like you wouldn't like it just hey let's [15:24] go to the owl and like it sound like a [15:26] human and it went through its top three [15:28] prediction and why based off practice [15:30] and what it knew. Now its predictions [15:33] went against the Vegas odds which I [15:35] thought thought was relevant like oh [15:37] that's interesting. It didn't it didn't [15:38] have the top three that everyone thought [15:40] that was close. And again, it didn't [15:43] judge the competition. The humans did. [15:45] And the humans didn't even know that it [15:46] was, you know, judging. Um, and it got [15:49] it right first, second, and third, which [15:52] is pretty remarkable. Now, I'm not [15:54] suggesting that this is going to be [15:55] accurate all the time. In fact, and saw [15:56] it like it wasn't. These are very [15:58] difficult sports to predict. But um it [16:02] was really interesting because it was a [16:04] you know a really thoughtful readout as [16:05] to to to why it predicted what it did [16:08] >> on the commentating side. Walk us [16:09] through that a little bit more. I mean [16:10] what does that mean when you say started [16:12] commentating right and when you talk [16:14] about you know the broadcast you guys [16:15] did from Salt Lake you know the [16:16] international broadcast is that you know [16:18] the entire audio of those broadcasts is [16:20] AI generated. [16:21] >> The entire audio of those four languages [16:24] were AI generated. Okay. [16:26] >> And you wouldn't be able to tell if [16:28] unless you were really looking for it. [16:30] And I think what's interesting about [16:32] this technology, it doesn't it has the [16:34] ability not just to take the English [16:36] broadcast and translate word by word. It [16:39] has the opport the ability to do it in a [16:41] in its own unique way. And so the things [16:44] that it would say is like, oh, Scotty [16:45] James burst onto the scene in 2019 with [16:47] a huge win in copper. And even us were [16:50] like, wait, he won in 2019 in copper. So [16:53] it, you know, again, it has the [16:54] opportunity to really get to know the [16:56] nuances and the history of these [16:57] athletes and provide a level of [17:00] sophistication and background that's [17:01] that's that's helpful to the audience. [17:03] The other thing that it it can do, [17:05] especially for, you know, endemic sports [17:07] like us, if you're a casual fan and [17:10] you're watching on Netflix in a year or [17:12] two, you're going to be able to say, [17:14] "This is the type of commentary [17:16] commentary I want at X Games. I'm a [17:18] beginner to these sports. commentate [17:21] these sports so that you can teach me [17:22] about these sports or you can say I'm a [17:25] novice or you can say I'm endemic and [17:27] eventually not too too far we will all [17:31] be able to program what type of [17:33] commentary we want to hear at at sports [17:35] especially in difficult sports to [17:37] understand like xgame sports [17:39] >> so we got a few minutes left I want you [17:40] to give us your vision for the future [17:42] right you guys stand up owl is a [17:44] standalone company now right this year [17:46] um so no longer part of the x games or [17:48] Google right it's its own [17:50] Um, you are now looking at different [17:51] sports. You're looking at different [17:53] properties within your sport. Um, walk [17:55] us through a little bit of maybe I don't [17:57] know how much you can share us [17:58] detail-wise in terms of who you're [17:59] talking to, but the types of people [18:00] you're talking to and and maybe the [18:01] kinds of solutions that you're [18:02] envisioning. [18:03] >> Yeah. So, I'm, you know, I'm founder and [18:05] executive chairman, so I'm not [18:06] day-to-day on the company. Josh is the [18:08] CEO. He's building a team. We raised 11 [18:10] million of of venture capital. And you [18:13] know, Josh and Tim are in conversations [18:15] from everything from pro- fighting to [18:18] basketball [18:20] uh to all the Olympic sports like figure [18:22] skating and gymnastics. And really the [18:24] goal is simple, to absolutely remove [18:27] human error from every judge to referee [18:30] sport. That is the goal and that's our [18:33] northern star. and some of the other [18:35] things like commentary or insights and [18:37] eventually coaching. Where I ultimately [18:39] see this going is this will be the [18:41] world's best coach in your pocket. [18:43] There's 60 million kids that play you [18:45] sports in this country alone. Imagine [18:47] their parents or those athletes having [18:49] the ability to have worldclass coach [18:52] judge literally coaching every move [18:54] where they upload their videos to. So [18:56] ultimately I think it it reaches the [18:58] consumer market but right now it's [19:00] really focused at removing human error [19:02] from the judge and refereed sports [19:03] >> and you want to essentially white label [19:05] it for for working with these other [19:06] properties. [19:06] >> We'll white label. We don't Yeah, sure. [19:08] We happy to white label as long as we [19:10] remove human error. Like the goal in [19:12] this company is not to make money. The [19:14] goal in this like you usually start a [19:15] company you want a big X. It's like the [19:18] goal of the company is to make sure that [19:20] the fans and the athletes get the right [19:22] outcome 100% of the time. That is the [19:25] goal and I think we can I think we can [19:27] accomplish it. [19:29] >> I think we can all agree we're looking [19:30] forward to that future. I was told I [19:32] have to stick to time and we are out of [19:33] it. But Jeremy, thank you so much. [19:34] >> Thank you.