Building the future, a founder's perspective
summary
Built and sold an AutoML deep‑learning company and then joins DataRobot after an acquisition; argues industry has an 85% failure rate because models rarely deliver business value. Shows concrete builds: a reinforcement‑learning Xbox bot trained on a server with 12 TB RAM 196 Xeon cores that achieved autonomous kills, an AI influencer detector that scaled sourcing and unlocked new customers, and prototypes for pool drowning detection and continuous ICU heartbeat monitoring — some worked and transformed workflows, many smaller projects did not.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:38Snowboarder to semiconductor engineer
- 8:30Startup struggles and acquisition
- 11:58Defining intelligence across species
- 12:40AI playing Xbox and emotions
- 22:54Why AI models fail in industry
- 26:49Transformational AI project criteria
- 30:20Future of AI in daily life
- 35:34AI unlocking human potential
- 37:43Q&A
related talks
26mbuilding reliable AI systems in production
23mDelivering business success with AI
56mData, AI, Hype and Fear - Recipe for Disaster?
33mExecutive Summit Roundtable Discussion
25mBuilding A Successful Data Science Team
20mWhat's changing and what hasn't in AI
26mCompleting the machine learning loop
42mAI-powered products and product-powered AI