Why intent prediction needs more than an LLM
About this episode
Ryan discusses with Frank Portman the limitations of next-token prediction for forecasting human behavior and how Yobi uses alternative models.
Ryan sits down with Frank Portman, CTO at Yobi, to talk about why next-token prediction, though great for language, isn’t the right inductive bias for forecasting human behavior. They discuss how Yobi builds a “foundation model of behavior” using transformers and graph neural networks instead of chat-style LLMs, and what it takes to run millions of personalization decisions per second while keeping consumer data private. Episode notes: Yobi is a behavioral AI company building foundation models that predict future behavior for ad tech, marketing, and more. Connect with Frank via fportman.com or at yobi.ai . Congrats to Hooked on winning a Populist badge for their answer to Removing whitespace around a saved image . See Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info .
People in this episode
Host: Ryan
Guest: Frank Portman
Topics covered
- intent prediction
- behavior forecasting
- machine learning
- personalization
- data privacy
Keywords
- intent prediction
- LLM
- transformers
- graph neural networks
- behavioral AI
- personalization decisions
- consumer data privacy
Mentioned in this episode
Organizations: Yobi, fportman.com, yobi.ai, Hooked
More episodes of The Stack Overflow Podcast
- The future of development is full-stack
- Developers who move fast still need to do it together
- Your AI is only as responsible as you are
- Building more than just an agent harness
- What's left for infrastructure-as-code after AI moves in?
- Agent orchestration is so two-years ago
Explore listener stats, chart rankings, contacts and more on the The Stack Overflow Podcast podcast page.