/ About this event
Voice AI has a problem: the smartest models think too long. The fastest models don’t think enough. Humans expect conversation to happen in real time. But the models capable of the deepest reasoning are often too slow for natural dialogue and the models fast enough to keep up can struggle with the intelligence, judgment, and nuance the moment demands.
So how do you build a model that can think deeply and answer instantly? In Healthcare this is tablestake, life and death.
Join research scientists from leading Voice AI start ups for a fast, unscripted conversation about one of the coolest and hardest problems in model training today: pushing intelligence and latency forward at the same time.
Learn in advance: https://hippocraticai.com/llm-research/
One demo. No slides. No hackathon. Just 25 minutes on the technical problem, what’s changing, and where the breakthroughs might come from.
Then we open the bar. Host drinks for engineers, researchers, and model trainers who want to keep the conversation going.
Come talk models, training, speed, intelligence—and what it is taking to make AI truly conversational.
This event is a part of #SFTechWeek—a week of events hosted by VCs and startups to bring together the tech ecosystem. Learn more at www.tech-week.com.
