This is an excerpt from Alex Heath’s Sources, a newsletter about AI and the tech industry, syndicated once a week to Verge subscribers.
Reinforcement learning (RL) is the next frontier, Google is growing, and the party scene has completely gotten out of hand. They were through the lines of this year’s Neurops in San Diego.
NEUROPS, or “Conference on Neural Information Processing Systems,” began in 1987 as a purely academic affair. It has since ballooned the hype around AI into a massive industry event where labs come to recruit and investors come to find the next wave of AI startups.
I was sadly unable to attend Neurops this year, but I still wanted to know what on earth people have been talking about in San Diego over the past week. So I asked engineers, researchers, and founders for their take. The list of answers below includes Andy Konivinsky, cofounder of Databricks and founder of the Loud Institute. Thomas Wolfe, cofounder of Hugging Face; The Rune of Openei ; And participants from Meta, Vimo, Google DeepMind, Amazon, and a handful of other places.
I asked everyone the same three questions: What is the biggest theme of the conference? Which labs look like they’re growing or struggling? Who had the best party?
The consensus was clear. “RLRLRLRLRL is taking over the world,” Lamarina CEO Anastasios Angelopoulos told me. The industry is converging around the idea that tuning models for specific use cases, rather than scaling the data used for pre-training, will drive the next wave of AI advancements. What’s clear from Lab’s speed question is that Google is having a moment. “Google DeepMind is feeling good,” the huggable-faced wolf told me.
The party circuit was naturally unbridled. Konivinsky’s Loud Lounge emerged as one of the week’s hotspots — with Jeff Dean, Yoshua Bengo, Einstuka, and about a dozen other top researchers coming through. An invitation-only cruise with 200 researchers, one of the cruise organizers, Nathan Lambert, told me. Ron was dry about the whole scene: “You can learn more from Twitter than literally being there … mostly my ground sense of ‘It’s too much.’ ‘
What attendees had to say about Neurops this year:
What was the oldest topic among participants that you think more people will be talking about in 2026?
Which labs feel like they’re picking up speed, and which feel more greedy?
What’s the best party you’ve ever been to or that busted the FOMO?
Yes, some people thought keynotes were a party. I think academia lives in Neurops after all.
