RAG, AI Agents, and VC Insights on AI
Catch up on our latest publishes and all that's new in the AI research sphere.
Hey Cogsers,
Happy Sunday. Let’s dive into our latest! In this edition:
A guide to our recent publishes
This week’s episodes with Anton Troynikov of Chroma, and Yohei Nakajima of BabyAGI and GP at Untapped
Nathan’s Reads: Research Edition!
First, a guide to some of our recent publishes if you don’t know where to begin.
Want to catch up on all things OpenAI, Amazon & Anthropic, and Google Deepmind?
Watch our deep dive and live player analysis with Zvi Mowshowitz here
Curious about AI Security, what it takes to jailbreak an LLM, and defenses against jailbreaks?
Listen to Part 3 of The AI Scouting Report here, where Nathan dives deep into AI security, jailbreaks, and defenses
Listen to our interview here with Adam Wenchel, CEO of Arthur.ai, an AI security company with LLM-specific defense techniques
Been wondering what an AI engineer is, or what’s going on with the AI Pendants you may have seen on Twitter?
Watch our episode with Swyx on AI engineers, how to hire for them, how to become one, and all the rage on AI pendants here
This week on The Cognitive Revolution, hear from:
Anton Troynikov, CEO of Chroma
Yohei Nakajima, creator of BabyAGI and GP at Untapped Capital
Anton Troynikov - Nathan’s Notes
Anton Troynikov is the founding CTO of Chroma. We first had Anton on the show back on Episode 5 which was released in early March, before GPT-4 release, and he impressed me then and since as a real intellectual force.
He's someone for whom the fundamentals come easily, who codes all the time, and for whom it's natural to think in higher dimensions and abstractions.
He's also a super quick wit, a prolific and at times provocative commenter, and – amazingly – at the end of the day still a bit of a doubter – certainly more of a doubter than I am – when it comes to LLM reasoning abilities.
In March, I treated him as an expert tutor on core concepts of embeddings and retrieval, and tried to explore how he thinks about navigating the latent space as a way to develop my own intuitions. If you haven't heard that episode and want something more foundational from Anton, definitely check that one out.
This time, I wanted to talk about everything that's happened since, including the fact that RAG, or retrieval augmented generation, has been LLM's first big commercial application hit, with tons of companies spinning up all manner of document-backed bots for all sorts of labor-saving purposes, and in the process nearly as many spinning up and deploying a vector database for the first time. We talked about what's working today and also what's next, including for Chroma as a business.
Coming in, I was honestly wondering if Chroma and other vector DB startups might struggle to grow businesses given that incumbent DB providers are also racing to implement vector stores, but Anton did a good job of reminding me of one of my own maxims – that there's a good chance we're all still thinking too small. His observation that most of the data stored in Chrome has never been in a database before suggests that there will be plenty of growth to go around for a while yet, at least at the infrastructure layer, and his plan to bring more and more value into the database so that retrieval becomes as simple as dumping text into a textbox, not unlike how people use chatbots at present, but for memory, seems like one that can drive Chroma to a great outcome even as lots of incumbents get into the game.
Finally, I'm quite proud of this episode – it's a focused but fast-moving conversation conducted at an expert level, with a genuine intellectual leader, currently working as a wartime CTO.
I've learned an incredible amount from the process of making this show, and I'm grateful that thousands of you, from an incredibly diverse range of backgrounds, have become regular listeners.
I also want to thank the team at Turpentine including our Producers Vivian Meng and Natalie Toren, our editor Graham Bessellieu, and of course Erik for putting the whole thing together. Working with Turpentine does make for an unbelievably convenient operation, in which I am 100% focused on understanding, communicating, and interviewing as well as I possibly can.
Yohei Nakajima - Nathan’s Notes
Yohei Nakajima is a Venture Capitalist, founder of Untapped Capital, who's become best known over the last year for what he's personally built, in public, originally mostly for himself, but ultimately inspiring 10s of thousands with the potential of AI automation and agents.
Yohei, unreasonably modestly, describes himself as "lazy", but after this conversation, and watching all his project releases, I think his defining qualities are really curiosity, playfulness, and a love of exploration. Over the course of 70 public releases, he's helped the entire AI community get a glimpse of the sort of highly personalized AI-powered conveniences that seem destined to raise standards of living for all, and as much and perhaps perhaps more than any other creator or builder I follow, it's always been very clear that he's having a ton of fun along the way.
We cover a lot of ground here, starting with his recent TED AI talk, then running through the many apps that he's built and how he goes about it, before finally digging into how he thinks about the AI space as an investor.
The first half of this conversation is a great study for any aspiring AI operations specialists, especially considering what a key role no-code proficiency has played in Yohei's journey, and I personally found a lot of value in his answers to my rapid-fire investment framework questions in the second half.
Overall, though, if there's one thing everyone should take away from Yohei, it's the way that he's delighted in the process of incorporating AI into his life in a way that serves his own idiosyncratic and ever-evolving purposes.
If you find value in the show and want to help it grow, I'd suggest sending this episode to a person you know who tried ChatGPT and still says they couldn't figure out what they'd actually use it for.
📚 Nathan’s Reads: Research Edition
All of the following papers are covered in our research deep dives! Part one linked here. Part two linked here.
As always, thank you for listening! We’re in the podcast arena, always looking to improve. If you’ve got feedback, let us know by sending an email to tcr@turpentine.co or DMing Nathan on the social media platform of your choice.
Also, we have a Discord now! Hop in and say hello to other listeners of the show here