Yes - compute, when done well, is a config param that could also use another 0

The Pitch: 1. Humans are our only proof point of general intelligence 2. If you want to guide AI research, pinpoint what humans can do but AI cannot 3. Once you can no longer come up with problems that humans can do and AI cannot, then you have AGI

.Greg Kamradt kicking off the first ever meetup at LangChain's new HQ! Packed house!

.Sam Whitmore crushing it at AI Memory She is stepping through the evolution of memory she’s built at New Computer Recording coming soon - absolutely gold

.Yash Gaitonde is the lead engineer behind Cursor memories Stepping through how they’re prototyping multi agent memory when users still like control “I’ve spent weeks on the prompting”

Thanks to Greg Kamradt for organizing a great event, and to Sam Whitmore, Yash, and Nicole Hedley for speaking!

this was so much fun! thanks for organizing, Greg Kamradt !! excited to share recording soon :)

really enjoyed the talks from WFH Sam Whitmore Yash Gaitonde and Nicole Hedley - thanks for organizing it Greg Kamradt and LangChain! here's 3 things i found very useful from what they had to share 🧵

You can just believe SF. Just attended an incredible AI memory meetup today I’ve always believed two key factors drive AI product capabilities: multi-agent architecture and memory architecture. Today @LangChain and Greg Kamradt hosted Sam Whitmore from New Computer and

Wait, you’re using the economy to define intelligence?

who wants to co-host this in NYC?

attended a gem of an event organized by Greg Kamradt at LangChain last night. pretty rare to hear direct insights from top builders (cursor, langchain, newcomputer, headstart) how they actually get LLMs to "remember stuff". my raw notes 👇 1/8