This story made me very happy. Lol

50-100? How do you do this and for what purpose?

For eg. on a large codebase I do hierarchical review where higher level agents focus on modules and leafs on files. Each reports upwards a summary of its findings incl children summaries.

I don't know how one would build anything without knowing. If it was true, I would be building a competitor to Anthropic right now.

You build without knowing by delegating to a system that extrapolates what it knows about you and what you want, and is able to execute on that to deliver faster, better, and more completely than you ever could. You'll live as a ball of intent and values, observing in wonder as everything you desire springs up around you before you know you want it. You could choose not to live like this of course, but it would be like driving a DIY go-kart on the highway - you'll fall behind and get in others' way, and the rest of society will treat you accordingly.

That's the optimistic case, anyway.

I assume you don't ascribe to the thought that friction leads to growth. In this future of complete frictionless existence wont humanity stagnate? In this world of no effort, nothing has meaning.

More or less, yeah. I'm not a proponent of it, I just think there's enough inertia in that direction of short-term vastly-superintelligent AI. Maybe we can figure out how to become subordinate components of a machine like that, but I think that's the best we can reasonably hope for.

I mean only people with billions of dollars can make competitors to anthropic because of necessary compute costs

that's the barrier to entry

Only people with billions of dollars can train foundation models, yes.

But a competitor to Anthropic at the product level? With open source models, very little barrier.

Careful with using these small models. The other day, I asked it "Can dogs eat avocado" and answer was emphatic Yes.

This is not meant as a criticism, but people should be aware of their limitations.

well, technically they can ...

...once

Reference?

Was it written by one person?

The majority of it, yes.

I thought about this first when chatgpt 3.5 came on the scene. Yes, you _can_ at some time in the future, replace programs with AI which would be slow to an extent - if AI can write and manage the code, it _could_ be even faster.

But there is a kicker here. It is upto LLM to discover the right abstractions for “thinking” while serving the requests directly or in the code .

Coming up with the right abstraction is not a small thing. Just see what git is over cvs - without git no one would have even imagined micro services. The right abstraction cuts through the problem, not just now, but in the future too. And that can only happen if the LLM/AI managing the app is really smart and deal with real world for a long time and make the right connection - these insights don’t even come to really smart people that easily!

Or maybe just don't use LLM.

LLM is just a tool in the A.I. world. There are lots of other A.I. tools, such as Neural Network, Fuzzy Logic, Genetic Programming, and so on.

Didn’t the Google Translate AI invent its own intermediate language for translating between languages?

Highly recommend watching Dwarakesh interviewing Reich: https://youtu.be/Uj6skZIxPuI

You will learn more about human prehistory in that 2 hours than anything else.