flecomet

The greater the clock speed, the greater the energy consumption, but I guess compute would scale linearly with frequency.

However LLMs are not really tiny compared to real brains, brains FLOP/s have been estimated to be around 1e15 to 1e21, while a single H100 SXM can already do 1e15 BF16 FLOP/s.

The brain also has about 1e15 synapses, each holding the equivalent of about 4.7 bits of information, which would be about 590 TB worth of parameters, equivalent to 294T BF16 parameters (would need 7.4k H100s to hold those, at 80GB per H100).

So in my opinion, it's more of a memory bandwidth problem than a computing power problem, as H100s have comparatively very little memory bandwidth per FLOP than the brain, which holds information directly inside the computing units (synapses).

And I also think that detail recall is a double-edged sword. Memorizing details can certainly be useful, but it can also compete with pattern generalization. If you can perfectly remember the training set (overfit) you don't need to learn to generalize.

> The model's parameters are larger than the weights of the neural network used for in the model.

What do you mean by this ? The model is the neural network itself and the parameters of the model are the neural network weights (+ non-trainable parameters).

I don't agree that Opus 5.5 is necessarily a better communicator than older models. In fact, I ran my own tests, and the difference wasn't very clear, and in terms of personal preference, the writing style of newer Anthropic models has really let me down. It doesn't reason like a person that imagines what the person they're explaining to might know or not know, or what they might think, as one usually does when trying to explain something to someone. They just fact-dump, like the problem solvers they are, since they are trained to solve user's problems correctly, efficiently and comprehensively, but they probably don't have a good internal representation of the user they're talking to, they are not really trained to imagine what we may think or feel, and thus, in my opinion, make pretty bad communicators (cf. the ocean of vibe coded READMEs on github that are horrendous to read).

For example, when asked to write a report on Docker and explain how it works, one of the first paragraphs was : "1. *Docker Engine*: The core component, which is a client-server application that uses a server component (the daemon) to manage containers and a client interface to interact with the daemon."

This is perfectly correct and makes sense for someone that is familiar with software engineering and already knows what containers are. Not so much for someone wanting a introduction to it. Do you imagine writing something like this at the very beginning of your explanation of Docker ? It even said: "I wrote it in the neutral register from your publication preferences, with no em dashes, and defined terms on first use.", when it didn't even define what a container is. I chose Docker at random, because I could judge if its explanation made sense, since I know how it works.

Although I agree with all your points, I think there is a potential caveat with grep: it is not very token-efficient as it can return lots of matches, filling up the agent's context.

A potential solution for these indexing systems is to maintain several indexes (one per branch for example). Idk how they handle code changes though.

But I think the bigger negative signal is that these tools have been around for roughly as long as LLMs have, yet they are still barely used. Maybe because they are not that useful ?

you are my hero

I believe this is most clearly visible in mathematics. Who is going to review the 700 proofs put out by OpenAI in a single day ? And even if mathematicians could, how will they handle the exponential growth of LLM-generated proofs ?

Not to say these proofs are slop, even if the LLM-generated work is of good quality, what happens when no one knows how it works anymore ? Even if you ask the LLM to explain, which it does quite badly, the time to understand the explanation is incompressible.

So in the end, productivity will probably reach a ceiling that we can estimate as the product of humans, their cognitive capacity and their time. And that ceiling may be lower than what AI companies valuation expect, regardless of the compute and they can pump out and the RSI level they can reach.

"Open windows and tabs, navigate, read page HTML, and expose the API remotely with Cloudflare — from any language that can speak HTTP." What does that mean "from any language that can speak HTTP" ? Did you mean from any program that can make HTTP requests ? It seems pretty interesting and useful, though I'm no expert, but the README is hard to parse. A short GIF showcasing usage and capabilities wouldn't hurt. Also perhaps explaining which problem exactly is being solved compared to other existing solutions.

"The first thing to understand about Akhetonics’ optical computer is that it’s not one thing yet." Great, that is such a good explanation. Thanks ChatGPT I guess.