almostuseful

If Astra uses some sort of looped transformer architecture (and if Fable doesn't), then that could explain why Astra can do more computation per token. It is essentially simulating a much deeper model.

But yea, I was also quite surprised by the low token usage.

You have to start somewhere. But looped transformers https://sebastianraschka.com/llm-architecture-gallery/looped... are more similar to your idea.

Give it a few months and open-weight self-hosted models will be able to do similar reverse engineering work. It is an extraordinary future we live in.

But the first time someone came up with that idea it was an act of creativity.

If at some point a competent senior software engineer can be automated away, I think we are so close to a possible 'AI singularity' in as much as that concept makes sense, that nothing really matters anyway.

I don't know what will be automated first of the competent senior software engineer and say, a carpenter, but once the programmer has been automated away, the carpenter (and everything else) will follow shortly.

The reasoning is that there is such a functional overlap between being a standard software engineer and an AI engineer or researcher, that once you can automate one, you can automate the other. Once you have automated the AI engineers and researchers, you have recursive self-improving AI and all bets are off.

Essentially, software engineering is perhaps the only field where you shouldn't worry about automation, because once that has been automated, everything changes anyways.