joelthelion

I'm not an expert, but maybe updating an existing open-weights model to give it up to date knowledge would be more tractable?

Looking at power output, or worse, power capacity, is not a good way to compare solar to nuclear.

While solar is definitely useful, nuclear is available at all time, during the night, winter, and bad weather. Solar as it is right now (even with a lot of batteries) is not a direct replacement for nuclear energy.

As a vim user, I feel neovim was useful for one thing: it shaked regular vim development a bit. That said, I see no reason to switch. Good old vim is rock solid and featureful. I hope it doesn't die due to neovim's competition.

They seem to be better at solving concrete problems than designing new things (at least for now)

Could the Chinese mission just take the US samples without the US's permission?

I've been wondering about this for a while. Maybe it doesn't work? Or maybe frontier labs just prioritize benchmark scores in spite of all their safety talk.

> Providers probably serve quantized versions without disclosing it

It should be OpenRouter's responsibility to protect you against it, by regularly benchmarking providers and giving you the control to avoid bad providers. In fact, that's a big opportunity for them, since it justifies their place as a middleman between users and inference providers.

Excellent post, thanks for the article. This is very relevant for anyone using openrouter and similar services seriously.

I feel there is still a lot of progress to be made before we can really trust LLM providers.

Why are the lamp posts blinking in a weird way?

Is GitHub really better? If you're a small project, you could always migrate to something else if this guy disappoints.

That's exactly what I meant . LeelaKnightOdds is specifically trained for that, and it makes a world of difference.

One issue is that models are only one part of the equation. Search is also extremely important, and for that you pretty much have to depend on a third-party service.

Is there an equivalent of LeelaKnightOdds for Go? That might be harder to tackle.

This is very cool, and at the same time, this is how they plan to lock people in. The moat isn't really in the models, it's in their users and how much they are invested in their ecosystem.

Once you have started working seriously with these tools, it stops being so easy to switch. What are the open-source alternatives to this? How usable are they for non-technical users?

Another comment I have about this is that unfortunately, there is always an LLM in the loop for each prompt. I feel LLMs should automate themselves away, meaning that for repetitive tasks, user prompts should go directly to deterministic, previously built scripts. This would be both a better user experience (more predictable), and a lot cheaper to operate.

> calculating the volume of a prism-like polytope between it and its projection

There is the key insight that you don't need to explicitly compute this projection.