hobofan

Whether it's on purpose or incidental doesn't matter.

"using a free browser" = "take any kind of privacy preserving actions"

If you use an odd/fringe browser you are more likely to be caught by CF Turnstile and be locked out from websites.

It's the same distinction as "open source" vs "source available" (when speaking about source code). It's not that widely in use because few people in the LLM space are diligent enough with IP to make the distinction, but the concept is easily transferable.

Yes, sodium polyacrylate is used in both expansion microscopy and baby diapers.

However you also need a few other ingredients to ensure isotropic expansion, or otherwise the expanded sample isn't scaled uniformly in all axis and you get a skewed view of the sample.

I find it fascinating that some of the main areas people are interested in are using benchmarks with quite few questions: ~200 for coding and ~150 for science.

As long as that's the case, it's no surprise that there is often a perceived difference of quality between benchmarks and real-world use.

I know that running the full composite benchmark set is probably already quite costly, and also that the benchmarks have to be rotated regularly, but it feels like a significant untackled problem.

Is your question "why don't decentralized humans appropriately self-regulate?"?

> 3. Post IPO, will Open AI continue allocating similar levels of capital to further these causes?

I think the scarier alternative may be OpenAI forever gating their super-frontier models, capturing a significant portion of the sciences and leveraging it for equity in businesses that build on scientific advances. e.g. "Hey CATL, how about you guve us 5% equity in exchange for a battery chemistry breakthrough our AI has found?".

It's roughly equivalent to the playbook Palantir had for the public sector when they marketed themselves as a company that had cracked data science. At least that's what they've been pitching back then. I can't find a lot on public disclosed deals based on that.

For coding, no.

I have some business automation process for some customers that I maintain, where Terra with low reasoning ended up with the best price/performance & latency. With an equivalent Luna with high reasoning, price was usually better, but quality had a higher variance (very much depends on where the thinking takes the model) and latency for high reasoning was worse.

> well, Claude/Chatgpt is still marginally better for me now

Enterprise doesn't care about marginally better. They care about cost.

Most organizations that run our self-hostable chat platform with cloud models are for most stretches running 6 months-1 year old models, and are strongly eyeing (or have already switched) to more commodity models for most of their workforce. Frontier models are mainly relevant for engineering tasks.

I'm not asking him to open source anything. I've given up any illusions that that would be happening ~10 years ago.

I'm just cautioning other people to put any stakes into a language that's from its organizational bones doomed to fail for any outside party.

If he raises interest in the language by saying it will be open source (which he did), I think it's fair for people to dismiss the language until he does actually open source it.

Putting stakes into a language that will essentially be DOA, like the author does seems like a crazy decision.

This thing has the "the natural language nailed" to the same degree that a BERT has. That's a far cry from the intelligence level required to output text of any sensible length, and even farther from the structural adherence required for tool calls.

> uv sync

You can also just do "uv run", and it will automatically sync as necessary!

I think that's one of the things that uv does great and often goes unnoticed: It's hard to use wrong for a layman. Forgetting to run the equivalent sync command in most other Python package mangers could have resulted in a myriad of different bugs caused by out of sync packages, leading to hours of wasted debugging.

But the thing that makes this special, is explicitly cutting things out that make LLM's capable of that. As gpm's sibling comment points out, if you would add that back in, you would just end up in the converging to the same place (with the same downsides).

The only way around that would be to somehow get all world knowledge and tool calls into a audio space, which seems like quite the stretch.