esafak

Some have both, despite using an auto-close bot! OC is at 4.6K open and 24K closed issues. https://github.com/anomalyco/opencode/issues

Was the mentality different when Imperial China was rich, and will it change again as it becomes richer?

Deepseek 4.1 Flash and Mimo 2.6 Flash.

Flash is plenty good for planning and reviewing, for my needs. In fact, I use it for that because it's too slow for execution, despite the name.

edit: I subscribe to z.ai, I don't host.

LLMs don't struggle with Svelte. Why wouldn't you use it?

Just read the paper instead: https://automatictransmission.khoury.northeastern.edu/paper....

I use it headless for reviews in CI.

If you modify it for use in a commercial backend (not a hosted db service), does the AGPL-3.0 license mean you have to share the source of project it is used in or only of the forked repo? On a related note, have you changed your stance on accepting issues and pull requests like some orgs?

I feel bad for the Jev guys. I wonder if they anticipated this much competition?

Sandboxing is not a substitute for alignment. A big part of the utility of these models is in their interaction with the real world. Entirely so when they are embodied. By definition, this means puncturing the sandbox.

That's part of alignment.

I would expect a flash model, with its reduced size, to suffer on tail tasks. That is the trade-off you make.

Because 'graphical' TUIs are pale imitations of GUIs. I don't have a TUI fetish, despite having grown up with them.

You say that because the 'most' existing models have done is hack governments and companies. Can't you think of worse things a model could do; accidentally or by instruction?

The present leapfrogging is not a contraindication because companies are not necessarily releasing their best models; we know they have smarter internal models. Furthermore, humans are still involved in model creation. Human involvement is expected to decrease over time, and when model iteration is completely automated, progress will happen at the machine's pace, leading to runaway intelligence, barring any ceilings.

I use a variety of models for various subagents. I don't want to change my harness every time I change models, or be beholden to companies for something the open source community can handle better.

Maybe they just use it less. If you code all day you can go through a billion tokens.

How fast is it? Looks like they don't offer 5.3 Flash.

I have not tried it yet but this looks as risky as openclaw, which I also won't use. What if it does something I would not have approved and I only found out about it later? Knowing how often agents go off the rails when I'm coding, I would hesitate to let one do other tasks. I would prefer to white list tasks one at a time as I gained trust.

Productivity is increasing as models get smarter; we are ascending the singularity. I'm serious.

The late author.

Jev ought to offer a flex mode that uses their spare capacity for a discount.

So is Israel the homeland of the Jews? If so, what does that make Palestinians? Tourists?

What a burn!

Ella: I think we just looked for podcasts that didn't have many comments.

Ira Glass: I see. So you picked NPR because it didn't seem very popular.

Ella: Yeah...

Of course it is an option. There is no prize for being the first to cause a disaster.

If you use subagents your main agent won't need to compact as often, with the loss of information that entails.

I don't think so. We probe people before entrusting them with risky decisions. We ought to be able to do the same of AIs. Even better, in fact, since we know everything about models down to their weights. The only thing we shouldn't do is to let them evolve at their own pace and make decisions without any oversight. If that means sacrificing some productivity that's fine. Aren't we getting amazing productivity out of what we already have?