scosman

I would have guessed that was just a bad title here but no, article states it as "opted in by default".

even the color match...

beautiful. But looks like an Aston Martin

yup.

yes exactly. Too many people ask AI to one-shot complex tasks, and wonder it behaves like a junior asked to rush something.

I have my own skill: 5 rounds of research/planning/test-planning. Interactive with me in loop for all important decisions. Starts with high level shape, then details. Planning can take 2-3 days of my time, then the implementation agent can take many hours (Opus 4.7). It splits the implementation across many phases/commits, each with its own code-review fix loop. Deep code review at the end can take another hour or two. It opens a PR, Gemini reviews, it reads out and resolves those issues.

Projects still take days or weeks, but 5x faster than doing it all myself.

Edit: the skill - https://github.com/scosman/vibe-crafting

The README has some renders. Some fun ones: 100/100 Acid test without a JS engine, convincing Google SERP even when blocked by bot-detection, makes my homepage look better than in real life.

If you steal my car, no who knows it's stolen would say it's "yours".

We're not talking abstract language concepts, this is a specific case. The data was taken without license/rights/approval. It's stolen. AA calling it "our data" is disingenuous. Legally it isn't theirs. While you could use "ours"/"theirs" loosely in English, they knew it wasn't true in a legal sense when publishing this.

Sure, but the difference here is the pirate is claiming it's "their data" and asking for donations.

We won’t be doing it in 2 years. By then my side will have won!

Doesn't work on Chrome or safari for me

"Join the livestream" button does nothing?

> I am optimistic Kimi K open models soon will outperform Opus models

Hard to outperform the model you distill...

I believe GDPR says you can access a copy of data they store about you, not that they can't sell it.

Did anyone else read this in Anthony Bordain voice?

Why so narrowly eval just with/without skill?

Same approach is useful for everything: model, params, prompt, sub-agents, skills, rag, etc?

Not sure I follow the logic. The list of profiles I visit feels like it’s my data, not the owners of target profile. By that logic can I GDPR chrome for the browsing history of anyone who has visited my site? IANAL but I thought GDPR is about getting a copy of your data, not others.

But calling it OpenAlacritty would be worse, which is what happened here.

2 commits, 8 hours old....

Mythos prompt: Hey Mythos, make me 20,000 H100s.

We're officially down to one 9 of uptime over last 90 days: https://status.claude.com

I think the thesis is it's not solely about returns. Get a 7% return investing a dollar domestically, plus add a taxable dollar to economy (and some recurring benefit to tax base) beats getting a 9% one-time return investing internationally.

this is what the unsloth post is for

not if I can help it!

https://github.com/scosman/pelicans_riding_bicycles

Making watches. You can start simple ("modding") which is essentially assembling a movement, case, hands, dial, etc. You can keep going deeper: try to fix or build a movement gears.

Agreed. They had a rough patch around the 4.7 to 5 upgrade. New architecture required hardware migration. The 5 to 5.1 upgrade was much smoother (same architecture new weights). As you say, little rough around edges, but still great value. Trick I learned is that it's max 2 parallel requests per user. You can put a billion tokens a month through it, but need to manage your parallelism.

Check out z.ai coder plan. The $27/mo plan is roughly the same usage as the 20x $200 Claude plan. I have both and Claude is a little better, but GLM 5.1 is much better value.

also: inference time scaling. Generating more tokens when getting to an answer helps produce better answers.

Not all extra tokens help, but optimizing for minimal length when the model was RL'd on task performance seems detrimental.

Is there a way to view IG pages without logging in? I would love to delete the app and setup privacy redirect.

> Where it breaks down is any task where you discover the requirements during implementation

Often you can find these during design. Design phases felt like a waste of time when you could just start coding and find the issues as you go. Now I find it faster to do a detailed design, then hand it over to agents to code. You can front-load the hard decisions and document the plan.

Sometimes the agent might discover a real technical constraint during dev, but that's rare if the plan is detailed. If so you can always update the plan, and just restart the agent from scratch.

This is my (cheesily named) process: https://github.com/scosman/vibe-crafting

I opened a startup's page the other day, and their streaming demo video was 550mb