cheeze

This is the thing I mention often.

"Am I arguing against the shuttle loom!?"

Then I realize that the shuttle loom led to the rise of unions because of unfair treatment in factories and realize that we have a _long_ way to go.

Single AZ S3 has other benefits. The point isn't the price, it's that it's _highly performant_ since you can keep all of your reads in the same AZ

Isn't openrouter the exact opposite of caring about security and privacy?

I guess you can choose your provider still? But isn't the point that the lowest bidder is doing inference?

> suing a torrent aggregator

Which has also been done :|

Bedrock is an option. Multiple platforms with solid data sovereignty.

If you don't want them slurping your data, you're gonna have to pay more.

Same as it ever was.

The classic slashdot comment!

Simple solution here, use a high quality OS.

Windows has never felt more archaic than now. I can't imagine trying to do work on a Windows machine in the age of LLMs.

If there is anything that LLMs are rock solid at these days, it's unix terminal usage.

It doesn't fit the narrative that they need to adhere to. I'm a skeptic through and through, but yeah... This article reads like propaganda.

I got here and... I agreed with everything up to this point.

The author makes a good point. If you don't know what you're doing, AI accelerates that. No question.

But they put a whopping... ten bucks into using DeepSeek and weren't impressed with the initial results.

I know they try to cover this with "you just aren't prompting correctly!" but if, in 2026, you aren't able to have an LLM generate decent quality code... IDK what to tell you. Good luck I guess?

What part of this was lying about what she had said?

I think the entire thing is absurd but libel usually requires falsehoods to be published. Slander is similar. It isn't just "damaged reputation"

I use something like OpenClaw. For me the big benefit isn't so much the autonomous nature but the organizational structure. The one I use exposes an MCP and lets me control it via talking to an LLM _in the tool_ which is really powerful. EG "audit this", "in X thread we were discussing Y, can you pull that information", etc. And the organizational structure just... makes a ton of sense to me. Don't have to set up a ton of steering and whatnot, it just kinda works.

Being able to use the search to find and resume an old thread is fantastic. The thing I use lets me use whatever underlying CLI that I want, which IMO the biggest limitation on Claude Code or whatnot is that you... have to use Claude models. Sometimes I wanna use sol, sometimes a cheap chinese model, etc. And I want sol with 1m context.

I know there are other tools out there that can do this. I know some people love their tmux/cmux, but IMO the thing that matters the most is being comfy and knowing the tool well. Hell, James Gosling used NetBeans when I talked to him in 2023. Know the tool well and that matters the most, IMO.

For me, it helps with my ADHD brain. And I could spend all day just researching and trying new tools, they are a dime a dozen these days.

Pulling some stats, I have 25 "tabs", and across all of those, I have 1,298 unique threads. 9,539 unique messages.

It works for me, but I don't care about the autonomous part.

IMO it's not. It's benchmarking GPT 5.4 and Opus 4.6. It's also missing Claude Code... one of the most popular harnesses (the most?)

Fable is great minus it's annoying writing style. It's nowhere near as bad as Opus 5 but it's very "AI trope-y" even with heavy steering. I can spot the writing immediately.

On the flipside, how many individual lives have been changed by the books provided to families for free? Impossible to quantify but I'd argue extremely important.

Similarly to the cars, I won't use it on moral grounds.

> I have tried many ways to get it to stop writing in its stupid punchy linked-in marketing-team voice, and I can't.

One of them writes better by default. One of them doesn't. That's the thing.

With enough steering, I can get a cheap low-capability model to do things correctly in most cases as well. But why bother?

You can tell it to write high quality code, to test things, and to come up with a proper rollout plan of a feature. Or it could just do it by default.

I know where I'm putting my money in that case.

I don't know. I think it makes sense to have a policy like this.

I haven't written a line of code myself in 6 months, but the point is that _you shouldn't be able to tell_

What I tell people at work: Using AI is good. AI can help you do things that you otherwise wouldn't have. It shouldn't be a crutch for thinking.

In that sense, people shouldn't be able to tell that you're using AI unless the tell is that it's _higher quality_ than if you had implemented it by hand.

A good example here is a well-written document that has a _ton_ of deep research behind it. Or code that has extensive tests that would have taken too long to write by hand for the task at hand.

It's bad when it's just "do this thing and post a PR" but I write my code with AI, look at it as a reviewer, offer suggestions, put those into my steering if needed, and I refine it. The way I see it, I'm the first reviewer on everything now before passing it off to another human for review.

There have been points where having AI rewrite something was slower than me doing it, but I'm already in the harness. The cost of waiting is almost zero. What stops me from contributing to the JVM using AI written code that I understand and spent time manicuring? Can a human really tell that it's AI written?

What is an example of something that is unsafe to consume via API or CLI?

Isn't MCP _an API_?

Is the em dash in the title meant to be ironic?

To say that it's American money in specific is laughable. The internet hasn't worked well on a matchday for quite some time in Spain due to their draconian laws around blocking piracy.

The EU can control companies that operate within the EU, they can't stop a Chinese company operating in China from destroying things. They could prevent the sale in EU without proof that the companies don't do that, but this is a good start.

This also isn't targeting the waste that comes from temu. This is targeting luxury brands that create artificial scarcity and destroy items rather than sell them at a discount. Think Gucci, Louis Vuitton, etc.

Ooh, anything Shackleton is a read for me.

Endurance: Shackleton's Incredible Voyage is a top 5 all time book for me.

│ When he saw the three men he stepped back and a look of disbelief came over his face. For a long moment he stood shocked and silent before he spoke.

│

│ "Who the hell are you?" he said at last.

│

│ The man in the center stepped forward.

│

│ "My name is Shackleton," he replied in a quiet voice.

│

│ Again there was silence. Some said that Sørlle turned away and wept.

Counterpoint, 80% of the code I write is not hard and I don't care about it past it being close to a reference implementation

Why wouldn't OpenAI offer the same?

I'm with you. If Dario posts in his blog about regulation of AI, I absolutely assume that is anthropic's position.

Or LA

Deepseek has some models in Bedrock. There is definitely a huge market for a "good enough" model running within the country of the company

While I agree with you in spirit... were you expecting that you could... burn the film to a DVD or something?

Of course buying a movie on itunes means you can only watch it on capable devices. You can't play a youtube video on a VHS player either.

It's a bit performative to say "x is my last day" when there is a ton of evidence to the contrary.

I agree with you, it's more that... it is performative in itself.