You use so many big words for nothing. All you need are backups. When it dies you restore. Nobody will care.

Now using CPU limits in k8s with cgroups v1 does hurt performance. But doing that would hurt performance without k8s too.

The last few times I bought a phone for myself or the company my main question was how low the price per years of software support is. Then I pushed the proposals up for approval and they asked me... the same thing.

Neither can single humans.

If you introduce some reasonable constraints AI will come out ahead most of the time, especially for optimization cases where AI will run circles around your average programmer and is perfectly happy to inline some ASM for you.

You still have bespoke cordwainers/cobblers 100 years after that process has been well and truly automated. But they're rare and almost nobody cares.

The choice between between teaching some midwit the law and going home in one piece seems crystal clear to me.

In a couple of years of riding I think the horn would have very slightly helped maybe... once or twice. If the other guy would have heard it at all which is doubtful.

There will likely be some internal reasoning going "I wonder if the user meant spell check, I'm gonna go with that one".

And it'll also bias the reasoning and output to internet speak instead of what you'd usually want, such as code or scientific jargon, which used to decrease output quality. I'm not sure if it still does

Many people with skin in the game will be spending tokens on hardening OSS bits they use, maybe even part of their build pipelines, but if the code is closed you have to pay for that review yourself, making you rather uncompetitive.

You could say there's no change there, but the number of people who can run a Claude review and the number of people who can actually review a complicated codebase are several orders of magnitude apart.

Will some of them produce bad PRs? Probably. The battle will be to figure out how to filter them at scale.

The only difference is that before AI the number of low effort PRs was limited by the number of people who are both lazy and know enough programming, which is a small set because a person is very unlikely to be both.

Now it's limited to people who are lazy and can run ollama with a 5M model, which is a much larger set.

It's not an AI code problem by itself. AI can make good enough code.

It's a denial of service by the lazy against the reviewers, which is a very very different problem.

It's a policy update that enables maintainers to ignore low effort "contributions" that come from untrusted people in order to reduce reviewing workload.

An Eternal September problem, kind of.

The community noticed he rarely blinks and he ran with the gag and edited all of them out.

The drive to achieve that level of success often comes from weaponized poor self esteem.

Well adjusted individuals just chill out after a few million and work on whatever is fun/important for them.

Only rarely does this also happen to be something that can take you from 10M to 1B. (and if it can it would take a lot of work you can't be bothered to do unless it's some core value like helping the poor beat malaria)

You actually get better through placebo, as long as there's a pathway to it that is available to your body.

It's a really weird effect.

The fight isn't against triggering placebo, it's against letting it muddle study results.

Look at the globe some day from that angle and compare it to the Mediterranean.

Somewhat indirectly but still.

I didn't even know there are multiple types of dashes.

I did know about multiple types of quotes because they kept breaking code on blogs. Still didn't care, but at least I learned how to spot and fix them.

Really looking forward to having the wrong kind of dash in code, but at least with current tech that seems like it won't happen.

Compared to a Dockerfile it's just too hard to follow