Do you genuinely believe this is true for even just 10% of gamers? 10% of people? I’m not saying I disagree, I just don’t believe any relevant amount of people give a damn about the purity of art for most things.

Most art the average person interacts with is mainly utilitarian. Marketing and porn is probably like a good 80% by mass. As long as the art does it’s job, I just don’t believe most people give a fucking damn.

To try to come up with an explanation that doesn’t pretend these peoples brain fell out of their heads:

Some people are crafters. For them the process and even the tools around that process are important, maybe even more important than the outcome. It makes sense that they would be wary of AI, to put it mildly.

Other people are builders. For them, all they want is to solve an issue, build a solution, create an art. The way they get there is an implementation detail they don’t care about. They don’t care whether it’s real oil, or digital oil, as long as the oil painting in there head becomes a real thing somehow. It’s understandable why they might at least try to use AI for art.

I’m sorry, but I’m certain that the average AI summary beats/matches what the average person gets from a normal search. You’re acting like people before AI didn’t just believe whatever the first or second link told them, for most cases.

If you’ve ever even read the summary of a study, you’re in the top 10% of “people that research online”.

Same is true for meat. Meat is to nutrition what AI is to intelligence. An extremely wasteful, ultimately unnecessary version with better alternatives.

The main difference for all the complainers is of course that they already don’t use AI. Not using meat would, in contrast, require actual effort beyond moaning and bitching.

If your morals depend on whether it requires personal effort, you’re not moral, you’re a hypocrite.

We aren’t, though, not even close. And we’ve also been holding back over the years. Trump even paid a german company to abandon an already approved offshore windpark.

Even still, I beg people to stop falling for the same sensationalist trick that makes every conservative foam at the mouth about immigrants. PUT ABSOLUTE NUMBERS INTO RELATION TO SOMETHING!! What is the estimated 34 million in 2030 in percent? What is the impact of non-veganism in the same time? What percentages are we looking at?

Our outrage and effort should be in relation to the negative impact of each problem. Animal AG has a gigantic impact in every sensible metric, and it’s far far far bigger than AI. Be honest, do you think even a fraction of the loudest complainers about AIs resource use are vegan?

Can we really believe people whose morality stops wherever they’d personally need to actually do more than bitch and moan?

I’d love to see any kind of sources on this. From what I can tell animal agriculture emits roughly 20–30× more CO2e than all data centers combined.

Animal agriculture: The FAO’s most recent estimate puts livestock supply chains at roughly 6.2 Gt CO2e per year, about 12% of global greenhouse gas emissions Data centers (including AI): The IEA puts all data centers at about 0.5% of global CO2 emissions today (~180 Mt CO2), projected to reach 1–1.4% by 2030

We really have literacy crisis on our hands. Their estimates put the industry (if we pretend it’ll not crash and keep growing) at 34 million tonnes of CO2. What percentage of USAs yearly emissions is that? 0,6% of net emissions.

It’s not the animal AG that’s doing it, not the oil industry, and not the car industry. It’s the 0,6% that MAYBE might happen.

I say this not to ignore those things, but to sensitize your valid impulse to fix global climate change issues to the real culprits. Most importantly, I’d like people to adjust their levels of effort and outrage to the actual impacts instead.

If you cared as much as you pretend you do, you’d already be vegan.

Well, sure, but when everyone loses their fucking minds about the tiniest thing, that even requires us to buy all the lies from billionaires about how much the build out will be, while having never done anything about the biggest thing they could personally change, it starts sounding mightily hypocritical.

It seems like the personal responsibility to boycott polluting and unnecessary industries is limited to whatever doesn’t require any personal effort from the person speaking.

Napkin math:

• 40% of americans are very or extremely concerned about AIs environmental impact
• To save 34 million tonnes of CO2e, you need about 5.2 million vegans from now until 2030
• So only 5% of ONLY the complainers would need to go vegan to make up for this “tremendous” impact
• That’s 2% of all americans

That’s the thing we’re even attacking random strangers online for. The thing that we pretend is going to bring about the end times. But sure, veganism is the extreme position…

By the way, what other resources are we saving?

• Water — about 8.8 trillion gallons (~33 billion m³)
• Land — around 1–1.5 million acres of forest/land
• Animals — roughly 8 billion animals including fish and shellfish

It wasn’t actually “bad” research that triggered it. It was the absolute INSANE fact that a single goddamn sentence of Hanks sounded like AI. Of course, Hanks videos trained all these models, so that’s not actually surprising to anyone with a brain. Which most fervent AI haters lack.

He made the mistake of treating his audience like grown adults, and explained the situation honestly. He uses AI for some research, but the text in question was written by humans. Now the witch hunt was off to the races. Anything not entirely perfect he had done (like the crocheting thing) was ripe picking to pretend that no single human had ever made a mistake. Surely, it must be the evil AI. Surely, Hank must be lieing.

Ironically, Hank should’ve lied. He treated an online hate mob like grown adults, which is always a mistake. They never actually reward honesty or accountability or positive change. All they want is outrage and hate.

What traditional tool can run an automated code review on every PR? Maybe even fix (or propose fixes) for the obvious issues?

A very large chunk of the improvements in the last year have come not from categorically better models, but from the circumstances of the models massively improving. For example, reasoning is just automatic prompt engineering, and eats a fuckton of tokens. Harnesses give LLMs tools, making it easier to turn nondeterminism into determinism (does this code compile is a decision the compiler can answer definitely). Then there’s subagents, which is just automatic context engineering.

Basically, the price per token might not have changed, but in practice, the amount of tokens used to get “SOTA” performance has massively increased.

No, at least in the case of Germany, keeping them running was simply not an economically viable option. They were ANCIENT, so would’ve needed a fuckton of work, which would’ve taken forever AND cost a fortune. This is the opinion of the energy industry THEMSELVES, not from bleeding heart liberals.

There’s basically no universe where building nuclear is a better option RIGHT NOW than just building more renewables and batteries and a smarter net.

The actual price is hard to really know, but I think training should also factor in. The hype of LLMs is based on the fantastical idea of continunous improvement forever, so you need to keep training. Even ignoring the hype part, you still need to retrain simply to update the data inside the LLM.

I guess we’ll only know for sure after the crash/readjustment.

From my memory of their financials, they actually get enough donations for the browser. The issue is the for profit companies fetish for a million different side quests.

There’s following the law, and then there’s giving away data to government agency just because they asked nicely, and could MAYBE get a warrant in the future. It is the equivalent of letting police into your house without a warrant, because maybe they’ll get one.

What exactly are we talking about? Doing Windows related development on Windows is roughly as decent as doing Linux related development is on Linux (or Mac).

It’s just that because like 90% of servers are Linux, 90% of development benefits far more from being developed on a Linux-y system.

For example, the Windows filesystem is very different. Over and over I’ve had issues with permissions being different, with paths being inconsistent (this happens esp. with WSL) and with limits on path length.

You can develop on Windows, but having the test env closer to the real env takes care of so many little headaches.

Ignoring prices, Mac is definitely the second best option after Linux for Linux-y development flows. None of my issues were huge, but still enough to ask for a Linux laptop for a replacement.

1. Very little customization, compared to Linux. I’m talking horizontal tiling window managers like Niri
2. Docker does not run natively, so you pay a hefty performance penalty with the VM
3. File name case insensitivity caused a bunch of Git issues

Sure, if you just use Linux for dev, with a Windows hypervisor, you won’t notice the difference.

We devs also have a serious issue of performance blindness, because generally work and test on pretty beefy machines. Windows 11 is undeniably heavier on the system than Linux, and Mac hardware flies anyway. If your dev machine is beefy enough, you won’t really notice though.

It obviously depends on the environment, but if I am supposed to develop tools that, in theory, can fuck up everything, then I also need access to everything (on my machine). There’s no point in testing, if the elevated access rights on the server suddenly surface a fuckton of extra bugs.

Heck, I need admin just for the basics of installing developer tools and opening web ports.

They tried to lock our stuff down once. After a couple of days of absolutely zero work being done because all our tooling was missing, and the poor IT guy had to somehow learn how to install every tool we needed and taking forever, we just got sudo rights.

Cool, and then there’s NEVER any problems with different paths? With back and forward slashes? With the limit on path length? With missing permissions on the file system requiring weird workarounds?

Most importantly, your server is likely not Windows, yet you test on Windows, and that’s never ever been a serious source of issues?

And don’t say WSL. That’s like saying the fix to using Windows is to use Linux, but fiddlier. Not to mention you still get issues with the mounted file system.