And it's not even working. Not one of the AI companies is profitable. So they're putting the hope for profits some time in the future over sanity and safety.
I have a work-supplied laptop with Windows on it. I use it maybe once or twice a month, just for the things requiring a VPN. The rest of the time it sits there gathering dust while I get real work done on my Linux laptop.
The specs on the work laptop say it should be a performance beast, but my Linux machine (with half the RAM) runs circles around it.
I'm watching that happen in my industry (software development). There's this massive pressure campaign by damn near everyone's employers in software dev to use LLM tools.
It's causing developers to churn out terrible, fragile, unmaintainable code at a breakneck pace, while they're actively forgetting how to code for themselves.
A while back, I was thinking about upgrading my living room entertainment PC. It's got a decent video card in it, but some of the other hardware is getting long in the tooth.
Now, my plan is to focus on software tweaks to squeeze the absolute best performance I can out of it, and keep the hardware as-is until it starts physically breaking down. And when that happens, I'll find refurbished hardware to upgrade it with, rather than spending the exorbitant fees to buy anything new.
What mystifies me about all this is that it's obvious what the end goal is: No more PCs, and everyone just rents dumb terminals connected to AI data centers that run everything and have all the compute power. The problem is that literally no one but AI companies want that. Not consumers, and not other companies that sell software and services to consumers.
When cars replaced carriages, it was because people actually wanted them. Cars had real-world benefits over horses. But this shit? No one wants it. Gamers want game performance you simply can't get with streamed games. People who work with computers for a living don't want their ability to do anything to vanish if their ISP has an outage.
We can't vote with anything but our wallets here, so this is a thing we can do to at least reduce their income. And considering they're losing close to $100 million every day already, I can't help but suspect nickel and diming them a bit still hurts.
I wonder how OpenAI's investors feel about that expense, considering they're losing somewhere around $80 million every single day.
At the beginning of the year, they projected a $14 billion loss for 2026. They'll almost certainly exceed that by quite a lot, and may have lost $12 billion in the previous fiscal quarter.
They aren't bleeding money. They're a veritable cash volcano, blasting it into near-Earth orbit.
Neil Gaiman. I fucking loved Sandman and damn near everything else he wrote. Finding out he's a total scumbag has basically ruined a lot of very dear memories.
Yes, it's base64. And what's behind it could be anything that can be attached to an email.
In this case, it's a PDF. If the base64 text can be extracted accurately, then the PDF that was attached to the email can be recreated.
The challenge is basically twofold:
There's a lot of text, and it needs to be extracted perfectly. Even one character being wrong corrupts it and makes it impossible to decode.
As the article points out, there are lots of visual problems with the encoded text, including the shitty font it's displayed with, which makes automating the extraction damn near impossible. OCR is very good these days, but this is kind of a perfect example of text that it has trouble with.
As for my approach, I'm basically just slowly and painstakingly running several OCR tools on small bits at a time, merging the resulting outputs, and doing my best to correct mistakes manually.
I'm not having trouble with it as such, it's just a slow and painstaking process. The source is crappy enough that an enormous number of characters need to be checked manually, and it's ridiculously time-consuming.
Some of the emails in the file dump had attachments.
The way attachments work in emails is that they're converted to encoded text.
That encoded text was included - badly - in the file dump.
So it's theoretically possible to convert them back to the original files, but it will take work to get the text back. Every character has to be exactly correct.
Source: I'm a software developer and I'm currently trying to recover one of these attachments.
Dude. I mean this in all sincerity: Eat my entire ass. Conflating voting with support for genocide might not be the stupidest thing I've ever encountered, but it's gotta be in the top twenty.
Welcome! Isn't it a breath of fresh air to use an OS that isn't trying to turn your computer into an advertising and upselling platform? It has its issues, but it's a huge relief to escape the constant inundation from Microsoft.
They can tell the difference. But you'd be amazed at the mental gymnastics someone with extremely strong motivations can engage in.
You gotta remember that these are mostly people who went back to the Trump trough three times in a row, even as he's been steadily destabilizing and destroying their lives. If they admit they fucked up now, they have to admit they've been fucking up for over a decade, non-stop. Many of them would literally rather die than do that.
In order to get there, they need to recognize the fact that if they make themselves a disposable rubber-stamp for everything a dictator does, inevitably the dictator realizes he doesn't even need them anymore and disposes of them.
I hope they are at the point of realization, but I have doubts.
God, that's so frustrating. I want to shake them and shout, "No, your code is 100% ass now, but you don't know it because it passes tests that were written by the same LLM that wrote your code! And you have barely laid eyes on it, so you're forgetting what good code even looks like!"
And it's not even working. Not one of the AI companies is profitable. So they're putting the hope for profits some time in the future over sanity and safety.