“Just” read documentation, says someone assuming past documentation is accurate, comprehensible, and relevant.
I taught myself QBASIC from the help files. I still found Open Watcom’s documentation frankly terrible, bordering useless. There’s comments in the original Doom source code lamenting how shite the dead-tree books were.
Due to some disagreements—some recent; some tolerated for close to 2 decades—with how collaboration should work, we’ve decided that the best course of action was to fork the project
Okay, that was always allowed!
Programming is the weirdest place for kneejerk opposition to anything labeled AI, because we’ve been trying to automate our jobs for most of a century. Artists will juke from ‘the quality is bad!’ to ‘the quality doesn’t matter!’ the moment their field becomes legitimately vulnerable. Most programmers would love if the robot did the thing we wanted. That’s like 90% of what we’re looking for in the first place. If writing ‘is Linux in dark mode?’ counted as code, we’d gladly use that, instead of doing some arcane low-level bullshit. I say this as someone who has recently read through IBM’s CGA documentation to puzzle out low-level bullshit.
You have to check if it works. But if it works… what is anyone bitching about?
Windows ME broke DOS support just to pretend it wasn’t 9x in a new hat. At the time - this was kind of a big deal. 98 was objectively better for any typical suburban setup.
XP was also NT in a new hat, but they had it fake all the bugs that popular games expected. Plus the hat was nice. That Fisher-Price UI era was genuinely great, especially compared to modern ultra-flat nonsense. Windows 95 had instantly visible hierarchy in sixteen colors. Nowadays you can’t even tell what’s clickable without guesswork and memorization. Or at best you get a dozen indistinct monochrome icons.
Windows 7 was the only time Microsoft nailed everything. Each XP service pack broke and fixed a random assortment of features. Everything from 8 onward is broken on purpose, first for the stupid tablet interface (when WinCE’s 9x UI worked just fucking fine on 3" screens with Super Nintendo resolutions), then to openly betray all trust and control. I would still be using Windows 7 to-day if modern malware wasn’t so scary. It’s not even about vulnerability - I must have reinstalled XP once a month, thanks to the sketchiest codec packs ever published. But since I can’t back up my whole hard drive on five dollars worth of DVD-Rs, the existence of ransomware pushed me back to Linux Mint.
Oh my god, they Windows ME’d it.
Oh no, statistical modeling about published works allows weird new shit. We must ban this entire class of software because we all care so deeply about copyright.
Do such models exist? Yes. Are they the big-boy models anyone’s really using? Ehhh not really.
There are in-use models that are “here’s a thing do whatever good luck,” which is at least as open-source as any MIT project. (Permissive licenses being “here is the code, have a nice life.”) Very few models are properly reproducible, because even when their training data includes DVDs you probably own, it also includes a ton of random internet pages that maybe don’t exist anymore. The push for ever-larger models, trained on as much stuff as possible, makes the use of “open source” regrettable or even deceptive choice. But quite a few are unrestricted for whatever weird shit you want to get up to.
Apparently Hunyuan just released some big-ass video model, and it’s air-quotes “open source” with a bunch of finger-wag restrictions. One of them is ‘you may not train your thing on our thing.’
Yeah I’m sure the companies that shrug off copyright concerns for Disney movies give a shit about Tencent’s pre-laundered intellectual property.
Humans will always outsmart the chatbot. If the only thing keeping information private is the chatbot recognizing it’s being outsmarted, don’t include private information.
As for ‘how do I…?’ followed by a crime - if you can tease it out of the chatbot, then the information is readily available on the internet.
Dreamcast: the first AI console.
It really was ahead on everything!
Oh it’s definitely not just Google. Apple’s been this fucked since 2007. But since this is the Android community, it’s helpful to stay on-message.
Shatter this corporation.
‘LLMs specifically won’t work.’
‘No, see, LLMs won’t work.’
Okay.
Right, should say deep neural networks. Perceptrons hit a brick wall because there’s some problems they cannot handle. Multi-layer networks stalled because nobody went ‘what if we just pretend there’s a gradient?’ until twenty-goddamn-twelve.
Broad applications will emerge and succeed. LLMs kinda-sorta-almost work for nearly anything. What current grifters have proven is that billions of dollars won’t overcome fundamental problems in network design. “What’s the next word?” is simply the wrong question, for a combination chatbot / editor / search engine / code generator / puzzle solver / chess engine / air fryer. But it’s obviously possible for one program to do all those things. (Assuming you place your frozen shrimp directly atop the video card.) Developing that program will closely resemble efforts to uplift LLMs. We’re just never gonna get there from LLMs specifically.
People have run LLMs on a Raspberry Pi.
The bubble is straining to burst because there’s not much difference between the high end and the low end.
Neural networks will inevitably be a big deal for a wide variety of industries.
LLMs are the wrong approach to basically all of them.
There’s five decades of what-ifs, waiting to be defictionalized, now that we can actually do neural networks. Training them became practical, and ‘just train more’ was proven effective. Immense scale is useful but not necessary.
But all hype has been forced into spicy autocomplete and a denoiser, and only the denoiser is doing the witchcraft people want.
Why are you even paying for it now, instead of doing it locally?
Honestly, the comparison with the rest of those icons was the only part I respected. It’s a suite of cult minimalist bullshit. Which still doesn’t explain how nobody went, “that’s a sperm.”
Or at least, “why isn’t it blue or yellow?”
That logo is fucking terrible. Cult minimalism produced a red sperm and a disconnected semicircle. It reads as nothing.
The ‘classic look’ in dark mode still puts 90% of the interface in light mode.
No other complaints.
All software beneath my web browser is also free, and none of it would dare.
Fortunately a lot of early Windows shit runs in Wine, since the most stable Linux API is Win32. Anything older than that either works in 86box or was broken to begin with. Okay, that’s not fair - WineVDM is necessary to bridge the gap for the dozen Windows 3.1 programs that matter. I am never allowed to write those off when one of them is Castle Of The Winds.
What Intel learned with Itanium is that compatibility is god. They thought their big thing was good chip design and modern foundries. They were stupid. AMD understood that what kept Intel relevant was last year’s software running better this year. This was evident back in the 486 days, when AMD was kicking their ass in terms of cycles per operation, and it caused division-by-zero errors with network benchmarks taking less than one millisecond.
But software has won.
The open architecture of RISC-V is feasible mostly because architecture doesn’t fucking matter. People are running Steam on their goddamn phones. It’s not because ARM is amazing; it’s because machine code is irrelevant. Intermediate formats can be forced upon even proprietary native programs. Macs get one last gasp of custom bullshit, with Metal just barely predating Vulkan, and if they try anything unique after that then it’s a deliberate waste of everyone’s time. We are entering an era where all software for major platforms should Just Work.