Interesting.
I hope for a new paradigm in web searching. I wouldn’t even mind if a search took 5-10 minutes, if it meant a handful of quality results. I easily waste that much time or more sifting through garbage ai and ad-driven results as it stands.
All I want is a search engine that 1. doesn’t make moral judgements on the results relevant to a search, 2. filters out ai and ad farm results by default, and 3. can be toggled to effectively search web 1.0-style forums.
Microsoft OS workload on an AI-optimized chip:
(5%) consumer benefit - users can get access to Clippy+ with a Microsoft premium account subscription, that if users aren’t subscribed, they’re reminded every time they go into the settings application
(15%) anti-piracy & copyright protection
(70%) harvesting and categorizing all user activities, for indiscriminate internal use, sale to other companies, and delivery to governments
(10%) Uninstallable OEM bloatware that does the same, but with easily exploited security flaws that are never effectively patched
Not sure if you already know, but - sophisticated large language models can be run locally on basic consumer-grade hardware right now. LM Studio, which I’ve been playing with a bit, is a host for insanely powerful, free, and unrestricted LLMs.
I think this type of anthropocentrism extends to chess too actually. I’m not an expert on the subject, but I’ve heard that chess AIs are finding success doing unintuitive things like pushing a and h file pawns in openings. If, 10 years ago, some chess grandmaster was doing the same thing and finding success, I imagine they would have been seen as creative, maybe even groundbreaking.
I think the average person under-rates the sophistication of AI. Maybe as a response to the AI hype. Maybe it’s because we’re scared of AI, and it’s comforting to believe that it’s operations are trivial. I see irrationality and anger cropping up in discussions of AI that I think stem from a fundamental fear of its transformative power.
NOC’s vocalizations were recorded and studied by a team of biologists from the National Marine Mammal Foundation (NMMF) led by Sam Ridgway. In 1984, Ridgway and others at the NMMF began to hear peculiar sounds coming from the whale and dolphin enclosure. They were reminiscent of two people talking in the distance, the words just beyond the limit of comprehension.[5] Later, a diver working in the enclosure came to the surface after he heard someone cry “out, out, out!”[1] After he asked his colleagues “Who told me to get out?”, they realized it had been NOC.
$330m is not nothing. But, with a funding split between a telecom CEO, and a shipping & logistics CEO - person has to wonder what sort of direction & tuning the team might be encouraged to explore. How will they stack up against existing & proven open source non-profits with impressive releases like EleutherAI?
These open source projects are neat, in that they give the average person the opportunity to peek under the hood of an LLM that they’d never be able to run on consumer level hardware. There are some interesting things to find, especially in the dataset snapshots that Eleuther made available.
In general, kind of cool to see France being on the cutting edge of these things. And I think it’s worth saluting any project that moves to decentralize power from states and megacorps, who seal wonderful, powerful things in black boxes.
One variable that I think doesn’t get looked at seriously is class size and school funding. Ask any North American teacher, and you’ll get a grim assessment on the trajectory of schooling since the 90’s. When teachers have more students than they can handle, it’s no surprise that things get out of hand.
I’d argue that part of the solution is more teachers per student. This enables better relationships between faculty and students, and better opportunities for mentorship. Build more schools, hire more teachers, pay them well, make school a place where teachers want to be, and where kids can thrive.
But reforming the existing system is a hot potato that neither the left nor right wants to hold, so, here we are. The system itself is degraded to the point that it doesn’t have the resources to self-correct. We need vision, wisdom, funding, and leadership, to steer things in a new direction. I think that would go a long way in preventing a misguided kid from fermenting the idea that murdering people, or their own classmates, is an answer to their problems.
I don’t mean to paint school shootings as simply a rebellion against a malfunctioning system, but, we really need to look at the system and make sure it’s serving the students that have no choice but to be there.
Thanks for the links, I’ll check them out. Persistence of data seems like an interesting issue.
Nobody would defend copyright if it wasn’t already in place
I don’t know about that. Say you take a few years to write a handful of poems, and it turns out people in your neighborhood really like them. You compile the poems into a book, and sell it for $5, and it sells well. Seeing this, your neighbor buys one, copies it, and starts selling it one neighborhood over for $2, and representing themself as the author. I would think most people in that situation would want to say, ‘hey, that’s not fair’. I don’t think that’s sick or rooted in greed, copyright can be a check on greed.
The problem really is that storage for video media is insane compared to storage of document or even photo data.
Yep, and add to that, 500 hours of video is uploaded to youtube every minute & they serve over 2.5 billion monthly users. The scale really is unfathomable.
If people here haven’t read into it, it’s incredibly interesting to look into the way the Internet Archive works. In particular you have to begin to concern yourselves with how long it takes for HDs, SSDs, and other media to degrade in time.
Where can I read more about this? It sounds interesting.
Absolutely. There’s nothing special about YouTube’s frontend - it can be replicated by someone with no coding experience, in an afternoon, for free, via a Softaculous module. On the inside, it’s the Library of Alexandria. And unfortunately, it’s owned by a company that understands that reality only as a means to a nefarious end, which is to develop a detailed psychological profile on its users that can be sold to advertisers.
My hope is that the cost of server storage and delivery will become inexpensive enough that YouTube can be forked and maintained by a nonprofit like the Wikimedia Foundation, who sees user generated content as a means to the enrichment of human experience. I’m not optimistic though, the history of the Library of Alexandria is instructive.
“Those people are snowflakes”
Following the theory that the leadership at twitter actually hate the users and are decimating the platform on purpose for the lols, maybe the outcome you suggest is the plan.
Part of me believes this theory, because it’s hard to imagine how someone even with the explicit stated purpose of destroying twitter could have topped the recent developments. It’s almost as if what they’re trying to do is embarrass and degrade the users.
I don’t know anything about anything, but part of me suspects that lots of good funding is still out there, it’s just being used more quietly and more scrupulously, & not being thrown at the first microdosing tech wanker with a great elevator pitch on how they’re going to make “the Tesla of dental floss”.
They exist. Inform yourself on the Internet Research Agency, one of Russia’s state sponsored troll farms. A handful of their activities are well documented in factual records. ‘Dems’ weren’t crying about it, every rational person who doesn’t want foreign interference and disinformation flooding our spaces is concerned about it. This should not be a partisan issue whatsoever.
Black lights
Nice.
Mmm. I grew up in a different time too. Makes me ponder how the software circumstances of that time built in us a very different idea of what an iteration actually is, when it comes to writing. The fact that we couldn’t go back and atomically dissect the history of a piece. That a draft, and an edit, were something heavier. Maybe we’d have to think a bit more slowly and carefully before irreversibly casting a previous version into the ether.
Don’t get me wrong, I’m not making a “gen z bad” post. Just reflecting on how things are different these days, and maybe it leads to a different kind of work.