"You don't get how it works it's just like a little baby robot looking at pictures or reading books!!!"
Normal people:
08.03.2026 00:33
👍 7
🔁 3
💬 0
📌 0
"You don't get how it works it's just like a little baby robot looking at pictures or reading books!!!"
Normal people:
Me sowing (leeching fuckloads of money by entwining myself with the US military): Haha fuck yeah!!! Yes!!
Me reaping (being a valid military target): Well this fucking sucks. What the fuck.
As someone required to interact with them I do not even grant them that
Embracing/rejecting LLMs continues to be an extremely strong scruples indicator
Orgs pushing this stuff also have managers instructing staff that they should "try to save time" and not ask their peers questions but the chatbot.
Getting young persons, especially developers, to actually ask questions is already difficult and these morons just make it worse.
The luddites were executed for insisting that laws be followed.
You have no idea what you're talking about.
llms aren't sentient, they aren't capable of sentience. but you should be very troubled at how ai evangelists a) sincerely believe they are and b) should continue to be treated as property
Definitely not trying to play into that idea. The near-constant invoking of the past as comparison for this stuff is itself tell-tale for how stagnant things are.
Constantly declaring "It's the glory days again!" (but the glory days in question were a nightmare)
The fact any of us can enjoy the fruits of the industrial revolution is because of the people who "fought against it".
If you truly believe we're in a parallel to the industrial revolution you aren't going to live to see its benefits without fighting.
You will be a victim of its horrors and, maybe, once people finally fight back, as happened last time, will people start to see their standards improve.
"With well over a century of hindsight and in a country that mightily benefited why yes the horrific treatment of the masses did have to happen the way it did"
If they're even fully aware of the horrors of the period it's taken for granted that the only manner in which the industrial revolution could have happened is the way it happened.
People really like to use their inadequate and myopic understanding of industrialization as cover for their deeply held beliefs that any sort of societal progress necessitates exploitation and extreme cruelty
Uhhh??? You want us to not put our admin points into the next tech? It's a +5 hammers in each factory wtf???
Things idiots that think real life has a videogame tech tree say
This shows how Open Source will die. With anyone just being able to pipe existing code through an LLM and claiming that to be "clean room" (which is hogwash) no licensing can protect your work from being taken and monetized by anyone. The commons are actively being shredded in front of our eyes.
a collage of these three tweets https://bsky.app/profile/why.bsky.team/post/3lzrkcsk3xc24 https://bsky.app/profile/why.bsky.team/post/3lzsgnvof4c2n https://bsky.app/profile/why.bsky.team/post/3lzubxv6q722o
i for one am shocked that the "i go in the pool to cool down, why can't my phone?" guy is bad at programming and offloading his job to the anthropic api
Lol you can find the math behind transformers and ANNs in general neat and fun and still acknowledge that this is toy bullshit where the most profound achievement is in the expediting and obscuration of plagiarism.
Also people that don't know anything use it and are blown away by the fact it can regurgitate well-rendered css it stole from somewhere.
They're "good at coding" because the bar for the people lauding them is "my code compiles and runs at least once" and model devs realized they could brute force that with a wrapper that runs the model over and over until that happens.
The code doing what it should is someone else's problem.
That has to be a bot or someone doing a bit. TREWS is a predictive algo from years ago that has fuckall to do with LLM """guardrails"""
I am dumber for having read any of that.
the result of UBI is not a better society. It's a bunch of people barely surviving while corporations and the billionaire class consolidate their economic and political power, and workers lose any ability we might still have to resist their efforts to destroy us, our environment, etc.
Tack on that those with little dev experience are the most likely to be wowed and oh look everyone's saying it's great for coding
It's gained a lot of traction because it's also a compelling narrative outside of tech - even for those that otherwise detest the genAI industry.
"The programmers automated themselves out of a job!" is cathartic for obvious reasons.
These are also the docs that wind up in hospital administration and talk excitedly about how AI will save them money- I mean clinicians' time
Ultimately, yes. Most of the problems that can be genuinely handled have alternatives that are not only actually reliable but faster. The remaining portion is just plagiarism but this industry has a culture of just break rules and worry about it later because 9/10 times no one comes after you.
"people who weren't saying it before have said the models from december are so much better"
Lol no they aren't. This isn't true! The latest marketing push just encouraged dorks that tried walking a ~balanced~ stance with crap like "I only use LLMs for boiler plate and UTs" into dropping the facade.
Okay but if it weren't for the capitalism might there be a path forward for leaded gasoline and cigarettes?
Especially if the criteria is "I possess the copyright to material trained on".
While I doubt even Disney has enough data meeting that criteria on their own to produce a competent video or image model, *were* it to work, the above would just shift the issue of exploitation over to Disney.
Like on one hand there's probably a conversation for works like ours having legal protections as artwork if you're strict about it but we can't have that conversation while the trillion dollar bubble exists to steer that conversation in a very wrong direction.