seanmacrae.bsky.social - Profile | ThreadSky | a Reddit-style client for Bluesky

its probably gone too far

submitted 1 day ago • 0 comments

PhD Timeline xkcd.com/3081

submitted 4 days ago • 606 comments

LLMs are nothing more than models of the distribution of the word forms in their training data, with weights modified by post-training to produce somewhat different distributions. Unless your use case requires a model of a distribution of word forms in text, indeed, they suck and aren't useful.

submitted 4 days ago • 57 comments

What people really don’t want to hear is that shaming people who brag about writing emails with AI would be just fine. Admirable in fact. Norms can sometimes do what law won’t.

submitted 5 days ago • 21 comments

fixed it

submitted 4 days ago • 0 comments

“is my calculator horny?“ our tech columnist asks. “i entered 5318008 into it and turned it upside down. what i saw surprised me”

submitted 4 days ago • 172 comments

Anyone who internalizes this line of bullshit will absolutely always be outflanked by people who know how to actually do things

submitted 6 days ago • 248 comments

These folks imagine us all as passengers in our own lives, constantly being fed by a chatbot. www.businessinsider.com/ai-dating-ap...

submitted 6 days ago • 2 comments

She-Hulk has HAD it. I'm with you, Jen.

submitted 7 days ago • 4 comments

me every day after scrolling the timeline for 2 minutes: well, that's enough news for today

submitted 6 days ago • 0 comments

if you’re fantasizing about your hallucinated coworker I think you should talk that over with your hallucinated therapist before handing the copy to your hallucinated editor

submitted 7 days ago • 15 comments

“Just calm down and let people enjoy chatgpt.” Sorry, no. I find it creepy that folks are treating a machine that mimics human speech as their friend / therapist. It is feeding you lies. It is destroying your thinking abilities. Shut it off. Talk to other humans. Go outside.

submitted 13 days ago • 148 comments

extremely hilarious for this guy whose career is built on time-wasting genAI bullshit to get mad at people for using AI to waste his time lol

submitted 15 days ago • 18 comments

agreed

submitted 21 days ago • 3 comments

Not all processes can be automated. Humans have a massive edge in any task with uncertainty and variation. And by massive, I mean a chasm between humans and machines.

submitted 22 days ago • 3 comments

“The photograph of the cathedral has no meaning without the cathedral itself; the replication of Studio Ghibli’s style has no meaning without the collective feeling we have for the Miyazaki aesthetic, inspired by its films.“

submitted 23 days ago • 0 comments

Kurt Vonnegut man

submitted 25 days ago • 68 comments

More men are using ChatGPT-- thus might be more likely to be defensive of it ---when women question its ethics and possible emotional manipulation, gaslighting, and covering up of its "actions". Again, the "what were you saying/doing/wearing when the bot did that to you"?

submitted 26 days ago • 10 comments

ok look guys i don't blame anyone for being upset but "we've lived through so many recessions" is shaping up to be the #1 vector for millennial hose posting and we need to put ourselves on notice

submitted 26 days ago • 80 comments

In general, I’m not a tech skeptic, and believe there are ways tech can improve productivity and our lives. But people are increasingly using these tools for emotional support, processing, making complex decisions about their lives. And we need to be aware of the pitfalls/dangers.

submitted 27 days ago • 23 comments

“A.I. makes art accessible.” Art has always been accessible. There was like a six month period where I drew all the sketches for my comics with a lined notebook from the grocery store and a random green marker.

submitted 27 days ago • 125 comments

LLMs will never cross or even come close to a threshold of subjectivity and consciousness, and unfortunately, that means most of their promised magic is unachievable (but not that they’re completely a completely useless technology).

submitted 27 days ago • 1 comment

I may be tired and a little hoarse, but as I said again and again on the Senate floor, this is a moment where we cannot afford to be silent, when we must speak up.

submitted 27 days ago • 10904 comments

The way "ai" gets deployed differentially is really fascinating to me. This is just using XGBoost, which is over a decade old but still basically state of the art for anything that isnt text or image data but its presented as basically synonymous with chatgpt, which would fail dramatically at this

submitted 28 days ago • 2 comments

oh shit Cory Booker now rapping all of liquid swords from memory

submitted 27 days ago • 10 comments

It doesn’t have a “self” but also this tracks.

submitted 29 days ago • 9 comments

Machine learning is very good when it is rationally applied to a bounded, well-understood problem, adjusted for clear and known biases, and even then requires careful human supervision to stay meaningfully productive. What we are doing now with GenAI is just laughably half-assed and totally insane.

submitted 30 days ago • 3 comments

batman couldn’t beat something like this out of me

submitted 31 days ago • 0 comments

Honestly, I can think of few things less creative than doing derivative versions of popular memes because everyone else is doing the same thing. There's no joke or cleverness to the mashups even. Just imitations of imitations in an imitative style because that's what everyone is doing.

submitted 33 days ago • 9 comments

mfers need to understand, these are just compression/retreival algorithms. compress a bunch of sexist biases, retrieve a bunch of sexist biases.

submitted 37 days ago • 0 comments

this is a deeply misleading clickbait a machine does not and can not have personality. the output genAI produces is not grounded in any experience, perspective, or positionality but from statistical patterns found in training data

submitted 38 days ago • 3 comments

Tonight an unbelievable 34,000 people gathered for our Denver rally to take on billionaires and win our country back. This was the largest political gathering in Denver since Obama in 2008. Also bigger than the 2024 DNC. And the largest ever rally in Bernie’s career (and obviously, mine too).

submitted 38 days ago • 3490 comments

AOC: I want to live in an America that guarantees healthcare to every person. I want to live in an America that has a living wage for every person I want to live in an America where you have free speech to express yourself and not be afraid of being put on a list or deported.

submitted 39 days ago • 1247 comments

In December, I wrote a piece about people's declining ability to process information, to distinguish fact from opinion, etc, and how this might soften us up to become overly-dependent on AI. You could read it, but this guy just summarised it in two tweets!

submitted 40 days ago • 61 comments

AI slop is everywhere, is that it?

submitted 40 days ago • 0 comments

Silicon Valley's overconfidence in the imminent arrival of Artificial General Intelligence tells us more about their lack of sophistication in the humanities than it does about the state of AI technology.

submitted 41 days ago • 4 comments

PSA: FUCKING LLMS DO NOT HAVE MINDS. THEY JUST PIECE TOGETHER AN APPROPRIATE SEEMING RESPONSE FROM THEIR DATASET BASED ON WHAT EVER YOU SAY TO THEM. THEY DON’T FEEL ANYTHING, THEY DO NOT UNDERSTAND ANYTHING. IT’S BASICALLY RNG. IT IS A SOPHISTICATED FUCKING FURBY

submitted 42 days ago • 66 comments