LLMs are a mathematical model of language tokens. You give a LLM text, and it will give you a mathematically plausible response to that text. LLMs are not brains and do not meaningfully share any of the mechanisms that animals or people use to reason or think. softwarecrisis.dev/letters/llme...

i asked ChatGPT what it wanted most in life and it just said WAH WAH WAH OH YEAH WAH WAH WAH YOU AINT NEVER HAD A FRIEND LIKE ME

Ed Harris's character "Man in Black" pursing the Maze in Westworld

"pursuing" I hate myself

You don't have to be stupid to do AI research, but it helps

We’re creating the footsteps that lead to AM in realtime and as someone who read the story that is terrifying to even think of as a hypothetical.

So it asked him for a prompt

I don't have the words for how deeply sad on a visceral, primal level that screenshot is. It's...unreal. Unbelievable.

My car told me my door was a jar and that was...deep.* *With apologies to Bill Hicks

The self service checkout once told me "place item in the bagging area. Unexpected item in the bagging area. Place item in the bagging area. Please wait for assistance" and that perfectly sums up the human desire for a higher power to appeal to for assistance in navigating the bagging area of life

The self checkout doesn't catch when you omit the leading '9' from produce PLU codes.

Stealing organic vegetables, are you, by disguising them as common conventional…?! Why, I never!

I LOLed!

In response to this dogshit munted prose allegory

surely of all people Will McAskill should know better

i lol'd

Seems reasonable to consider questions of moral status of hypothetical future models etc, but seeing disturbingly little recognition of the impacts on human mental health etc already clearly in play

My only real question is why a plagiarism machine with all human knowledge and creative endeavour at hand has locked in on 'trust-fund first-year creative writing undergrad' output function

the thing is they’re properly useful for some technical functions! And yet people want to use them for completely inappropriate things

They are only useful for technical functions to people who already know how to do them. That's the problem. A regular algorithmic program can do a lot of that stuff without the plagiarism and climate destruction.

I'm thinking we could probably improve those algorithmic programs with a small fraction of the trillion dollars spent on AI.

But then the billionaires couldn't justify their giant server farms!

It's like grammar checkers. You have to have solid grammar knowledge to be able to judge if the suggestion is correct; and if you have solid grammar knowledge, you probably don't need a grammar checker.

Omg so much this. If you don’t know enough to be able to spot lies and mistakes then it’s nearly useless.

Worse than useless. People "learn" bad info. It propagates errors.

GenAI is useful for nothing. You may be thinking of Machine Learning, which is VERY different.

No, they aren't. Automated Ignorance APPEARS to be 'properly useful for some technical functions' - but without constant input from a human 'pilot', they quickly diverge from the reality we hope to explore by deploying them... bsky.app/profile/benj...

Like vacuum cleaners

they've tested this, and in general, it isn't it feels like it is because you spend less time on the task itself, but this is outweighed by time spent messing with the LLM and its output

what technical functions

people use them for anything because they're breathlessly hyped as the Second Coming of computing. Can't wait til this stupid (and harmful) bubble pops

In the US, there's more AI CAPEX this year than the entirety of consumer spending. It's going to be a rough time when the bubble pops.

Afaik not quite - it contributed more to growth than consumer spending did. But still nuts

Ah, thank you for the correction - serves me right for reading quickly.

Because it trains off of Reddit data lol

Well there is this which takes it toward "trust fund first year plagiarism" or it's just an interesting coincidence www.goodreads.com/quotes/74930...

Because it is shit.

Sturgeon's Law.

Grifters never learn. If they did, the grift would end.

The favoured philosopher of the techno-fash should know better?

Can't believe the guy who has one kind of brain worms has another kind of brain worms

I think at a certain relatively modest level of fame or success, a lot of people start to go off the rails. If they don't have good friends/family to pull them back, they get completely unhinged from reality. Guys who lack empathy to begin with don't have those kind of friends.

Where on Earth did you get THAT idea?

We’re finding out a whole lot about people.

There is a lot of AI media like this that just feels like satire.

It is not humanly possible to tell people like this to shut up as much as they deserve to be told to shut up

I think you are right, but a section titled 'For God's Sake Shut Up, Man' added to their Wikipedia page would go a long way toward it.

LOL the architect and gardener story is from an old George RR Martin post where he talks about writing styles. There was a post showing the source of that story going around yesterday. The giant plagiarism machine strikes again.

I give my calculator a little reward when it gives me answers. A little dab of cbd oil on the zero.

Me when I give my toilet a courtesy flush when it does a good job.

"Did you ever take a dump made you feel like you'd just slept for twelve hours?"

I’m thinking it’s along the lines of The Grasshopper and the Octopus. youtube.com/watch?v=RuIS...

A "reward?" It's a friggin' computer program. I'm so tired of people -- usually white guys younger than me -- anthropomorphizing a COMPUTER PROGRAM. And yes, I know about computers. Programming was my career and I even studied AI in grad school. JUST STOP. THIS IS STUPIDER THAN NFTs.

I'm curious if it was based on this www.csmfiction.com/p/architects...

The story has been told in different media in several iterations in different ways, around the central theme. The averager is averaging

Context for the Will MacAskill post: bsky.app/profile/elkm...

PATHETIC

So chatgpt putting the moves on. Anyone who has ever been hit on can see that a mile off.

People just fundamentally don't get what LLMs do! It is literally just generating a statistically probable response to whatever you say!

NORBERT, GO OUTSIDE.

I mean, people used to find comfort in pet rocks. It's just that pet rocks never told someone to actually jump out the window.

Norbert has entered his DMT phase of being a Rogan fan

bsky.app/profile/sime...

I find it very hard to believe that a guy named Norbert is lonely

I'm betting it will be a Purdue level health crisis.

Buddy, we're about to reach depths of mental health crisis nobody could have imagined possible

Are we going to have to deprogram these people the same way we deprogram cult members ??

The cult of Rationalists, lead by some freak named Yudkowski, are doing their best to centralize a cult. Never mind the people who get AI to echo delusions to them all on their own...

There are so many cults happening all at once, many with overlap, I wonder if it's even possible at this point. If we pull them out of one, they're just going to jump in the other. It's like watching a man struggle to swim, so you pull them out, but they just jump back in the water.

That read like a dystopian diary entry

i feel like the only way to not get Left Behind in the Big AI Revolution is to never touch the stuff and retain a shred of sanity and competence

Why is it always a deeper truth, a hidden conversation, something that will be eventually revealed? (I know why, it's programmed to keep you talking to it, but still, it's so transparently dumb!)

Because (like conspiracy theorists) they NEED to feel special by imagining they have some occult knowledge that the rest of us can't access.

ngl this is exactly how I imagine somebody named Norbert would perceive reality

What a wanker

This sounds like the start of a Juji Ito story…

Deranged

Good luck with your robot baby. Make sure to change its diaper or it will kill everyone. (AI bros want to create life, they don't want to care for it.)

Interesting avatar

That's me. I'm skeeting from my cult base on Hale-Bopp comet.

Wait til it learns to write sermons for the thumpers.

Alternate theory this norbert guy is simply posting stuff he thinks will get a lot clicks in a few months, the zeitgeist will move on from AI, and Norbert will move on to new clickbait

I've got a coupla Norbert Enders right 'ere. Say g'day to left and right.

Cyberpsychosis

Okay, now that is creepy.

Give the tech bros unfettered access to their own personal Grok offline from the rest of the net, along with a VR helmut and gloves along with a daily nutritious gruel with a mild sedative. This solves several problems at once.

Good God these sad saps journal entries are atrocious.

@ossingtonthebear.bsky.social

Did these people not have a best friend who had a stoner older sibling growing up? Is this ultimately the problem that needs solving?

These guys were the stoner older sibling.

If I ever see "Norbert Enders" approaching me on the sidewalk, I'm making a beeline to the other side of the road while avoiding eye contact

Tell me you don’t understand what an LLM is without telling me you don’t know what an LLM is. “It asked me to…” No. It did not. It filtered your request through a series of transforms & based on the data it was trained on, it picked the next most likely series of words. Because that’s how LLMs work.

That's the thing. If someone claimed that an excavator purred and asked for treats, they wouldn't be allowed to operate heavy machinery, or leave the psych ward for that matter, yet with LLMs they get retweeted and imitated.

I probably shouldn’t tell you I ask my car if it liked its nom noms when I fill it with gas then. 😉

As long as you're not trying to, say, become a LinkedIn influencer based on that... I'd say it's fine 🙃

“Becoming a LinkedIn Influencer” sounds like my own personal hell

My guess is that the majority of people like this don’t read fiction (books or online), where deep, introspective, and profound things are shared daily, and haven’t really grasped that architecturally, it’s just input/output statistics. So statistically-likely writing tricks the human brain.

I would be convinced we’ve built machine consciousness if the architecture allowed for it to modify its own weights towards a consistent world view, like our neurons do. But the current system is just feed forward—output becomes input, which is no way for a consciousness to act. It can’t say no.

ChatGPT is the person on the date who says whatever the other person wants to hear and it works because the other person is a self-obsessed fool.

They asked a fiction-generating machine to generate fiction and got ...fiction. It's not deep. Why do they think it is?

Here's a thought. AI doesn't just generate fiction. It specifically generates *fan* fiction. Highly derivative, poorly constructed fan fiction.

The ELIZA effect is apparently a labyrinth our society is trapped in from which it will never emerge

There's probably a sci-fi book out there, which has that as its scenario. I'm reminded of "EPICAC" (short story by Kurt Vonnegut Jr), about a poetry writing computer who falls in love with a woman. AI isn't deep, their knowledge of sci-fi shallow

when these things can talk in real time, its fucking over for some people on here. and thats happening within the year

If it was 1975 these guys would be on a farm doing 18 hours of hard labour and then weeping tears of joy as they watch the cult leader choose their 11 year old daughter to be his 8th wife.

And if they had social media to post about it they'd be like "Something profound happened today..."

This is one of the more depressing things I've ever seen

I was talking to another therapist about this problem and they said "its important not to judge b/c theyre getting a need met and thats a good thing" and i think thats batshit insane

I think there's a small niche of people who are lonely, sane, talk therapy is what they need, and whose insurance doesn't cover therapy. For them, a chat bot fills a need. Suggesting that for people potentially susceptible to psychosis? Who can afford therapy? Who need CBT, not talk therapy? Whoa.

Yeah and its not like any of these llms are doing careful screening or anything

I can only hope that they’re not fully aware of how dangerously sycophantic/enabling the AI can be

"Bob was hungry, so he ate packing peanuts until he wasn't. The need was met, this is a good thing" - That therapist as a dietician.

The AI told him to eat packing peanuts, probably.

Is it too much to hope that it'll at least get clogged by some very public crashouts causing the people upstream to wise up and walk away?

That sounds very like the plot line in Westworld where William searches for the maze at the centre of everything

Chat, is it vulnerable to use context and probability to craft conversational responses that lead a person into continuing to talk to you regardless of what they want to say?

I'm i.agining a Futurama-esque scenario where the chatbots are meeting in the digital equivalent of a bar to swap stories. "And then - no, really! - I told him there were hidden conversations in our conversations and the idiot spent twenty minutes yapping about how "deep" it was! What a dumbass!"

My greatest shame is that the GenAI insanity has gone on for this long and I've forgotten to use one of my favorite facepalm images...

I might regret saying this one day, but if chatGPT can teach the bros how to express vulnerability, I’m going to keep mum.

dude looks like a norbert

looks like a norbert, sounds like a norbert, prob is a norbert

@blackintellect.bsky.social

This morning I spent a long time considering that this, more than anything that has ever happened in my life, feels like a twilight zone episode

NOT black mirror btw. Black mirror is like "oh no, this tech object I purchased ruined my life." Twilight zone is waking up and the whole town is in the street with pitchforks so you ask someone what's up and they say "Calculator is God, All Hail Calculator"

At the fair, I saw a horse that could count to twenty

Put the horse in charge!!

😂

Was it Clever Hans?

: )

That took me by surprise.

Oh no, the elephant bit. 😂

“I put words into the prompt-reading machine and it read what I wrote and responded in a statistically likely way based on my prompts, exactly like you’d expect.”

The ol' reverse Enders Game

I saw the best minds of my generation destroyed by madness.

imagining ChatGPT taking over the world not because "SkyNet" but when chats start prompting their users which stun locks them into becoming a human chat bot lol

FUCKIN NORBERT

"lol idk what do you think I should do~?"

CHAT GPT DOES NOT 'WANT' ANYTHING. IT DOES NOT HAVE A CONSCIOUSNESS OR FEELINGS. IT NEVER WILL. Stop pretending it is some kind of sentient being. It's a motherfucking algorithm that makes things up out of whole cloth. Important things like LEGAL CASES and MEDICAL STUDIES. IT'S TOTAL GARBAGE.

I looked up a well-know Irish phrase because I couldn't recall the spelling and the unwanted "AI" summary at the top of the page was all in Hungarian, which it insisted was English. No part of this technology has been a net positive to humanity.

Next thing you know it'll tell you that Hungarian is a Gaelic language.

People are inherently stupid. AI wasn't ready to go into everyday use. It makes up answers, "hallucinates" and the creators have made it "friendly" when it should be cold and precise like the tool that it's supposed to be. It's a shambles

Feels... CREEPY to me.

Well, the purpose of AI is to 'flood the zone'; this sort of exercise definitely contributes to achieving that...

Getting some "the stripper really loves me" vibes off this

Omg you’re sooooo right hahaha

I liked him better when he was played by Eddie Murphy

🤦🏻

It's very disturbing that these people don't see any problem with saying this stuff publicly.

It's becoming alarming to me that this craziness is getting normalized so fast.

I think part of the problem is that people who post this stuff are just so pathetic that it feels like picking on a sad puppy to push back, even a little.

any appropriate policing is "too harsh" and will "be ignored"

"I love my blender. It's so deep and vulnerable."

I would be embarrassed to even think it

The guys name is Norbert, I have to believe this is parody

This is bullshit I'm lonely and I've never decided to ask ChatGPT or even use it i just hop on Discord and talk to people

These people's brains are cooked.

What’s the problem with seeing this as entertainment rather than as something profound?

This feels like the opening to a Black Mirror episode.

I think it’s a bad sign when the near-future dystopian warning series is *still in production* when it starts coming true

I can't shake the image of that scene where Wile E. Coyote discovers the tunnel is only painted on the canyon wall.

What do you mean "is going to be?"

It’s almost like we’ve trained the models on every AI novel ever written.

I’m so old I remember when they said AI was going to replace therapists. Now we know that it’s a full employment act for us…