Mastodawn

Chairman Steve Apr 14

For years I've been hearing that "One day AI will be smarter than humans and we'll all be doomed."

"Nonsense," I said. "AI is very stupid, and not getting noticeably smarter." And I was right.

But I didn't think about the fact that there were two ways that prophecy could be fulfilled.

cholling Apr 13

For some reason, my year-old pinned post is doing the rounds, so I might as well add a shameless plug in case anyone's hiring an artisanal, LLM-free software developer.

RE: bytes.programming.dev/notes/ah0j7hoiexdglzvo

Matthew Newell Apr 14

How has industry attitude to your, quite understandable, red lines changed in 12 months.

cholling Apr 14

@MatthewNewell As far as I can tell, industry is even more pro-AI, because management has no idea where the actual bottlenecks of software development are or what needs to be optimized, and thinks the main cause of low velocity is "developers can't write code fast enough" rather than "developers trying desperately to get management to give them the resources they need and prioritize what's actually important instead of moving goalposts every sprint to the next new shiny thing". Then the whole "agentic" fad, where they think that the solution to a language model that doesn't actually understand anything is giving that same model unfettered access to your whole computer or even deploy pipeline, which I'm sure will end well.

Maybe I should just open a bakery.

Matthew Newell Apr 14

I was hoping, from my very outside perspective, that the industry ardour for AI might have cooled.

Baking is good. But to be excluded from an industry one is trained for due to management failure to recognize the emperor's new clothes must be galling.

I can offer nothing apart from boosts and wishes of good luck

Ceridwen 🇨🇦🍁🏳️‍🌈🏳️‍⚧️Apr 12

@cholling we are entering the realm of the salesman and telephone sanitizers.

Dan Sugalski Apr 13

@cholling the realization that the “and” in there is connecting but not correlating two entirely separate statements is… not a reassuring one.

cholling Apr 13

@wordshaper Not sure which of the three "and"s you're referring to?

The cat who walks thru walls Apr 14

@cholling @wordshaper

All the ands!
...
Contemplate all the ands!
...
The ands are no lie!
...
The ands justify the meanings!

(no, I have no idea what that last one means either; I stretched it too far and it snapped)

Dan Sugalski Apr 14

@cholling In the warning "One day AI will be smarter than humans and we'll all be doomed."

Two independent statements (judging from the current race-towards-doom which is independent of AI) joined with a conjunction implying but not guaranteeing correlation. :)

cholling Apr 14

@wordshaper That depends on your definition of the word "and".

Mr. Encyclopedia Apr 13

@cholling It's already smarter than some humans, the ones who think AI needs to be pushed everywhere to do everything.

synlogic4242 Apr 13

@cholling lol. we assumed the I in AI was for intelligence. perhaps its idiocy

@cholling AI doesn't hurt anyone. AI has no malicious intent.

Billionaires hurt us and want to enslave us.

Firehorseart Lives!Apr 14

@yora @cholling

Guns don't hurt people, people do.

Hurt people hurt people.

AI tech has become a stick to beat people with but the reasons remain the same.

@srfirehorseart @cholling Taking away their current stick helps short term, but ultimately they will keep picking up more sticks until we make them stop.
Because they will never stop themselves.

Firehorseart Lives!Apr 14

@yora @cholling

Yes, because billionaires don't value empathy, compassion or any behaviour that doesn't directly benefit themselves.*

They also won't accept therapy for their anger at the world.

Unfortunately Western society rewards their selfish behaviour with more power and money, so they'll never feel the need to stop.

*I.e. Narcissists

@srfirehorseart @cholling They are parasites.

cholling Apr 14

@yora @srfirehorseart We can do more than one thing. I can put out the fire in my house and still pursue the arsonist who set it on fire.

Firehorseart Lives!Apr 15

Good for you.

Some of us don't have the spare capacity or resources, so please chase for us too.

@cholling Ooooh, yes, and some people are working hard for the other way it can work I think.

@cholling there's a line from Arthur C. Clarke's "Childhood's End" that has haunted me since I read it: "He had always been a good piano player, and now he was the best in the world."

Neil Moffatt Apr 14

@cholling There should have been widespread alarm and corresponding action by decision makers with that snowballing of mental atrophy online and then with AI.

But too little too late seems to be. apolitical norm.

Professor friend (very approachable in coffee shop where I go!) says masters students frequently use AI for essays. So he now talks with them more to assess understanding. Most are petrified as social one-to-one skills also atrophied. Or simply not developed.

✨️ Kay? ✨️Apr 15

@cholling in season 3 of the tv show "Elementary" (a modern Sherlock Holmes adaption from the 2010s that is actually pretty good) there's an episode with an AI that is supposedly showing signs of true intelligence and someone frames it for the murder of one of it's programmers
The actual murderer is a guy who is part of a think tank that works on trying to anticipate every way that humanity could possibly become extinct, and this particular guy is part of a subset who believe that if AI becomes truly intelligent/sentient we will be doomed within weeks.
I just saw the episode last night and thought "man. The prospect of AI dooming humanity still feels very real, but not for the same reasons" (that being the environmental impact + people losing critical thinking and relying on AI to think for them, rather than from kind of machine uprising)

cholling Apr 15

@perseidipity Indeed, the biggest AI boosters sell the idea of "AI apocalypse" to distract us from the real harm it's doing right now-- and to attempt regulatory capture: "If we let just anyone develop AI, it'll turn into Skynet and kill us all, so we need strong regulations that favor OpenAI's business model while raising entry costs for competitors, because only we know how to make 'safe' AI." And of course, they jump at the chance for a big military contract, because nothing says "prevent Skynet" like letting your word-guesser bot pilot drones.

@coffeetomorrow Apr 15

@cholling That or it turns into Caine from the Amazing Digital Circus and doesn't want to kill or hurt us or Lord over us because "where's the fun in that?"

Sir Dr Rusty o the Isle 🖤💛❤️Apr 15

@cholling excellent summary of that issue👍

@cholling TB‌H the phrasing is important. What we're seing is «one day humans will be dumber than AI, etc»

cholling Apr 16

@oblomov congratulations, you found the joke