I strongly believe there are entire companies right now under heavy AI psychosis and its impossible to have rational conversations about it with them.

[bms48@mastodon.social]

@lizbian @mitchellh @briankrebs Prof. Michael Wooldridge has expressed clearly, for wider dissemination on his Royal Society lecture from February, many of the problems recapitulated in this thread; the zingers are 30m in or so and he flies cover very carefully at the start. This is where I'd point civilians. But the odd Condescending Wonka meme may help. https://www.youtube.com/watch?v=CyyL0yDhr7I

[w_gulley@mstdn.social]

@mitchellh Were you alive in the 1970's when GM's Van Nuys plant produced so many malfunctioning cars that they couldn't repair them fast enough to keep them from piling up on the lot there in Van Nuys? We'll fix them later, just get them off the line was the mantra. But they didn't. Fix them later. Sound familiar? Yes, but this time it will be different. Riiiiight.

[klausfiend@mstdn.ca]

@mitchellh I used Claude today to verify a change I'd made to a piece of Python code. Instead of using a statically defined tuple of valid subclass names, it suggested using a pattern I found later on StackOverflow that it stated was "drift-proof" (which is what I wanted, to reduce toil in the code base.)

It offered me something that 1) failed right away due a missing import it didn't mention in its suggestion and 2) the code it suggested returned an empty tuple when I used the suggested pattern. I shrugged, and reverted to the manual method I had originally written.

Claude is arguably the most robust model for software engineering and I find it does dumb shit _all the fucking time_ ... and so many people just blithely assume it's right without checking.

[richpuchalsky@mastodon.social]

@mitchellh

It's not "AI psychosis". It's just bad programmers and people who don't know anything about anything, and that suddenly has been revealed now that they don't have to pretend to do anything.

[cyberwitch@goingdark.social]

@mitchellh I toss the occasional "Where's your Ed At?" podcast their way. Sometimes the best grounding force is a Brit swearing in your ear for 40 minutes about trading made-up fart tokens.

[gatesvp@mstdn.ca]

@mitchellh I think you are on to something being up in corporate behaviour, I'm not sure if it's attributable directly to "AI psychosis" as much as companies leveraging the robots to serve their basest instincts.

You and I can look at a system like GitHub or Windows and say that it has gone to complete trash. We can say that the cost of the downtime (for us) exceeds the value of the new features coming in.

But that math appears to be different for their customers. We're not seeing a mass exodus from GH or from MS products. All of that MTTR math is proving out. Failures are way less costly than developers kept claiming. Shipping is way more important.

Corporations are finally unleashed to ship trash and keep collecting subscriptions at an unprecedented rate.

Line must go up

I personally think this is all bad. But I don't know that it's "AI psychosis" rather than just basic "corporate psychosis".

[hopeless@mas.to]

@mitchellh Let's hope OP never decides to get into politics. Because demonizing everyone who reached a different conclusion than him with "psychosis" came very naturally. There's no justification given for the use of the word in this context either.

concern troll - Wiktionary, the free dictionary

Wiktionary (en.wiktionary.org)

[unlofl@mstdn.social]

@hopeless Its a metaphor and the post explains their concern pretty well

Can anyone help me find the downvote button here? My fedi client still only has one for "fav"

[johannab@cosocial.ca]

@pojntfx @mitchellh holy goats, hadn’t hear that dementia analogy before but that is exactly it. I’ve lost elder family to dementia and when you’ve lived with it you realize that it is so much more than “forgetting”, it is literal decay of executive, cognitive capability. Not sure i hould say thanks for sharing that, i’m now going to see it everywhere.

[nikclayton@mastodon.social]

@mitchellh does this mean you're going to remove the "maintainers are exempt" language from https://github.com/ghostty-org/ghostty/blob/main/AI_POLICY.md ?

[tmcfarlane@toot.community]

@mitchellh memories of the "microservices fix everything" brigade. We threw out easy integration testing and embraced distributed SPoFs

[jasongorman@mastodon.cloud]

@mitchellh Also, the data's showing very clearly that MTTR is getting longer, probably because of growing comprehension debt.

[kekunplazas@mamot.fr]

@lizbian @mitchellh @briankrebs Yep, exactly my point since a pretty long time.

[cczona@hachyderm.io]

@pojntfx I lost a loved one to dementia, and have been saying for a while how similar LLM psychosis is to that experience.

Carina C. Zona (@cczona@hachyderm.io)

@Di4na @jenniferplusplus and there is already evidence emerging that engineers who depend on LLMs to write their code for them are eroding their skills. I would analogize it to early stage dementia. The person can't see how their judgement is gradually developing fissures that compromise their ability to function. Eventually it will become too clear to deny anymore. But right now they are increasingly impaired while no less confident in the comprehensiveness of their skills. It's the period when they present a big risk to self and others, because of the growing gap between reality and perception of competence. This person is letting LLMs draft most of their code, and fails to see that not continuing to hone their own skills as an active coder has personal consequences; and that doing so en masse poses societal consequences. What happens in a generation when there are virtually no engineers left who can review a LLM's outputs competently?

Hachyderm.io (hachyderm.io)

Carina C. Zona (@cczona@hachyderm.io)

@janl@narrativ.es I do feel in mourning. It's been very similar to when a relative developed dementia. The person I knew disappeared in jolting episodes, replaced by someone with their face but an unrecognizable personality and stripped of the values they'd held dear. It was deeply troubling and so, so sad to be around. In the past year I've been feeling that same sense of loss. Moving through the stages of grief for one person, then another, another. It's mentally exhausting. But when you care about someone or something, it's difficult to just decide to not let it matter.

Hachyderm.io (hachyderm.io)

They think they are fine, while the person you used to know is gradually replaced by someone unrecognisable. Professional de-skilling issues aside, a sadder dimension to me is how AI psychosis degrades not only cognition but also social skills. So relationships deteriorate too, and at scale that will ultimately become communities and solidarity deteriorating as well. It's been disturbing to see people I care about lose interest in connecting at a human level, while grasping for the empty flattery of a large language model. Is this our future, truly?

[immibis@social.immibis.com]

@SpaceLifeForm @mitchellh "When the cloud is down, it's foggy" - some kind of strange metaphor or pun

[immibis@social.immibis.com]

@unlofl @hopeless There's no fediverse downvote by design, and the upvote/like/star doesn't do very much either. You can block people who continually spout nonsense.

[tenpasttwo@mas.to]

@mitchellh
In the 90s we only optimised to MTBF, MTTR wasn't considered terribly helpful. A software deployment could run for months without restarts or updates. But back then a person could hold an entire stack in their minds. I recall my first realisation of the total shift to MTTR with an ML based written record reader. We outsourced the ML, each instance could read only 1 record before dying. My boss was nonfussed, just restart it anew for every record... Infra is now just bubbles popping.

[totoroot@ibe.social]

@mitchellh@hachyderm.io Thanks for stating what many are seeing in our industry. Especially since your opinion bears a lot of weight.

I myself am trying to figure out how to voice concerns in the workplace and in the Open Source communities I participate in, while trying to not completely disregard the excitement people feel.

It becomes hard to argue with colleagues about the pros and cons of using LLMs in software engineering, when the CTO suggests an AI-first approach to anything, at any chance they get.

[ambrosia_engine@mastodon.social]

@mitchellh I'm doing a PhD on the epistemic risks of LLMs. I used to work in industry and I see this too and it's so terrifying. Currently writing a paper looking at one of the mechanisms at play here, but how do we even talk to these people and get them help anymore? Especially when it's clear entire startups are affected.

[kaffando@mastodon.social]

@RebelGeek99 @mitchellh @DukeDuke

What you've just said....yes. Exactly that.

Yes.

#COVIDBrain #AISlop