AI fails

[icowden]

That does appear to be standard across most car brands. VW's is in my experience, is somewhere between useless and willfull disobedience.

Indeed "Hi ID - please warm my feet"
No NOT my seat. My FEET
Hi ID turn off seat heater.
NO DON'T TURN OFF THE HEATING!!

<presses heating touch buttons instead unless it's dark when it's nigh on impossible>
<reaches halfway across car to turn off steering wheel heater because it's still set up for left hand drive cars>

 

[briantrumpet]

Indeed "Hi ID - please warm my feet"
No NOT my seat. My FEET
Hi ID turn off seat heater.
NO DON'T TURN OFF THE HEATING!!

<presses heating touch buttons instead unless it's dark when it's nigh on impossible>

 

[briantrumpet]

https://www.businessinsider.com/ai-tools-are-deskilling-workers-philosophy-professor-2025-11

Berg's point is that AI doesn't merely automate tasks — it automates the very processes through which people develop their skills.

Once workers grow dependent on AI, they lose the friction that strengthens their ability to reason, problem-solve, and make decisions.

"We have them compromising their most basic levels of their ability," she said. "The threat to the highest level of their ability is just tremendous."

If companies continue to push AI into every workflow under the banner of efficiency, she said, they may end up with a generation of employees who appear more productive on paper but lack the ability to perform without digital hand-holding.

In other words, AI might not be enhancing the workforce. It might be slowly dismantling it.

You can download the whole paper here: https://zenodo.org/records/17065099 - I've just done so, and am about to digest it and the liver & bacon I've just eaten. EDIT - it's taking a clear stance against the blanket use of AI in academia - i.e. it is biased.

 

[briantrumpet]

It's forthrightly combative. Here's one of the more general broadsides quoted:

Even if you disagree strongly with the thrust ('Against the Uncritical Adoption of AI Technologies in Academia'), there are a number of threads running through its reasoning which are worth considering — philosophical, ethical, cognitive, educational et al — and how the risks of AI (particularly in education) need deep thought and thorough assessment for the risks. They make an analogy with smoking, and how the tobacco industry pushed a product on the public which they knew was addictive and dangerous.

I hadn't realised quite how long 'AI' had been making grandiose claims...

Anyway, tl;dr - there's lots of thought-provoking stuff in there.

 

[briantrumpet]

Ha, I've just had a rather fun dialogue on Bluesky with the author of that, with me suggesting that the preposterousness of someone learning a musical instrument via AI (i.e., there is no part of the learning process which would be meaningfully accelerated via its use without a loss of essential skill/understanding) could highlight the problematic areas she's concerned with in developing academic understanding in students.

I suggested that my pupils might be receptive to my worries about using AI unwisely because they know that learning a musical instrument in a long, hard process, for which there are no shortcuts. And that's exactly what makes it rewarding, when you stick at it. No easy fixes.

 

[Pblakeney]

And that's exactly what makes it rewarding, when you stick at it. No easy fixes.

I'm not sure that you are up to speed with modern life. Everything available, now! And easy with as little effort as possible.
Just to be on the safe side, 😉

 

[briantrumpet]

I'm not sure that you are up to speed with modern life. Everything available, now! And easy with as little effort as possible.
Just to be on the safe side, 😉

Well, all those South West youngsters who have learnt via to play the 1st trumpet part for Bach's Mass in B Minor must be hiding, cos there are just three of us old fogeys in the whole of Somerset, Devon and Cornwall who will do it... to be honest, it would be nice if there were more!!

 

[Pblakeney]

Well, all those South West youngsters who have learnt via to play the 1st trumpet part for Bach's Mass in B Minor must be hiding, cos there are just three of us old fogeys in the whole of Somerset, Devon and Cornwall who will do it... to be honest, it would be nice if there were more!!

That's cos it's not easy. 😉

 

[briantrumpet]

That's cos it's not easy. 😉

Seems you might be right.

 

[briantrumpet]

Sceptical me doth like this quote.

https://www.thenewworld.co.uk/james-ball-cory-doctorow-the-big-fake/

 

[briantrumpet]

Sceptical me doth like this quote.

View attachment 11398

https://www.thenewworld.co.uk/james-ball-cory-doctorow-the-big-fake/

It's actually a really good article that ranges more widely than just AI. Worth a read.

 

[briantrumpet]

 

[briantrumpet]

Another long polemical read.

https://www.currentaffairs.org/news/ai-is-destroying-the-university-and-learning-itself

In classrooms today, the technopoly is thriving. Universities are being retrofitted as fulfillment centers of cognitive convenience. Students aren’t being taught to think more deeply but to prompt more effectively. We are exporting the very labor of teaching and learning—the slow work of wrestling with ideas, the enduring of discomfort, doubt and confusion, the struggle of finding one’s own voice. Critical pedagogy is out; productivity hacks are in. What’s sold as innovation is really surrender. As the university trades its teaching mission for “AI-tech integration,” it doesn’t just risk irrelevance—it risks becoming mechanically soulless. Genuine intellectual struggle has become too expensive of a value proposition.

The scandal is not one of ignorance but indifference. University administrators understand exactly what’s happening, and proceed anyway. As long as enrollment numbers hold and tuition checks clear, they turn a blind eye to the learning crisis while faculty are left to manage the educational carnage in their classrooms.

The future of education has already arrived–as a liquidation sale of everything that once made it matter.

 

[briantrumpet]

In case you don't make it that far (as I say, it's a long read)

When critics push back against this corporate evangelism, the reply—like Roy Lee’s—is predictable: we’re accused of “moral panic” over inevitable progress, with the old invocation of Socrates’ anxiety about writing to suggest today’s AI fears are mere nostalgia. Tech luminaries such as Reid Hoffman make this argument, urging “iterative deployment” and insisting our “sense of urgency needs to match the current speed of change”—learn-by-shipping, fix later. He recasts precaution as “problemism” and labels skeptics as “Gloomers,” claiming that slowing or pausing AI would only preempt its benefits.

But the analogy is flawed. Earlier technologies expanded human agency over generations; this one seeks to replace cognition at platform speed (the launch of ChatGPT hit 100 million users in two months),while the public is conscripted into the experiment “hands-on” after release. Hoffman concedes the democratic catch: broad participation slows innovation, so faster progress may come from “more authoritarian…countries.” Far from an answer to moral panic, this is an argument for outrunning consent.

 

[Pross]

The latest Michael Connelly Lincoln Lawyer novel is a good story based around AI development. Basically a legal case against a company for failing to put enough guards in place / ignoring concerns of an ethicist resulting in the prejudices of the programmer leading to a person to committ murder. I couldn't think of anyone that might be behind the storyline idea, obviously completely made up fiction.