this post was submitted on 10 Mar 2025

27 points (100.0% liked)

TechTakes

1750 readers

57 users here now

Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.

This is not debate club. Unless it’s amusing debate.

For actually-good tech, you want our NotAwfulTech community

founded 2 years ago

MODERATORS

[email protected]

Foreign Policy: The U.K. Pivot to AI Is Doomed From the Start (foreignpolicy.com)

submitted 3 weeks ago by [email protected] to c/[email protected]

18 comments fedilink hide all child comments

at frickin last - I wrote this in January and it's finally been published

archive: https://archive.is/RSU91

top 18 comments

sorted by: hot top controversial new old

[–] [email protected] 6 points 3 weeks ago* (last edited 3 weeks ago)

The more I use various LLMs, the more I've come to realise that they have a tendency to confidently lie. More often than not, it seems an LLM will give me the answer it thinks I want to hear, even if the details of it's answer are factually incorrect.

Using these tools to decide and affect real peoples lives is a very dangerous prospect.

Interesting article. Thanks

[–] [email protected] 5 points 3 weeks ago (1 children)

Thank you for this much needed reality check. I don't understand why the Government are doing venture capital's bidding.

[–] [email protected] 6 points 3 weeks ago

Labour are all about the billionaire donors these days.

[+] [email protected] -9 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

Presumably 'AI' can make simple rules based decisions, if done properly (unfortunately, being the UK government this is a big 'if').

But what exactly is sacking a million people supposed to do to the economy?

[–] [email protected] 12 points 3 weeks ago (1 children)

Presumably ‘AI’ can make simple rules based decisions, if done properl

honest question: was this meant seriously, or in jest?

[+] [email protected] -8 points 3 weeks ago (4 children)

Serious.

Fill in form online
AI analyses it, decides if applicant is entitled to benefits.

Why do you ask the question?

[–] [email protected] 12 points 3 weeks ago (1 children)

why do you think hallucinating autocomplete can make rules-based decisions reliably

AI analyses it, decides if applicant is entitled to benefits.

why do you think this is simple

[–] [email protected] 10 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

You should not need an AI to do that if it's not a freeform text input?

[+] [email protected] -6 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

Who knows what the British government acutally means when they say 'AI'? I doubt they do.

And I assume there is a sliding scale from what I described my excel speadsheet is capable of today, to whatever it is they hope AI will eventually interpret. And in fact barring the inevitable fuckups AI probably can eventual handle a lot of interpretation currently carried out by human civil servants.

But honestly I would have thought that all of this is obvious, and that I shouldn't really have to articulate it.

My main point is that they claim they will save one million jobs (so sack one million people) and this will somehow boost the UK economy. I don't see how it can.

Covid really showed that most jobs are not essential. I would suggest most jobs exist because the people exist to do them: a situation an increasingly small number of people have got themselves into a position to profit from.

What happens when AI takes away swathes of bullshit jobs but the people still exist? Silicon valley is hoping to make AI functional enough essentially leave somebody else holding the bag, after emptying it of goodies. For some reason, the UK government seem to think this will be a boost to the UK economy: I'll assume basic cluelessness until I see some other reasonable suggestion.

[–] [email protected] 13 points 3 weeks ago (1 children)

And in fact barring the inevitable fuckups AI probably can eventual handle a lot of interpretation currently carried out by human civil servants.

But honestly I would have thought that all of this is obvious, and that I shouldn’t really have to articulate it.

you keep making claims about what LLMs are capable of that don’t match with any known reality outside of OpenAI and friends’ marketing, dodging anyone who asks you to explain, and acting like a bit of a shit about it. I don’t think we need your posts.

[–] [email protected] 8 points 3 weeks ago (1 children)

the post history is very infosec dot pub

[–] [email protected] 10 points 3 weeks ago (1 children)

a terrible place for both information and security

[–] [email protected] 9 points 3 weeks ago (1 children)

and a terrible pub

[–] [email protected] 5 points 3 weeks ago

Jury's still out on the Damage Over Time effect.

[–] [email protected] 6 points 3 weeks ago

citation/link/reference, please

[–] [email protected] 6 points 3 weeks ago

"AI" in the context of the article is "LLMs". So, the definition of not trustworthy.