Artificial Intelligence

[ri Alban]

Let's keep it to ourselves. 💰

[henrysmithsgloves]

In English 👍🏻

Edited June 8, 2023 by henrysmithsgloves

[JFK-1]

22 hours ago, Ulysses said:

An interesting article here that sets out some reasons why the techlords want to spook governments into worrying about future developments in AI, while ignoring the real and much more immediate dangers of the technology.

 

Artificial Intelligence and the Ever-Receding Horizon of the Future (techpolicy.press)

 

I don't dwell on an existential terminator style risk from AI, though obviously that's not impossible. I think one of the most immediate risks to large scale society is something I have mentioned before, the ability to create convincing material depicting anybody in any inappropriate situation.

 

It will happen to neighbours the instigator can then watch and even talk with about it. It will happen to work colleagues, it will happen right throughout the society and I can see this being devastating. Millions will believe anything they see on the internet. It's always been that way, grainy photo evolves into alien invasion, yeti and so on.

 

These inappropriate videos and pictures or whatever will be out there in the wild permanently and will stick to the victims. They might move away to escape it, then somebody in the new location recognises them as the pervert in that video. I think that's immediate and urgent.

[Ulysses]

1 hour ago, JFK-1 said:

 

I don't dwell on an existential terminator style risk from AI, though obviously that's not impossible. I think one of the most immediate risks to large scale society is something I have mentioned before, the ability to create convincing material depicting anybody in any inappropriate situation.

 

It will happen to neighbours the instigator can then watch and even talk with about it. It will happen to work colleagues, it will happen right throughout the society and I can see this being devastating. Millions will believe anything they see on the internet. It's always been that way, grainy photo evolves into alien invasion, yeti and so on.

 

These inappropriate videos and pictures or whatever will be out there in the wild permanently and will stick to the victims. They might move away to escape it, then somebody in the new location recognises them as the pervert in that video. I think that's immediate and urgent.

 

That's the kind of stuff the regulators want to regulate.  That's the kind of stuff the techlords want the regulators not to regulate.  Why?  Because there's more money for the techlords in letting ****ers loose with their software to do the kind of things you describe.

 

That is why we've heard so much shite in recent weeks from the tech firms talking about the fake and fantasy dangers of "AI" rather than its real and immediate dangers. 

[JFK-1]

12 minutes ago, Ulysses said:

 

That's the kind of stuff the regulators want to regulate.  That's the kind of stuff the techlords want the regulators not to regulate.  Why?  Because there's more money for the techlords in letting ****ers loose with their software to do the kind of things you describe.

 

That is why we've heard so much shite in recent weeks from the tech firms talking about the fake and fantasy dangers of "AI" rather than its real and immediate dangers. 

 

If the machines take over I hope it's like the Matrix, hell I would happily be a battery for the chance to leap from building to building and running around walls. Learn how to fly a helicopter in seconds, i'm up for that. Even a date with the girl in the red dress though apparently she doesn't say much. Bring on the AI and the Matrix. Morpheus is a terrorist. 😉

[Ulysses]

11 minutes ago, JFK-1 said:

 

If the machines take over I hope it's like the Matrix, hell I would happily be a battery for the chance to leap from building to building and running around walls. Learn how to fly a helicopter in seconds, i'm up for that. Even a date with the girl in the red dress though apparently she doesn't say much. Bring on the AI and the Matrix. Morpheus is a terrorist. 😉

 

 

Sounds cool.    

[JFK-1]

8 minutes ago, Ulysses said:

 

 

Sounds cool.    

 

Did you miss the part where I mentioned she doesn't say much, she would be deadly boring. What are you thinking of? Were you listening to me or looking at the woman in the red dress?

 

[Ulysses]

Just now, JFK-1 said:

 

Did you miss the part where I mentioned she doesn't say much, she would be deadly boring. What are you thinking of? Were you listening to me or looking at the woman in the red dress?

 

 

 

Actually she looks like quite a good conversationalist, dontcha think?   

[JFK-1]

12 minutes ago, Ulysses said:

 

 

Actually she looks like quite a good conversationalist, dontcha think?   

 

First time I noticed everyone around her is wearing black and white. Hell I must have been looking at the woman in the red dress. 😄

[Ulysses]

Just now, JFK-1 said:

 

First time I noticed everyone around her is wearing black and white. Hell I must have been looking at the woman in the red dress. 😄

 

Likewise.  That's natural intelligence for ya.   

[henrysmithsgloves]

3 hours ago, JFK-1 said:

 

Did you miss the part where I mentioned she doesn't say much, she would be deadly boring. What are you thinking of? Were you listening to me or looking at the woman in the red dress?

 

 

[Gizmo]

20 hours ago, JFK-1 said:

 

If the machines take over I hope it's like the Matrix, hell I would happily be a battery for the chance to leap from building to building and running around walls. Learn how to fly a helicopter in seconds, i'm up for that. Even a date with the girl in the red dress though apparently she doesn't say much. Bring on the AI and the Matrix. Morpheus is a terrorist. 😉

Nu-huh you don't. It's only the free and the few rescued from the Matrix that can harness it's power fully like instant learning etc. 

It's why Cypher wanted to be completely restored as someone important, otherwise he'd just be a battery and another drone stuck in the system. 

 

[Gizmo]

This is an interesting take, a little sideways from the usual "oh **** Skynet" talk.
 

 

[Ulysses]

4 minutes ago, Gizmo said:

This is an interesting take, a little sideways from the usual "oh **** Skynet" talk.
 

 

 

In a nutshell, what does he say?

 

(Too lazy to spend 4-5 minutes listening to it.)

[Gizmo]

7 minutes ago, Ulysses said:

 

In a nutshell, what does he say?

 

(Too lazy to spend 4-5 minutes listening to it.)

It's more how he looks at it, honestly worth the 5 minutes despite the zoomer video ratio. 

[Ulysses]

1 minute ago, Gizmo said:

It's more how he looks at it, honestly worth the 5 minutes despite the zoomer video ratio. 

 

Some day.  Maybe.   

[JFK-1]

2 minutes ago, Ulysses said:

 

Some day.  Maybe.   

 

I liked it, an interesting prelude going back to the Greeks and their treatment of the Gods, right through to a prediction that AI will replace us. But then what's new, that's not an uncommon view. Fry is into Greek mythology and I always was too. 

[Ulysses]

10 minutes ago, JFK-1 said:

 

I liked it, an interesting prelude going back to the Greeks and their treatment of the Gods, right through to a prediction that AI will replace us. But then what's new, that's not an uncommon view. Fry is into Greek mythology and I always was too. 

 

I know, and I quite like Fry.  I highlighted a really important word in your post, by the way.

[Ulysses]

[cannonfoda]

Used bard earlier today to listen to music then write out a composition for the kids brass band.  Would normally take hours to get everything sorted out for each instrument but bard was able to do it in seconds.  

 

One quick export into notational format later and jobs done.

 

 

[il Duce McTarkin]

2 hours ago, Ulysses said:

 

 

 

[JFK-1]

 

[Gizmo]

An illustration of how quickly AI image generation is improving. Image > video
 

 Also another wee example, from a wave shot I took at St Abbs.


AI was used to animate it, results here: https://imgur.com/a/PaBsNOP (couldn't get it to embed directly) - with zero input other than the base image. It's not perfect by any means but still quite impressive. 
 

[Lone Striker]

1 hour ago, Gizmo said:

An illustration of how quickly AI image generation is improving. Image > video
 

 Also another wee example, from a wave shot I took at St Abbs.


AI was used to animate it, results here: https://imgur.com/a/PaBsNOP (couldn't get it to embed directly) - with zero input other than the base image. It's not perfect by any means but still quite impressive. 
 

Very impressive.   

 

Then you see stories like this - https://www.bbc.co.uk/news/uk-england-merseyside-66348332

 

My guess is that  AI will mimic the internet itself in terms of how it develops and gets used.  80% benign & useful .... 20% criminal, dangerous  and downright sick.   The 20% can result in ruination for some unlucky folk.

 

    

 

 

[Gizmo]

6 minutes ago, Lone Striker said:

Very impressive.   

 

Then you see stories like this - https://www.bbc.co.uk/news/uk-england-merseyside-66348332

 

My guess is that  AI will mimic the internet itself in terms of how it develops and gets used.  80% benign & useful .... 20% criminal, dangerous  and downright sick.   The 20% can result in ruination for some unlucky folk.

 

I'm wondering how post-reality (ie knowing true reality from fake) is going to play out. Some of this stuff will in fairly short time become nigh indistinguishable from base reality. 

[Ray Gin]

2 hours ago, Gizmo said:

An illustration of how quickly AI image generation is improving. Image > video
 

 Also another wee example, from a wave shot I took at St Abbs.


AI was used to animate it, results here: https://imgur.com/a/PaBsNOP (couldn't get it to embed directly) - with zero input other than the base image. It's not perfect by any means but still quite impressive. 
 

 

Incredible

[Lone Striker]

3 hours ago, Gizmo said:

I'm wondering how post-reality (ie knowing true reality from fake) is going to play out. Some of this stuff will in fairly short time become nigh indistinguishable from base reality. 

Me too.  Generating film scenes without cameras is all well and fine (although the actors, writers  and production folk don't agree !!) - but AI seems like a gift to criminals & dangerous wrong'uns.   We've already seen fake messages come out the "mouths" of Zelensky and Martin Lewis - and as you say, the videos are almost indistinguishable from reality.  

 

Nightmare scenario would be if criminals manage to install AI bots into institutions like banks and HMRC and big retail companies holding our bank & personal  details.  Stealing secure data and using ransome-ware attacks are bad enough at the moment,  but imagine how much worse it could get if a cleverly programmed  AI bot  can somehow be installed on to their servers.

 

 

[Ulysses]

I couldn't find the thread about this year's Fringe, so I'm posting this here.  AI goes to Edinburgh.

 

The Guardian: A bot walks into a bar...

 

[JudyJudyJudy]

19 hours ago, Gizmo said:

An illustration of how quickly AI image generation is improving. Image > video
 

 Also another wee example, from a wave shot I took at St Abbs.


AI was used to animate it, results here: https://imgur.com/a/PaBsNOP (couldn't get it to embed directly) - with zero input other than the base image. It's not perfect by any means but still quite impressive. 
 

impressive. well done 

[JFK-1]

Artificial Intelligence police van detects drivers using mobile phones

 

A police spy camera van using Artificial Intelligence (AI) has detected hundreds of people using mobile phones at the wheel or not wearing seat belts.

Hampshire and Isle of Wight Constabulary and Thames Valley Police carried out the operation from 17 to 21 July, targeting commercial vehicles.

It took place on the A34 and the A303.

Simon Gomer, head of the safer roads unit, said the technology showed how prolific the offences were.

The AI-equipped van uses two cameras to catch the footage.

 

Full BBC article

 

[JFK-1]

 

AI Utopia: A Conversation with Nick Bostrom

 

Sam Harris speaks with Nick Bostrom about ongoing progress in artificial intelligence. They discuss the twin concerns about the failure of alignment and the failure to make progress, why smart people don’t perceive the risk of superintelligent AI, the governance risk, path dependence and "knotty problems"

The idea of a solved world, Keynes’s predictions about human productivity, the uncanny valley of utopia, the replacement of human labor and other activities, meaning and purpose, digital isolation and plugging into something like the Matrix, pure hedonism, the asymmetry between pleasure and pain, increasingly subtle distinctions in experience, artificial purpose

Altering human values at the level of the brain, ethical changes in the absence of extreme suffering, our cosmic endowment, longtermism, problems with consequentialism, the ethical conundrum of dealing with small probabilities of large outcomes, and other topics.

Nick Bostrom is a Professor at Oxford University, where he is the founding director of the Future of Humanity Institute. He is the author of more than 200 publications, including Anthropic Bias (2002), Global Catastrophic Risks (2008), Human Enhancement (2009), and Superintelligence: Paths, Dangers, Strategies (2014), a New York Times bestseller which sparked the global conversation about the future of AI.

His work has framed much of the current thinking around humanity’s future (such as the concept of existential risk, the simulation argument, the vulnerable world hypothesis, astronomical waste, and the unilateralist’s curse).

He has been on Foreign Policy’s Top 100 Global Thinkers list twice, and was the youngest person to rank among the top 15 in Prospect’s World Thinkers list. He has an academic background in theoretical physics, AI, computational neuroscience, and philosophy. His most recent book is Deep Utopia: Life and Meaning in a Solved World.
 

 

[Ray Gin]

I recently completed a job utilising AI to do a task that otherwise would have been beyond both me and anyone else at the company. 

 

Great for us but shite for any freelance web developer who would usually be hired for such a task. Sorry pal, your job is now obsolete.

[il Duce McTarkin]

3 hours ago, Ray Gin said:

I recently completed a job utilising AI to do a task that otherwise would have been beyond both me and anyone else at the company. 

 

Great for us but shite for any freelance web developer who would usually be hired for such a task. Sorry pal, your job is now obsolete.

 

My job will be obsolete in a few years (at most) too, no doubt. It's a slightly concerning prospect but also an opportunity to do something else. Pros and cons.

[JudyJudyJudy]

13 minutes ago, il Duce McTarkin said:

 

My job will be obsolete in a few years (at most) too, no doubt. It's a slightly concerning prospect but also an opportunity to do something else. Pros and cons.

 Not everyone 😝

 

 

[il Duce McTarkin]

13 minutes ago, JudyJudyJudy said:

 Not everyone 😝

 

 

 

😁

[Tazio]

[Watt-Zeefuik]

Apple has been spending God knows how much money on AI and the upshot of it seems to be that Siri can no longer do the basic stuff I've been doing with it for years. Like it can no longer find my spouse's name in the contacts and its voice transcription has become comically bad.

[JFK-1]

4 hours ago, Watt-Zeefuik said:

Apple has been spending God knows how much money on AI and the upshot of it seems to be that Siri can no longer do the basic stuff I've been doing with it for years. Like it can no longer find my spouse's name in the contacts and its voice transcription has become comically bad.

 

So Siri is bimbo AI

[JFK-1]

 

How A.I. Could Change Science Forever

 

 

It's getting harder to harder to ignore the potential disruptive power of AI in research. Scientists are already using AI tools but could the future lead to complete replacement of humans? How will our scientific institutions transform? These are difficult questions but ones we have to talk about in today's episode.

 

Written, presented & edited by Prof. David Kipping.

 

 

[Bigsmak]

Will we all have AI glasses in the near future? 

 

https://www.kickstarter.com/projects/looktech/looktech-smart-ai-glasses-and-hands-free-hd-camera?ref=e83mqy&utm_source=PledgeBox&utm_medium=Newsletter

[ƒιѕнρℓαρѕ]

AI's approach to cartoons is something to behold

 

[Ray Gin]

23 minutes ago, ｕｎｋｎｏｗｎｕｓｅｒ said:

AI's approach to cartoons is something to behold

 

 

[redjambo]

AI is useless . It's a long way from being mature enough to be both useful and safe. And yet companies are falling over themselves to release half-baked products just to be in on the latest buzzword. Sometimes it's as if we're actually living in Idiocracy.

[hughesie27]

7 hours ago, redjambo said:

AI is useless . It's a long way from being mature enough to be both useful and safe. And yet companies are falling over themselves to release half-baked products just to be in on the latest buzzword. Sometimes it's as if we're actually living in Idiocracy.

AI is a pretty poor term imo. It encompasses such a vast variety of things that it's unhelpful to say AI is useless. 

Examples like the comic strip above are obviously funny/crap but there are dozens of other examples of it being put to incredible use. And anothe4 dozen where its stupid.

 

My phone for example can live translate a phone call I am on with someone speaking Chinese to me. And both of us will be understand each side of the conversation. That's AI at its best. 

[Ray Gin]

8 hours ago, redjambo said:

AI is useless . 

 

I've been using it at work so this is not correct.

 

[Watt-Zeefuik]

On 07/01/2025 at 20:15, redjambo said:

AI is useless . It's a long way from being mature enough to be both useful and safe. And yet companies are falling over themselves to release half-baked products just to be in on the latest buzzword. Sometimes it's as if we're actually living in Idiocracy.

 

LLMs aren't useless, but they're only actually useful for about 5% (at best) of the stuff they're being promoted for. If you need something to read a bunch of text and do a standard transformation on that text, and aren't too fussed over a decent error rate, they can do absolutely amazing things. We've got a couple deployments of them at work that are all about processing and recognizing terms in text, and yeah, we can really improve some usability in some stuff.

 

But that's a really limited set of conditions.

 

A good couple of studies that came out last year showed that teams that integrated AI into their work generally saw their overall productivity drop. It was in a couple of different fields but at least a couple of them were where you'd expect the generative models to be above average.

 

Bottom line, the "hallucination" problem will never go away with the current round of LLM models and the newer models which promised so much more aren't delivering. OpenAI, the biggest player in town, has no possible route to profitability and will end up getting absorbed by Microsoft and their models will get turned into the new Clippy.

 

The good news is that these companies are also dumping gigantic sums of venture capitalist money into solar and wind farms to power all the data centers that are going to go dark when this all goes belly-up, so then we can use them for other things! 😉 

[ƒιѕнρℓαρѕ]

17 hours ago, hughesie27 said:

AI is a pretty poor term imo. It encompasses such a vast variety of things that it's unhelpful to say AI is useless. 

Examples like the comic strip above are obviously funny/crap but there are dozens of other examples of it being put to incredible use. And anothe4 dozen where its stupid.

 

My phone for example can live translate a phone call I am on with someone speaking Chinese to me. And both of us will be understand each side of the conversation. That's AI at its best. 

 

I got AI to teach me how to make butane THC extract.

Useless eh?

[Gundermann]

Something else to *** the planet with.

 

https://www.newsbreak.com/news/3758590522243-voices-i-ve-been-working-in-ai-for-years-there-s-one-big-problem-no-one-is-really-addressing

[Lord Montpelier]

8 minutes ago, Gundermann said:

Something else to *** the planet with.

 

https://www.newsbreak.com/news/3758590522243-voices-i-ve-been-working-in-ai-for-years-there-s-one-big-problem-no-one-is-really-addressing

Heres what may happen. AI becomes so advanced - with quantum computers- that service providers find it impossible to create security algorithms that cant be cracked. Leading to the collapse of the internet as we know it as its rendered unusable perhaps towards its original intention of knowledge sharing and JKB c.1997 type forums. 

[Ulysses]

I watched a video today which purported to be an interaction between an Irish person and Deepseek AI.

 

I'm not sure which was the more concerning, the fact that it refused to discuss Tiananmen Square, or the fact that it did so in fluent Irish.   

 

Might be a fake, though.