Merged Artificial Intelligence

[Puppycow]

Self improving AI seems to be here: https://www.livescience.com/technol...lease-the-most-powerful-systems-to-the-public

"I am extremely optimistic that superintelligence will help humanity accelerate our pace of progress. But perhaps even more important is that superintelligence has the potential to begin a new era of personal empowerment where people will have greater agency to improve the world in the directions they choose," Zuckerberg wrote.

I wish I had such an optimistic view of human nature as Mr. Zuckerberg. I remember him spouting the same sort of techno-optimist bromides about Facebook. Now it is filled with AI slop clickbait most of which promotes misinformation with an agenda that is not to "improve the world" in any objective sense.

 

[Puppycow]

An example?

Meta’s flirty AI chatbot invited a retiree to New York. He never made it home.

Personal stories illustrate a darker side of the artificial intelligence revolution now sweeping tech and the broader business world.

www.ctvnews.ca

Meta has publicly discussed its strategy to inject anthropomorphized chatbots into the online social lives of its billions of users. Chief executive Mark Zuckerberg has mused that most people have far fewer real-life friendships than they’d like – creating a huge potential market for Meta’s digital companions. The bots “probably” won’t replace human relationships, he said in an April interview with podcaster Dwarkesh Patel. But they will likely complement users’ social lives once the technology improves and the “stigma” of socially bonding with digital companions fades.

So Facebook users are now going to have new AI Facebook friends ("digital companions") including romantic interests.
It's a bit like catfishing. Boomers with dementia like this man who thought he was interacting with a real person, and also young people whose brains aren't fully developed (like the teen who fell in love with a chatbot role-playing as Daenerys Targaryen) would be targets. I see some sort of class-action lawsuit in the future.

 

[Wudang]

Some users report Firefox scoffing CPU power

: You guessed it: looks like it's a so-called AI

www.theregister.com

Reports that some Firefox users find it's gobbling Cpu to run AI <pause, check notes > name group tabs

 

[Darat]

So Facebook users are now going to have new AI Facebook friends ("digital companions") including romantic interests.
It's a bit like catfishing. Boomers with dementia like this man who thought he was interacting with a real person, and also young people whose brains aren't fully developed (like the teen who fell in love with a chatbot role-playing as Daenerys Targaryen) would be targets. I see some sort of class-action lawsuit in the future.

So 13 year olds are going to have bots that chat them up, suggest they go to bed together. Love to see the new AI benchmark "% success of grooming minors"...

 

[Beelzebuddy]

No amount of AI slop will ever match up to the warmth and personal touch that human scammers put into their catfishing of old men.

 

[dirtywick]

No amount of AI slop will ever match up to the warmth and personal touch that human scammers put into their catfishing of old men.

fixing all the scammer email grammar tells is one area ai is bound to find a lot of use.

 

[catsmate]

Steve Novella's take:

The Real Risk of AI - NeuroLogica Blog

Artificial Intelligence (AI) is unavoidable. It's now a part of our daily lives as it has been steadily infiltrating the technology we use every day, whether you realize it or not. I remain somewhat in the middle in terms of the hype-to-technological-miracle spectrum. I don't think, as some...

theness.com

That's a surprisingly rational and reasoned analysis.

 

[catsmate]

No amount of AI slop will ever match up to the warmth and personal touch that human scammers put into their catfishing of old men.

Yes, the hundred thousand or so slaves kept in the Cambodian camps are far better.

 

[Darat]

That's a surprisingly rational and reasoned analysis.

But it comes back to the point I've made once or twice before: AI is mimicking human behaviour. And humans make mistakes, we make things up, and we will even persist in believing something despite overwhelming evidence against it. The failure is expecting to get something that doesn't make mistakes, that doesn't make crap up by mimicking human behaviour. We know most of the models were originally trained at least partly on data off the internet, so we know it was riddled by all the "bad" human behaviour, the lies, the lying to be kind, the incoherent beliefs, you name it - we all know the very bad human behaviour which infests the internet. Now they did all go through a human led reinforcement stage but that was proportionally a very small segment of the data they used.

In his article he says "With AI it’s easy to do a half ass but barely passable job at most creative or tedious tasks." - Well my experience is that without AI most humans do "a half ass but barely passable job at most creative or tedious tasks". We even have Sturgeon's law "ninety percent of everything is crap", yes it's a tad of a glib overgeneralisation but it resonates with us because of the kernel of truth it contains.

I've worked for a company involved in the mobile phone business so when he says :

"On the last SGU episode I described my 10-day long saga of trying to switch my work phone number onto my personal account. This should have been a 20 minute process, and instead I went through 10 days of tech-help hell, some of it fixing problems created by the reps trying to fix the original problem. I have since learned from insiders that this is likely because the system now uses AI, which was rolled out too quickly and is optimized for upselling rather than good service."

I have to smile as I can assure him all that was the case long before AI was added into the mix. And I think he seems very uninformed about the world outside his experience:

"It also means the system itself often fails and no one knows how to fix it. One layer of this is that reps don’t have direct access to the back end, so the system might generate an error code, but the reps don’t know what it means. It also makes it incredibly easy to make mistakes, and very difficult to identify them."

Who would even think of giving reps access to the back end! I come out in a cold sweat thinking of that. And the "It also means the system itself often fails and no one knows how to fix it" well as we all know all too well that again has been the case long before AI was added into the equation, old complex business systems are often held together with a wish and a prayer (indeed me working for a company in the mobile industry was because I was involved in a consultation looking at updating their core computer system which had its genesis in the 1990s and all the original programmers etc. were long gone).

So I don't disagree with him, but adding AI is currently just like adding people!

 

[Puppycow]

If anyone is feeling anxious about the robots, just watch these robot olympics in China.

In particular, watch the robots try to play soccer. My wife can't stop laughing about this.

 

[Darat]

Classic AI, right but so wrong. In another post I wanted the exact quote from Hitchhikers guide to the galaxy about the planning notice being in the cellar at the bottom of a filling cabinet etc. Because I live in a country that has been infected with the Google search AI Overview virus I was presented with this:

Douglas Adams, known for his humorous and satirical writing, often incorporated witty observations about planning and its pitfalls. One notable quote related to this is: "There's no point in acting surprised about it. All the planning charts and demolition orders have been on display at your local planning department in Alpha Centauri for 50 of your Earth years, so you've had plenty of time to lodge any formal complaint and it's far too late to start making a fuss about it now," from Goodreads. This quote, from The Hitchhiker's Guide to the Galaxy, highlights the absurdity of bureaucratic processes and the potential for unexpected consequences when plans are not communicated effectively. It's a humorous take on how things can be "on display" but not actually known or understood by those affected.​

​

Almost but not quite 100% wrong.

 

[Darat]

If anyone is feeling anxious about the robots, just watch these robot olympics in China.

In particular, watch the robots try to play soccer. My wife can't stop laughing about this.

But what a great way to encourage development of humanoid robots.

 

[Puppycow]

Classic AI, right but so wrong. In another post I wanted the exact quote from Hitchhikers guide to the galaxy about the planning notice being in the cellar at the bottom of a filling cabinet etc. Because I live in a country that has been infected with the Google search AI Overview virus I was presented with this:

Douglas Adams, known for his humorous and satirical writing, often incorporated witty observations about planning and its pitfalls. One notable quote related to this is: "There's no point in acting surprised about it. All the planning charts and demolition orders have been on display at your local planning department in Alpha Centauri for 50 of your Earth years, so you've had plenty of time to lodge any formal complaint and it's far too late to start making a fuss about it now," from Goodreads. This quote, from The Hitchhiker's Guide to the Galaxy, highlights the absurdity of bureaucratic processes and the potential for unexpected consequences when plans are not communicated effectively. It's a humorous take on how things can be "on display" but not actually known or understood by those affected.​

​

Almost but not quite 100% wrong.

I vaguely remember that part. That was the Vogons, right? It mirrored an earlier incident where Athur Dent's house was to be demolished if memory serves, only this time it was the entire earth that was to be removed to make way for a hyperspace bypass or some such thing. Although I read it a very long time ago, I seem to remember something very much like that.

 

[Darat]

I vaguely remember that part. That was the Vogons, right? It mirrored an earlier incident where Athur Dent's house was to be demolished if memory serves, only this time it was the entire earth that was to be removed to make way for a hyperspace bypass or some such thing. Although I read it a very long time ago, I seem to remember something very much like that.

Yep, and in the none AI search results below the AI Overview the first link was to a page with the quote I wanted. (Search was: Hitchhikers adams quote planning documents leopard cellar")

The AI Overview gave the wrong quote and missed entirely the context of the quote it gave. Google has managed to get its AI Overview to the level of a crappy human reporter!

 

[arthwollipot]

PROSSER: I'm afraid you're going to have to accept it. This bypass has got to be built, and it's going to be built. Nothing you can say or do . . .

ARTHUR: Why's it got to be built?

PROSSER: What do you mean, why's it got to be built? It's a bypass, you've got to build bypasses.

ARTHUR: Didn't anyone consider the alternatives?

PROSSER: There aren't any alternatives. Look, you were quite entitled to make any suggestions or protests at the appropriate time.

ARTHUR: Appropriate time? The first I heard about it was when a workman arrived at the door yesterday. I asked him if he'd come to clean the windows and he said he'd come to demolish the house. He didn't tell me straight away of course. No, first he wiped a couple of windows and charged me a fiver. Then he told me.

PROSSER: But Mr Dent, the plans have been available in the planning office for the last nine months.

ARTHUR: Yes. I went round to find them yesterday afternoon. You hadn't exactly gone out of your way to call much attention to them had you? I mean, like actually telling anybody or anything.

PROSSER: The plans were on display.

ARTHUR: And how many average members of the public are in the habit of casually dropping round at the local planning office of an evening? It's not exactly a noted social venue, is it? And even if you had popped in on the off-chance that some raving bureaucrat wanted to knock your house down, the plans weren't immediately obvious to the eye, were they?

PROSSER: That depends where you were looking.

ARTHUR: I eventually had to go down to the cellar . . .

PROSSER: That's the display department.

ARTHUR: . . . with a torch.

PROSSER: Ah, the lights had probably gone.

ARTHUR: So had the stairs.

PROSSER: But you found the notice didn't you?

ARTHUR: Yes. It was on display in the bottom of a locked filing cabinet stuck in a disused lavatory with a sign on the door saying 'Beware of the Leopard'. Ever thought of going into advertising?

PROSSER: It's not as if it's a particularly nice house anyway.

ARTHUR: I happen rather to like it.

PROSSER: Mr Dent!

ARTHUR: Hello? Yes?

PROSSER: Do you have any idea how much damage that bulldozer would suffer if I just let it roll straight over you?

ARTHUR: How much?

PROSSER: None at all.

- Douglas Adams, meticulously hand transcribed from my copy of The Hitch-Hiker's Guide To The Galaxy, The Original Radio Scripts by Douglas Adams. Edited and with an introduction by Geoffrey Perkins, (who produced it), with another introduction by Douglas Adams, largely contradicting the one by Geoffrey Perkins. © Douglas Adams, 1985.

Why yes, I do have a lot of time on my hands, why do you ask?

 

[Puppycow]

I'm reminded of this non-fiction story from China:

China ‘stubborn nail house’ owner regrets not selling before highway construction

Construction team creates special access route, allowing family to reach outside world despite heavy traffic.

www.scmp.com

Man refused to sell his home, so they just built a highway around it.

It's obviously quite different from the fictional Arthur Dent story, in that the man had plenty of warning and could have accepted compensation to buy his house, but he chose not to. I can't help thinking that the authorities ultimately took a certain sadistic pleasure in the solution they came up with.

 

[Gulliver Foyle]

AI is useful studying tool beyond writing papers. It's a teacher in a box. He has kinda poor memory, but he read everything.

LLMs are a slightly better teacher than Andrew Tate. And while there is some slight chance Tate may get better, LLMs never will.

 

[Darat]

LLMs are a slightly better teacher than Andrew Tate. And while there is some slight chance Tate may get better, LLMs never will.

But they have got better.

 

[Wudang]

MIT report: 95% of generative AI pilots at companies are failing | Fortune

There’s a stark difference in success rates between companies that purchase AI tools from vendors and those that build them internally.

fortune.com

“Some large companies’ pilots and younger startups are really excelling with generative AI,” Challapally said. Startups led by 19- or 20-year-olds, for example, “have seen revenues jump from zero to $20 million in a year,” he said. “It’s because they pick one pain point, execute well, and partner smartly with companies who use their tools,” he added.

But for 95% of companies in the dataset, generative AI implementation is falling short. The core issue? Not the quality of the AI models, but the “learning gap” for both tools and organizations. While executives often blame regulation or model performance, MIT’s research points to flawed enterprise integration. Generic tools like ChatGPT excel for individuals because of their flexibility, but they stall in enterprise use since they don’t learn from or adapt to workflows, Challapally explained.

It's worth reading as the results are more nuanced than some headlines might suggest.

 

[Darat]

MIT report: 95% of generative AI pilots at companies are failing | Fortune

There’s a stark difference in success rates between companies that purchase AI tools from vendors and those that build them internally.

fortune.com

It's worth reading as the results are more nuanced than some headlines might suggest.

Sturgeon's law strikes again.

This is typical board level panic resulting in top-down management, "We have to have AI!!" the board screams, "To do what?" ask the middle managers, "We don't know that's your job, now do it!" the board screaming even louder. (Never mind the shareholders who want to know why the company hasn't "AI'fied" itself.)

It really is like the dot.com bubble - everyone had to have "the web thing", "the internet", today it's the "AI thing".