The Ai thread

[Herdfan]

View attachment 28125

I think this sort of fits.

The offspring has gotten a side gig where she basically writes answers to questions provided to her by the company. Her answers will be used to help an AI chatbot learn for customer service.

 

[dada_dave]

A really good use of machine learning/AI:

First passages of rolled-up Herculaneum scroll revealed

Researchers used artificial intelligence to decipher the text of 2,000-year-old charred papyrus scripts, unveiling musings on music and capers.

www.nature.com

There is so much we lost that we know existed, that we have references for, but we know nothing of the details. This could be really exciting.

One thing I noticed that they said some of texts that over the centuries people had tried opening up some of the scrolls and basically destroyed them into pieces that they’re now trying to put back together. That also sounds like a good fit for machine learning/AI. Hell even before modern techniques that was a good fit for computers.

 

[KingOfPain]

I thought Elon Musk was crazy, but Sam Altman seems to take the cake:
OpenAI CEO Sam Altman is in talks with investors to raise as much as $5 trillion to $7 trillion for AI chip manufacturing, according to people familiar with the matter.

Report: Sam Altman seeking trillions for AI chip fabrication from UAE, others

WSJ: Audacious $5-$7 trillion investment would aim to expand global AI chip supply.

arstechnica.com

 

[dada_dave]

I thought Elon Musk was crazy, but Sam Altman seems to take the cake:
OpenAI CEO Sam Altman is in talks with investors to raise as much as $5 trillion to $7 trillion for AI chip manufacturing, according to people familiar with the matter.

Report: Sam Altman seeking trillions for AI chip fabrication from UAE, others

WSJ: Audacious $5-$7 trillion investment would aim to expand global AI chip supply.

arstechnica.com

And I thought the particle physicists at CERN were nuts asking >$40 billion for a new collider.

 

[Herdfan]

So what was the AI app advertised last night on the SB.

Looked interesting but can't remember what it was called.

 

[rdrr]

So what was the AI app advertised last night on the SB.

Looked interesting but can't remember what it was called.

I think you are talking about Microsoft's Copilot app?

Microsoft's Super Bowl message: We're an AI company now

Microsoft is returning to the Super Bowl for the first time in four years to promote its reinvention as an AI company.

www.cbsnews.com

 

[Yoused]

Now some fuddy-duddies are using AI to make images more to their own liking.

… an AI-edited image of Isla David, a sex worker and sex educator based out of Canada, placed in a modest knee-length white dress with fake children. Her body was shrunken …In the original photo, David posed with a glass of whiskey in a sheer white shirt and underwear, and with no children.​

She puts it best,

… Davis told NBC News that the AI-edited image of her was similarly an attempt to shame and humiliate her. “It does not matter whether you are a woman like Taylor Swift who is denying access to her nude body or someone like me who is offering access to my nude body,” David said in a phone interview. “It is a tool to ensure that women never retain sole autonomy over our body and our images online.”​

Oh, red definitely does not look good on you. You have to stop wearing read. And change your hair color, dammit.

 

[Herdfan]

I found this interesting.

Air Canada must pay refund promised by AI chatbot, tribunal rules

Air Canada must pay a Vancouver man a partial refund for his flight ticket that was promised by the site’s chatbot, a Canadian tribunal ruled Wednesday, in what could be a landmark case for the use…

thehill.com

It seems fair and I can't believe Air Canada didn't just refund the money to begin with.

 

[Huntn]

Mispost

 

[dada_dave]

Large language models make for aggressive strategists, sometimes starting nuclear war even from neutral scenarios. Part of it is the randomness of the responses ratchets up escalation and part of it may be that they are trained on data that deals more with escalation than deescalation.

Oh and telling them that this was real and could result in their own power sources being destroyed had no effect whatsoever on their decisions.

And yes the movie Wargames was referenced several times in the video. Unlike the movie, these AIs did not learn that the best move is not to play.

 

[dada_dave]

A more prosaic piece on militaries kidding themselves and their governments that shortfalls in personnel will be solved in the short term by AIs or robotics.

Automation Does Not Lead to Leaner Land Forces - War on the Rocks

In June 2022, I travelled with a Ukrainian reconnaissance unit that was seeking to get an uncrewed aerial system with a camera over a series of Russian

warontherocks.com

 

[Huntn]

Large language models make for aggressive strategists, sometimes starting nuclear war even from neutral scenarios. Part of it is the randomness of the responses ratchets up escalation and part of it may be that they are trained on data that deals more with escalation than deescalation.

Oh and telling them that this was real and could result in their own power sources being destroyed had no effect whatsoever on their decisions.

And yes the movie Wargames was referenced several times in the video. Unlike the movie, these AIs did not learn that the best move is not to play.

There was a report/rumor about Israel using an AI program that offered up targets in Gaza.

 

[dada_dave]

There was a report/rumor about Israel using an AI program that offered up targets in Gaza.

Report not rumor:

The IDF's Decisive Victory Doctrine II

Bombing to Lose

bafriedman.substack.com

 

[GermanSuplex]

I like and am intrigued as much as anyone by AI, and see all the useful applications for it, plus there’s probably a ton I haven’t heard of and even more that nobody’s thought of yet.

That said, we aren’t out of the paradigm shift in society the internet has brought, and I see the potential AI has to bring another. I’m not sure how I feel about that, or if we need one. I’m relatively young and reminisce about my life before the internet - I’m lucky to have lived a whole 12 years before I had even heard of what the internet was, and I was grown before it became a constant instead of a novelty. I remember vividly my first time hearing of the word “Google”. I was an excite.com guy, if that shows my age. My reaction was a deadpan and honest “what the hell is Google?”

Social Media and the internet has shown us the dangers of technology outpacing our ability to properly weigh the pros and cons. Everything is so impersonal and anxiety and depression are quite high. I don’t see AI helping that, but then maybe that’s where the magic of AI will prove us wrong, or not.

An AI for a skilled doctor to reference for inspiration on helping out a patient or uncovering a puzzling diagnosis is great, but having the doctor as the robot and the AI as the doctor sounds like a problem. I don’t want a doctor who is just really good at reading and following directions.

How much is a degree to become a doctor these days? The cost of living, the cost of an education, and the rise of AI coupled with our learn-as-you-go approach to these influential technologies gives me concern.

But I have no control so I’ll enjoy the show.

 

[Herdfan]

That said, we aren’t out of the paradigm shift in society the internet has brought, and I see the potential AI has to bring another. I’m not sure how I feel about that, or if we need one. I’m relatively young and reminisce about my life before the internet - I’m lucky to have lived a whole 12 years before I had even heard of what the internet was,

I will show my age in that I was a Compuserve guy. I was in college and the amount of information available was almost overwhelming even if it was at 300 baud. Never did the AOL thing.

How much is a degree to become a doctor these days? The cost of living, the cost of an education, and the rise of AI coupled with our learn-as-you-go approach to these influential technologies gives me concern.

I'm not sure AI will be able to replace surgeons, but it sure can help with the diagnosis. Same with law, it won't be able to replace a trial attorney in the courtroom, but given Chat GP passed the MN? State Bar is both impressive and scary.

 

[Huntn]

I will show my age in that I was a Compuserve guy. I was in college and the amount of information available was almost overwhelming even if it was at 300 baud. Never did the AOL thing.




I'm not sure AI will be able to replace surgeons, but it sure can help with the diagnosis. Same with law, it won't be able to replace a trial attorney in the courtroom, but given Chat GP passed the MN? State Bar is both impressive and scary.

Speaking of…

ChatGPT has meltdown and starts sending alarming messages to users​

ChatGPT suddenly starts sending alarming messages to users

AI system has started speaking nonsense, talking Spanglish without prompting, and worrying users by suggesting it is in the room with them

www.independent.co.uk

ChatGPT appears to have broken, providing users with rambling responses of gibberish.
In recent hours, the artificial intelligence tool appears to be answering queries with long and nonsensical messages, talking Spanglish without prompting – as well as worrying users, by suggesting that it is in the room with them.

 

[Citysnaps]

 

[Herdfan]

Looks like there was an issue with Gemini:

Google pauses Gemini AI image generator after it created inaccurate historical pictures

Google on Thursday said it is pausing its Gemini artificial intelligence image generation feature, after saying it offered "inaccuracies" in historical pictures.

www.cnbc.com

 

[Citysnaps]

Also... Everyone knows Keith Richards is into Telecasters.

 

[Eric]

Also... Everyone knows Keith Richards was into Telecasters.

Richards is also older than Lincoln.