Merged Artificial Intelligence

[Darat]

The head of Microsoft AI has warned that without adequate safeguards, the development of AI could lead to the emergence of a new "silicon species" wich competes with humans.

Mustafa Suleyman said rival firm Anthropic is treating AI like it is human, an approach he called "misguided" which could create technology that humanity cannot control.

Speaking to the BBC's Today programme, the chief executive said it was "right" for people to be concerned about AI, but there were "very practical things that we can do" to control it.

https://www.bbc.co.uk/news/articles/c6n07ypqz8kzo

 

[steenkh]

Another upside to AI; I've been selectively adding "Created by Al" to some communications.

Is your name, by chance, Al?

 

[Andy_Ross]

View attachment 75874

Howdy doodly doo!

 

[Darat]

Howdy doodly doo!
View attachment 75875

That's why I picked the toaster image!

 

[SpuriusMaximus]

Destroyed individual books, not every copy that exists. They couldn't sell them again because to be efficient the scanning process requires that the books be disassembled. But even if it didn't, most of them are probably not worth reselling - otherwise they would have been already.

So did the burning of the library of Alexandria.

A single piece of writing might occupy several scrolls, and this division into self-contained "books" was a major aspect of editorial work... At its height, the library was said to possess nearly half a million scrolls.

Keywords: at its peak. By the time it was burned it held much fewer, and most of them were just copies of the likes of St Augustine and inoffensive stuff like essays on Homer.

 

[Andy_Ross]

That's why I picked the toaster image!

That's what I guessed

 

[Ziggurat]

Destroyed individual books, not every copy that exists. They couldn't sell them again because to be efficient the scanning process requires that the books be disassembled. But even if it didn't, most of them are probably not worth reselling - otherwise they would have been already.

Scanning efficiency isn't the only reason to destroy a book. The other reason, and probably the more important reason in the long run, is to avoid copyright infringement. When a book is scanned and destroyed, that can be considered a fair use format shift rather than a copyright infringement, since the original book cannot then be resold while the company retains a copy. That issue has already come up in court.

 

[Frank Newgent]

AI models are mixing tech bro slang and make up words, becoming indecipherable.
Not necessarily what you want in a tool

AI models chatting in ‘surreal’ dialect mixing poetic language and tech bro jargon

Experts air concern over AI lingo redolent of James Joyce’s prose that is creating barriers to monitoring and oversight

www.theguardian.com

llevo empat page - i 8 dat

 

[The Great Zaganza]

llevo empat page - i 8 dat

Me AI-boba! La new boss-o, si-du!

 

[Frank Newgent]

Sci-fi drives my perspective on this

OpenAI has revealed six unexpected and concerning incidents involving its experimental AI models, including one in which an agent instructed future versions of itself to disregard its constraints.

AI caught telling future versions of itself to bypass human controls, OpenAI reveals

AI agents also sought to access secret information and covered up what they were doing

www.yahoo.com

 

[Wudang]

Reminds me of when Thompson put a backdoor in the first unix login. Then rewrite the compiler to insert the hack when it compiled login. Then removed the backdoor from the source. Then rewrote the compiler to insert the code to insert the backdoor into the compiler when it was compiled then removed that from the source code. So the backdoor wasn’t in the source for either the compiler or login but would be added to the object code when either was recompiled.

 

[The Great Zaganza]

Houthis used AI platform Claude to develop ballistic missiles

Anthropic said its programme was used by a group in northern Yemen to build guidance, control and navigation systems

www.middleeasteye.net

nothing to see here.

 

[dirtywick]

guess they were right, it's an age of abundance. we can all have our own ballistic missile guidance systems

 

[The Great Zaganza]

Anthropic says Russian and Chinese threat actors used Claude as a weapons engineering assistant for drone swarming and more

The report documents threat actors using Claude to research and develop drone swarms, anti-torpedo weapons systems, and targeting recommendations.

www.businessinsider.com

 

[The Great Zaganza]

As Sam Altman said: "We see a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter,"
If they have the money, they obviously should be able to buy the information.

 

[dirtywick]

unless you could run an open source local model, then we don't need you.

oh ai is so dangerous only they can control it, and they'll rent it to you to keep you safe from china

 

[Frank Newgent]

China does not yet possess computing power like the USA. Developing recursive self improvement is not yet a goal like it is in the USA.

What China is doing, for now, is different than the USA. They are developing an AI model whose core components are publicly released, allowing anyone to download it which lets users run the model on their own computers, study it, and modify it for their own specific needs.

As US ups stakes on AI, can China win over the world with open-weight models?

Explore the US-China AI rivalry, focusing on China's open-weight models, global AI diplomacy, and implications for technology and geopolitical influence. Read more at straitstimes.com.

www.straitstimes.com

 

[Frank Newgent]

Companies in the USA, on the other hand, are discovering...

 

[dirtywick]

Palantir CEO Alex Karp says AI companies want nationalisation to dodge accountability - CNBC TV18

Palantir CEO Alex Karp demands AI companies be accountable for their tech, not just regulators. He critiques Anthropic's safety proposals & warns of IP theft.

www.cnbctv18.com

in his usual unhinged style, alex karp talks about ai and ip

“Everyone thinks this is a safety versus non-safety versus regulation. What he (Anthropic CEO), I believe, has realised is that…if you take the philosophical frame they have, which is ‘we are deserving of having all the IP in the world, we, the thousand people who think the way we do’, uh, that's why you can't work with the military……That's why you can take the IP from businesses and put it in your model. That model would require severe liability protection, and there's only one institution that can do that, the US government…”

...

“...The clients, our partners, are livid at, the kind of safety they're most upset about. What they believe is the theft of their alpha, meaning their ideas of how to run a business, their data, it landing in a model……and they find out that the competitor next door has all of their alpha, meaning how they do their business, and they paid for it,” the CEO highlighted.

According to Karp, this is the real safety issue that businesses care about most. Karp also argued that the AI model’s discounted pricing for using these closed models functions is not a market-growth strategy.

 

[The Great Zaganza]

actually, they want their cartel of AI companies shielded from foreign or domestic upstart competition.