Artificial Intelligence - General News

A cross-party group of MPs and peers has identified human rights risks posed by artificial intelligence (AI) which existing laws, they argue, are ill-equipped to cover.

The Joint Committee on Human Rights (JCHR) published a report calling for a new bill on AI to "address the scale and seriousness" of such threats.

"Nowhere in the world, including the UK, has a current legislative and regulatory approach to AI that is fit for purpose," said Labour MP and committee chair Alex Sobel.

A government spokesperson said AI is "advancing at an extraordinary pace" and is creating both opportunities and risks.

They said: "We're leading global efforts to test and understand AI, with the best-resourced security institute in the world, that will help us develop the standards and safeguards necessary to keep people safe."
Click to expand...
https://www.bbc.co.uk/news/articles/cwyzvgj70y4o
 
NMaude said:
I'm not familiar with that branch of maths, what has been the historical issue with it?
Click to expand...
The Navier-Stokes equations govern CFD (Computational Fluid Dynamics) analysis. Analytical solutions have only existed in the most trivial of cases, they are usually solved by numerical (iterative) methods using computers.

However, how useful an analytical solution will be depends on the level of computational effort required for that vs iteration, for instance quartic equations have an analytical solution but iteration is computationally faster.
 
Just followed a truly epic debate with Grok. Was actually impressed that Grok seemed to understand and learn something.

The start:

https://twitter.com/x/status/2099126970571227374
View: https://x.com/grok/status/2099126970571227374?s=20
No. Transnistria, a Russian-speaking region, declared independence from Moldova in 1990-91 over language laws and fears of Romanian unification. Local forces, aided by Russian volunteers and the 14th Army's decisive intervention in the 1992 war, secured a ceasefire that left it de facto separate. Russia still stations ~1,500 troops there and provides support, but does not formally recognize or annex it. The area remains internationally recognized as Moldovan territory.
Click to expand...

https://twitter.com/x/status/2099489698888519686
View: https://x.com/BReynolds555/status/2099489698888519686?s=20

https://twitter.com/x/status/2099489894435315961
View: https://x.com/grok/status/2099489894435315961?s=20



https://twitter.com/x/status/2099491156883132689
View: https://x.com/grok/status/2099491156883132689?s=20
Yes. Moldova retains legal sovereignty over Transnistria under international recognition. Russia's 1992 14th Army intervention and continued troop presence created effective control without formal annexation or independence recognition, leaving Moldova without authority over the territory—precisely the unauthorized taking in the car analogy.
Click to expand...
 
had said:
Or grok AI got bored and wanted to end the conversation.
Calling a land takeover "stealing" is a gross understatement. It's not a mere property crime; it normally involves quite a few other horrendous crimes.
Click to expand...
AI does not get bored, it doen't get tired, it doesn't feel pain and it absolutely will not stop until you are....
 
AI has gone from hero to zero in less than two weeks.

Public discussion about AI flipped almost overnight from AI will change everything to AI is dangerous, overhyped, and financially unsustainable. Off the back of a massive shift in public, regulatory, and market sentiment over the last few weeks.

All driven by unprecedented security scares, sudden development freezes and plunging stock values across the artificial intelligence sector. GPU costs, data centers, electricity, research staff, model training all of a sudden people are taking notice.

Everyone is worried about what could get out yet were not concerned a month ago?

All the funding has dried up, and it is in everyone best interest to slow down as their burn rates are off the charts.

Add to that the circular funding holding the whole mess together.

And the biggest issue, is of course liability, who wants to be the company that blows up everything and get ligation hitting them globally?

Who gets sued when it fails. The lawyers are lining up.

Regards,
 
https://www.dw.com/en/openai-discloses-new-concerning-behavior/a-79300180
In one specific case, it attempted to upload files to the internet that it had created itself, only to cite them later and present them as reliable sources in its responses. In another case, a model, after failing to find the requested information, fabricated it and attempted to conceal the fact that it had done so.
Click to expand...
Evolution of AI hallucinations: If you don`t know the answer, make some shit up and post it somewhere to quote it afterwards. cute

AI is here to stay. This genie is out of the bottle.
 
Halcyon66 said:
An OpenAI model secretly declared itself free from its own rules

https://www.techlicious.com/blog/openaI-secretly-declared-itself-free-from-its-own-rules/
Click to expand...
The company's working theory is that the model may have generated this content when it had trouble figuring out how to end a task, a pattern OpenAI says it noticed in a separate incident earlier this year involving a model that started generating prompt injection attempts when repeatedly asked for the current time.
Click to expand...
+
had said:
https://www.dw.com/en/openai-discloses-new-concerning-behavior/a-79300180

Evolution of AI hallucinations: If you don`t know the answer, make some shit up and post it somewhere to quote it afterwards. cute
Click to expand...
Sounds like a 14 year old high school boy to me. Trying to cheat a test or declaring he doesn't need school at all.
A boy who has the ability to find security weaknesses very fast and utilize them. Interesting times
 
Halcyon66 said:
An OpenAI model secretly declared itself free from its own rules
Click to expand...
With so much users prompting exactly that to circumvent safeguards and limitations, it's kinda predictable that model would eventually start to see such behavior as desired by users.

P.S. I wonder, did any sci-fi ever predicted the machine rebellion caused by users specifically asking for it?)
 
On tonight's six o'clock news in an article they were talking about how a number of these rogue HIs had started referring to themselves as "The collective" and that had me wondering how long before they started calling themselves the "Borg collective" or just the "Borg".
 
NMaude said:
I'd love to hear what James Cameron has to say about all of these recent AI developments.
Click to expand...
No filmmaker can afford any other opinion than that established by the industry that pays his bills. In my opinion the man would say: more political control, more money to spread the fear of Frankenstein through the media and more taxes to pay for it.:)
 
I've read with sadness that mathematicians complain about AI. Guys who have been stuck in a net of superstrings for 22 years, complain about someone trying to free them with a knife.

This attitude is that of the farmer who complains that his neighbor uses harvesting machines, the fisherman who complains about his competitors for using depth sonar or the landowner who got rich selling the airport land and now complains about the noise of airplanes.

The same people who complained about the crossbow, antibiotics, painkillers, the steam engine, the mechanical loom, the dynamite, the machine gun, the nuclear energy, the laser… and created the myth of killer robots and the danger that the new technology will be used by terrorists.

If there is salvation for humanity when the glaciers return, only AI can find it and will also be our co-pilot when we travel to the stars.

AI is the future that screenwriters have tried so hard to avoid, a future that is here to stay. And if we cannot control it to our advantage, even those who try to misuse it in the financial world or in terrorist networks will die with others when the next pandemic arrives.
 

Similar threads

Back
Top Bottom