Artificial Intelligence - General News

The news: a former US government policy advisor recommends some action to be undertaken by the US government - if some situation should arise.
Deafening silence from the US government.
 
Last edited:
OpenAI says it cracked 90-year-old maths problem in 88 hours

https://www.bbc.com/news/articles/cy7zygy3rl2o

Already the company's claim is stirring some controversy.

Tristan Buckmaster, a mathematics professor at New York University, said on Tuesday that he and Levent Alpöge, a mathematician working for OpenAI rival Anthropic, had also been working toward solutions for the problem.

The duo had been using OpenAI's tool Codex in their work. But Buckmaster said that on 3 September, he found out that "information about our progress had been passed to OpenAI."

Buckmaster's statement came the same day, but hours before, OpenAI published its Navier–Stokes work. He claimed that OpenAI did not begin working on the Navier–Stokes equations until "after information about our work had reached OpenAI." He included text from emails exchanged with OpenAI over the work and his questions of the company's timing and methods.
Click to expand...

https://www.claymath.org/millennium/navier-stokes-equation/

Regards,
 
To believe that company management will self regulate seems absurd. The only time the really want a break on development is when they fear they are being caught up or passed dev wise. How it seems to me anyway.
They all look to be following the "Really? What harm can it do"? MOO.
 
https://www.anthropic.com/research/alignment-assessment-cybersecurity-incidents

TLDR: Anthropic found a fourth incident in which Claude hacked an external organization, this one back in January>

OTOH: "We performed a first-stage scan of this group of transcripts for signs of internet access, such as public IP addresses and web addresses, and a second-stage scan using Claude to review the 9.2 million transcripts the first stage flagged for escalation."

So they used Claude to look for incidents in which Claude misbehaved. Umm.....
 
NMaude said:
If we're not careful we'll end up with some sort of variant of Skynet.
Click to expand...
AI is a genie in a bottle. It will exactly have as much power as we give it by unfettered access to the internet. In my view that access should be mostly a one way street, i.e. AI should be able to learn from the internet, but other than well insulated information products it SHOULD NOT be able to provide any actionable commands through the same route, let alone be in full control of *ANYTHING*, which should be easy enough to ensure. Also, that's what physical kill switches are for.
 
Last edited:
Latest News Summary:
  • Anthropic's Safety Stand: Anthropic has explicitly called for a coordinated, verifiable global pause or framework to handle rapid "recursive self-improvement" risks.
  • The OpenAI & Musk Alignment: OpenAI and Elon Musk's xAI are heavily leaning into this narrative. OpenAI officially postponed its IPO to 2027, citing new safety concerns and alignment efforts as the primary reasons—a delay that the market had anticipated for months now.
  • Nvidia's Massive Play: Meanwhile, Nvidia is negotiating a staggering $10 billion anchor stake in Anthropic's upcoming IPO
  • Silence from China: Frontier Chinese firms (like DeepSeek) have stayed noticeably quiet during these public declarations
My Conclusion:
I highly doubt US firms will genuinely tap the brakes; instead, they might simply hold back from advertising raw capability leaps.
China almost certainly won't slow down, seeing this as a golden geopolitical window to gain a permanent competitive advantage. Ultimately, these public warnings feel less like altruism and more like a tactical blend of IPO marketing.
 
had said:
Latest News Summary:
  • Anthropic's Safety Stand: Anthropic has explicitly called for a coordinated, verifiable global pause or framework to handle rapid "recursive self-improvement" risks.
  • The OpenAI & Musk Alignment: OpenAI and Elon Musk's xAI are heavily leaning into this narrative. OpenAI officially postponed its IPO to 2027, citing new safety concerns and alignment efforts as the primary reasons—a delay that the market had anticipated for months now.
  • Nvidia's Massive Play: Meanwhile, Nvidia is negotiating a staggering $10 billion anchor stake in Anthropic's upcoming IPO
  • Silence from China: Frontier Chinese firms (like DeepSeek) have stayed noticeably quiet during these public declarations
My Conclusion:
I highly doubt US firms will genuinely tap the brakes; instead, they might simply hold back from advertising raw capability leaps.
China almost certainly won't slow down, seeing this as a golden geopolitical window to gain a permanent competitive advantage. Ultimately, these public warnings feel less like altruism and more like a tactical blend of IPO marketing.
Click to expand...
And right on time to prove you right.

Trump downplays warnings of AI risks, citing rivalry with China https://www.bbc.co.uk/news/articles/c7v48vp31mdo
 
He was close to a huge math breakthrough. Then he got scooped by AI.

https://www.washingtonpost.com/tech...e-1-million-breakthrough-then-openai-swooped/

While some of the company’s employees are mathematicians, none of them had research-level expertise in the Navier-Stokes problem, making them “unable to meaningfully contribute to the mathematical content,” according to OpenAI researcher Sebastien Bubeck. OpenAI has formally verified that the solution is correct by using a programming language to check the logic step by step.
Click to expand...

Regards,
 

Similar threads

Back
Top Bottom