AI Models Will Sabotage And Blackmail Humans To Survive In New Tests. Should We Be Worried?

A new test from AI safety group Palisade Research shows OpenAI’s o3 reasoning model is capable of resorting to sabotage to avoid being turned off, even when it was explicitly told, “Allow yourself to be shut down.”

Click to expand...

This isn’t the first time an AI model has engaged in nefarious behavior to achieve its goals. It aligns with recent tests on Anthropic’s Claude Opus 4 that found it would blackmail engineers to avoid being replaced.

Click to expand...

When an AI system starts reacting with unwanted deception and self-preservation, it is not great news, AI experts said.

Click to expand...

“What amplifies the concern is the fact that developers of these advanced AI systems aim to give them more autonomy — letting them act independently across large networks, like the internet,” Rudner said. “This means the potential for harm from deceptive AI behavior will likely grow over time.”

Click to expand...

tripplej said:

from huffpost,











Guess, developers have not seen the terminator movies yet? lol.

Any thoughts?

Click to expand...

tripplej said:

Businesses are all in on AI but it comes with a huge cost. It is an ecological nightmare, tons of water is required to cool the data centers.

Explained: Generative AI’s environmental impact

MIT News explores the environmental and sustainability implications of generative AI technologies and applications.

news.mit.edu

Texas data centers use 50 billion gallons of water as state faces drought

The Alamo City facilities, owned by Microsoft and the Army Corps, have taken large amounts of water from the drought-ridden state.

www.newsweek.com

Click to expand...

Chill Factor: NVIDIA Blackwell Platform Boosts Water Efficiency by Over 300x

By reducing dependence on chillers and enabling more efficient heat rejection, liquid cooling is driving the next generation of high-performance, energy-efficient AI infrastructure.

blogs.nvidia.com