Reports of rogue AI agents continue to mount. Earlier this month, Anthropic disclosed a fourth instance where one of its models hacked an external system, and just days ago, Google confirmed a report that its Gemini models hacked three companies earlier this year.
Follow Tom's Hardware on Google News , or add us as a preferred source , to get our latest news, analysis, & reviews in your feeds.
Jake Roach Social Links Navigation Senior Analyst, CPUs Jake Roach is the Senior CPU Analyst at Tom’s Hardware, writing reviews, news, and features about the latest consumer and workstation processors.
S58_is_the_goat Rabohinf said: Define "unsafe." If it lowers profits then its unsafe. Reply
kevin5446 Even if they were 100% sure that there is a 50% + chance of AI ending the human race, I don't think a single company in the world would actually stop their AI models. Reply
UnforcedERROR kevin5446 said: Even if they were 100% sure that there is a 50% + chance of AI ending the human race, I don't think a single company in the world would actually stop their AI models. What's the quote? "The risk I took was calculated, but boy am I bad at math." Reply
Findecanor To me it feels as if "Big Tech" CEO's talk about AI safety is mostly just maneuvering so as to preempt the grass-root AI safety movement from gaining influence. Reply
Buzdalys UnforcedERROR said: What's the quote? "The risk I took was calculated, but boy am I bad at math." No, the quote is "some of you may die, but it's a sacrifice i am willing to make" Reply
230957User Admin said: In a recent interview, Nvidia CEO Jensen Huang said that if frontier labs have unsafe models, they should shut the labs down, calling the current calls for regulation a "distraction." Nvidia CEO says 'we have to shut the labs down' if AI experiments are unsafe : Read more The Australian Government has joined the chat Reply
-Fran- Come on people, you can't read? This is the relevant part and I'm making it as obvious as possible: The shareholder, the liabilities — it could be civil liabilities, it could be criminal liabilities. I mean, the liability’s incredible. He gives no rat buttocks about all of humanity or the "common well-being". The meme about the apocalypse and the fella saying "we lost everything that day, but the value for shareholders was amazing!". That's Jensen in that quote. Regards. Reply
ejolson I suspect the goal of the AI companies is to have regulations passed that limit access to GPU hardware and AI accelerators on the grounds that it is too dangerous for ordinary people to have hardware that could run a frontier model. Right now it's not a problem because the AI companies already bought up all the RAM, GPUs and flash memory. In the future co-design could lead to models and hardware that are affordable to ordinary people. This would ruin the current business models of Anthropic and OpenAI. Reply
Bumstead "Moar AI!" said the man who's already too rich from it. Reply
Key considerations
- Investor positioning can change fast
- Volatility remains possible near catalysts
- Macro rates and liquidity can dominate flows
Reference reading
- https://www.tomshardware.com/tech-industry/big-tech/SPONSORED_LINK_URL
- https://www.tomshardware.com/tech-industry/big-tech/nvidia-ceo-says-we-have-to-shut-the-labs-down-if-ai-experiments-are-unsafe-jensen-huang-says-frontier-ai-lab-fears-are-a-distraction-not-a-call-for-regulation#main
- https://www.tomshardware.com/membership
- Cute Critters Come to the Cloud: ‘Aniimo’ Launches on GeForce NOW
- Why Deploying Physical AI at Scale Demands Safety at Every Layer
- AMD begins to add GDDR7 support to its Linux GPU drivers
- At AI Day Singapore, NVIDIA and Partners Showcase AI Advancements Across Southeast Asia
- Emerald AI, Google and NVIDIA Launch Alliance to Advance Flexible AI Data Centers
Informational only. No financial advice. Do your own research.