Yet again the USA making sure China will win the AI "war":
It's a long video and quite technical video about how the Deepseek team is innovating in the light of not having the latest GPUs etc.
... is long and you may not care or be interested in all the details of what happened with the Open AI-Hugging Face incident. But I recommend watching what the Open AI security and infrastructure has to say about the lessons learned here (starting at 30:25 in the video).
What he says is that now it has become known that AI agent orchestrated offensive attacks are totally possible. Outside threat actors can/will take advantage of this. AI agents, working in coordination on the offensive, on a huge scale, can move pretty fast.
Further development of AI intelligence will result in a dramatic increase of offensive capability for attackers.
Major problem is that there is no automated defense able to keep up with a potential wide scale AI agent-coordinated attack by a threat actor; why it is being suggested in the industry in the USA to start slowing the development of AI superintelligence before the threat grows larger.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
