Admin Blockchain Company
Follow
Recommended Watch: ‘Godfather of AI’ on AI “exceeding human intelligence” and it “trying to take over” (youtube.com)
The Nobel Prize winning ‘Godfather of AI’ speaks to Newsnight about the potential for AI “exceeding human intelligence” and it “trying to take over.”

Geoffrey Hinton, former Vice President of Google and sometimes referred to as the ‘Godfather of AI’, has recently won the 2024 Nobel Physics Prize. He resigned from Google in 2023, and has warned about the dangers of machines that could outsmart humans.

In May 2024, Faisal Islam spoke to the professor for Newsnight.

Newsnight is the BBC's flagship news and current affairs TV programme - with analysis, debate, exclusives, and robust interviews.

Website: https://www.bbc.co.uk/newsnight
X:  twitter_1x_v2.png / bbcnewsnight  
TikTok:  tiktok_1x.png / newsnightbbc  
Instagram:  instagram_1x.png / newsnightbbc  
Facebook:  facebook_1x.png / bbcnewsnight  
    • 1
    Francisco Gimeno - BC Analyst AI's existential threat exists, Geoffrey Hinton, the new Physics Nobel Prize or 2024 affirms. LLMs is just the beginning. The examples of what a LLM could do (and probably will be done) in terms of neural network is astounding. The next solution when AI goes beyond human intelligence? Experts go from nothing will happen, to the significant chance of AI to take control of the processes and even regulations, and we know with control, power comes, not necessarily responsibility. Hinton warns us to be prepared for what could happen, just in case (like when we have an AI specialised in defence and has autonomy to kill or lead a conflict). There are risks also in other sectors, like an AI go rogue and control something against our interests. Hinton is worried we don't have a global agreement to regulate to avoid bad consequences, just like we agree not to use chemical or biological weapons. But even the huge disruption in the work place and the economy (UBI), just left in hands of AI can be not a very good idea. For Hinton the period to confront the problem is from five to twenty years from now. We think probably it will be closer to us than those time periods. And there are not yet regulations or control or safety procedures in the companies competing for the AI market. They put more effort on profit than on safety. We live in interesting times.