Modern AI “Godfather” Yoshua Bengio Joins UK Safety Program

As artificial intelligence (AI) and generative AI-powered tools continue to grow in popularity and revolutionize industries, governments and AI experts are calling for stronger protective measures to ensure the safe usage of AI technology.

One such initiative taken by the UK Government’s Advanced Research and Invention Agency (ARIA) includes Safeguarded AI, which has just made an announcement regarding a prominent figure of the AI industry, Yoshua Bengio, that’s set to enhance its program.

So, what is ARIA’s Safeguard AI about and what did the agency say about Yoshua Bengio? Let’s explore!
 

What Is Safeguarded AI?

 

• According to its website, ARIA’s goal for its Safeguarded AI program is “to usher in a new era for AI safety, allowing us to unlock the full economic and social benefits of advanced AI systems while minimizing risks.”

• The program finds its purpose as AI is becoming more capable and current measures taken to mitigate risks of advanced AI systems are unreliable and limited.

• Furthermore, the agency feels there is a lack of R&D efforts conducted to provide quantitative safety guarantees for AI systems, as they’re believed to be impossible or impractical.

• \Backed by £59 million ($75 million) spread over four years, ARIA’s Safeguarded AI will combine scientific world models and mathematical proofs to construct a “gatekeeper” AI system that can understand other AI agents and reduce its risks.

• The program is split into three technical areas (TAs) to develop its intended AI system.

• TA1 is called Scaffolding and includes building an extendable, interoperable language and platform.

• TA2 is called Machine Learning and will use AI to develop best-in-class mathematical models of real-world complex dynamics and frontier AI to train autonomous systems.

• TA3 is called Applications and will include deploying a gatekeeper-safeguarded autonomous AI system to evaluate other AI systems deployed in critical areas.

• ARIA is also offering funding to organizations in high-risk sectors such as transport, telecommunications, supply chains and medical research, to help them build AI safety-focused applications.

• The first phase will consist of £5.4 million ($6.85 million), while the second is expected to provide another £8.2 million ($10.4 million).

• Furthermore, ARIA is looking for people who are interested in building Safeguarded AI’s safety mechanism through a nonprofit organization, which will be set up using £18 million ($22.8 million), for which ARIA will be accepting funding applications early next year.

 

What Did Safeguarded AI Announce About Yoshua Bengio?

 

• Through a blog post published on its website, ARIA announced that it was excited to welcome Professor Yoshua Bengio as Scientific Director for Safeguarded AI.

• Yoshua Bengio won the prestigious Turing Award (which is known as the “Nobel Prize of Computing”) in 2018 for his breakthroughs in artificial intelligence.

• Bengio is a world-renowned computer scientist and a pioneer in deep learning and is dubbed a “godfather” of modern AI.

• Bengio will join Program Director David ‘davidad’ Dalrymple in achieving Safeguarded AI’s goal.

• This will include working with the program’s team and its R&D Creators and providing scientific and strategic advice, focusing on TA3 and TA2.

• Through the blog post, ARIA said, “Yoshua’s work on the Cautious Scientist AI research agenda, which seeks to develop efficient methods for providing high-assurance guarantees that an AI system won’t take harmful actions, aligns closely with the Safeguarded AI program, through which we’re looking to construct a ‘gatekeeper’.”

 

What Did Yoshua Bengio Say?

 

• Yoshua Bengio, Scientific Director of Safeguarded AI, said, “We’re currently racing toward a fog behind which might be a precipice.”

• [Contd.] “We don’t know how far the precipice is, or if there even is one, so it might be years, decades, and we don’t know how serious it could be … We need to build up the tools to clear that fog and make sure we don’t cross into a precipice if there is one.” 

• “We need to bring the discussion of how we are going to address the risks of AI to a global, larger set of actors. This program is bringing us closer to this.”

Do you think other prominent figures in the AI industry should join this initiative? Do you think AI companies and governments should add to this initiative or create similar ones to work towards AI safety?

Let us know in the comments below!