Artificial Intelligence (AI) has become a huge topic of debate for everyone across the industry regarding its potential threats and regulation. Romesh Wadhwani, Chairman, Symphony AI, believes that AI is more difficult to regulate than a nuclear power plant. Wadhwani spoke to Tech Today Editor, Aayush Ailawadi about Artificial how AI is far more dangerous than a nuclear power plant.
Talking about the recent 22-word warning by top AI experts including OpenAI CEO Sam Altman, ‘Godfather of AI’ Geoffrey Hinton and more that says that threats of AI can cause extinction as great as pandemic or nuclear war. Wadhani believes that they are right to sound alarm across the world. He stated: “I think their vision has to be equal to the challenge. And the way people are talking about regulation is still old fashioned as they are thinking of it as it's a social network. Failure to regulate social networks is something that they perhaps want to catch up in this world of generative AI. But the harm that's been done by social networks is a miniscule fraction of what can happen if this is uncontrolled.”
He said further that if proper guardrails are not put in place for AI, it will be worse than Chernobyl. He explains, “To create a nuclear plant in the first place that has the potential risk of a future melt requires an investment of 5 billion or $10 billion to create. Dangerous AI requires an investment of $10,000 $100,000 $1 million.”
“The other factor that people are not paying attention to is this whole area of open sourcing of AI technology. You can't open source nuclear secrets. People may steal them, but it requires a huge amount of effort to steal them. With AI, every actor in the world --good actors, bad actors, all have access to the same large language model. And because of the way language models can be operated, you don't need to invest $500 million or $5 billion to take advantage of them. You can do it with much less expensive hardware technology," he added.
He further pointed out that China has been quick on its feet to get the situation in control and even made a roadmap to make the best out of it. He says, “China is building its own regulatory frameworks and they have the ability to enforce those frameworks better than most other countries do. The US currently does not have a meaningful regulatory framework. Although, EU is in the process of developing one but I feel like they are not approaching it with the level of rigor that's actually needed," he explained.
On being asked about the role India is playing in the AI revolution, Wadhwani noted that it is not only India's great opportunity, but actually it is India's responsibility to be one of the top three players in AI. He states, “To become one of the top leaders in AI, India is going to require massive national scale initiatives for the application of AI ministry. And just as digital transformation, has played a big role in India's progress over the last few years with Aadhaar and UPI. There is a need for an AI transformation of many programs that the Indian government has or could plan," Wadhwani said.
Also Read:
‘This call is from Delhi Police’: Scammers tricking people to steal Aadhaar, ATM, PAN details
Nothing Phone (2) confirmed to be ‘Made in India’: Check launch details, specs, price