Weekend Warp

For a change, India dominated the discourse in the week gone by.

The OG Rockstar of AI, Jensen Huang , was in India creating mob-like situations wherever he went, as also the more professorial Yann LeCun , the Chief of AI at Meta . Huang talked up AI (no surprise there), the need for Sovereign AI (none there too, 英伟达 is looking at markets beyond BigTech to keep up their scorching growth), and urged Indian business leaders and politicians to export ‘intelligence, rather than mere services’. He also bemoaned the lack of independent AI research labs in India, a sentiment I share. The consummate salesman that he is, he struck a big 1GW AI Data Centre deal with Mukesh Ambani’s Reliance Industries Limited , and smiled benignly as Tata Communications announced another one. His India trip was good for him, as Nvidia briefly became the world’s most valuable company again just pipping Apple. Nandan Nilekani, India’s own tech rock star, meanwhile poured cold water on India creating its own LLMs, saying that the “big boys in the Valley” should do that, while in India, we should build applications on top to “solve real world problems”.

He said this in a conversation with Yann LeCun, who also did the rounds in India, evangelizing the goodness of Open Source AI, that Meta has bet the farm on.

Read more about Jensen ‘Rockstar’ Huang’s and LeCun’s India trip.

Meanwhile, AI had its biggest moment on the world stage, when two The Nobel Prize went to AI researchers and scientists. It started off with John Hopfield and Geoffrey Hinton winning the Physics prize for their work in deep learning and associated fields. The very next day @David Baker, Demis Hassabis and John Jumper won the chemistry prize for applying deep learning and AI to solve the intractable problems around protein folding. Two of these names are the leading lights of AI – Hassabis who set up Google DeepMind , and Hinton, the ‘father of AI’ who is probably the only person to win both the Turing Award, the highest honour in AI, and the Nobel. While people did scratch their heads on the connection between physics and deep learning, the message was unambiguous. AI and its applications had reached the highest level of science, validating Demis Hassabis’ assertion when he set up DeepMind that the lab will win four to five Nobels.

I wrote about the import of this here

There is some trouble in the land of ‘big boys in the Valley’ – I am talking about Sam Altman’s OpenAI . Another safety researcher has left, ex-CTO Mira Murati continues to try poaching people to form her own startup, and there are rumours swirling around that OpenAI is trying to extricate itself from the tight embrace by its Big Brother 微软 . The story goes that their contract was valid until AGI had been achieved and OpenAI is trying to prove that it has, or something like that!. On the plus side of the balance sheet, super star Microsoft researcher Sebastien Bubeck ?(of the Sparks of AGI fame) has joined OpenAI. Also, the innovation factor continues to churn with the launch of another astonishing update to its core model – the Canvas feature, which could be the future of Microsoft Word.

The space is certainly heating up...... at warp speed.