Regulations too soon could stifle innovation, Tanuj Bhojwani, the head of People+AI argued during the AI Alliance Delhi NCR conference on May 17. “I mean, I don’t want to pick on anybody, but like, look at Europe, right? Like, I think people who were looking to build there [are] actually looking to get out,” he pointed out. He suggested that the government should generate demand by committing to buy AI solutions that address specific problems. “That signals private capital to come in and say yeah, you know what, we’ll go for it. Because the government is a large customer and [a] perennial customer,” he explained.
Bhojwani said that the conversation around regulations needs to be slowed down. To discuss regulation, one first needs to understand what will happen. “Of course, stuff like deepfakes, etcetera, you know, is immediate. But for the most other things that we worry about, like we were worrying about bioweapons manufactured with AI like sometime last year. I think it’s forgotten. Nobody cares anymore,” he argued emphasizing that reacting too soon can be a problem.
Other key points made by People+AI:
Should India focus on LLM creation or on developing AI use cases?
During the discussion, the moderator pointed out that at People+AI’s event in Bangalore, Nandan Nilekani (the co-founder of the EkStep Foundation of which People+AI is a part) said that India is not in the arms race to build large language models (LLMs). Rather, India would become the AI use case capital of the world. Commenting on this push for becoming the use-case capital, Bhojwani explained that the idea behind Nilekani’s statement was perhaps not to do one or the other but instead, to pick a focus between developing LLMs and use cases.
“And the point is, look, distribution has always been the game. A lot of people today say that, look, Google, Microsoft is the first new technology where, or at least one that I know of, where the incumbents are not necessarily getting disrupted, but instead getting cemented because they have the distribution,” Bhojwani explained. He added that if India focuses on the use cases and the applications for now, and gets people used to using AI, “the supply chain [of creating Indian LLMs], from that point on, we can start getting at later.”
On its Open Cloud Compute (OCC) Project:
Described as a digital public infrastructure (DPI) for artificial intelligence (AI), OCC is trying to build a grid powered by protocols and standards that will allow various users to log onto an interface and find compute that best fits their geographical and business needs. “Now, when we say connect diverse compute, it means enable not one provider or one cloud service provider. It means enabling many cloud service providers who exist today,” Tanvi Lal, strategy and operations, People+AI, explained.
Lal mentioned that for tech startups, the decision to look for cheaper compute comes in when their businesses start to scale up. “Our entire thesis is, why don’t we make a more informed decision sooner? Why don’t we give all the information that needs to be with customers sooner by actually transparently saying, this is the kind of compute that’s available and this is how you get it,” she said. She mentioned that People+AI has signed on 24 technology partners (including both consumers and producers of compute) who have shown interest in co-developing the OCC grid.
Also read:
STAY ON TOP OF TECH NEWS: Our daily newsletter with the top story of the day from MediaNama, delivered to your inbox before 9 AM. Click here to sign up today!
Source Link
