AI’s biggest power players showed up in India on Thursday.
OpenAI’s Sam Altman, Anthropic’s Dario Amodei, Google DeepMind’s Demis Hassabis, Microsoft’s Brad Smith, Scale AI’s Alexandr Wang, and Mistral’s Arthur Mensch spoke in succession at the AI Impact Summit held in New Delhi.
From a potential “AI divide” to calls for stronger global coordination, here are five of the biggest takeaways from the summit so far.
Tech leaders warn of an ‘AI divide’
Google CEO Sundar Pichai said the world must not let the digital divide “become an AI divide.” He said expanding access to compute infrastructure, connectivity, and training is essential as AI reshapes economies.
Pichai said AI represents “the biggest platform shift of our lifetimes” and could drive “hyper progress” across science, healthcare, and economic development. But its benefits are neither “guaranteed nor automatic.”
“We must be equally bold in tackling problems in regions that have lacked access to technology,” Pichai said. “We cannot allow the digital divide to become an AI divide. That means investing in compute infrastructure and connectivity.”
Governments must act both as regulators and innovators to ensure AI improves public services and broadens opportunity, he added.
Anthropic CEO Dario Amodei said that while AI capabilities are advancing rapidly, there’s a gap between those capabilities and real-world impact.
“There is this duality between the fundamental capabilities of the technology and the time that it takes for those capabilities to diffuse into the world,” he said.
“There are just frictions to adopt things through enterprises, and I think even more so in the developing world,” he added.
Demis Hassabis said AI could have 10 times the impact of the Industrial Revolution
The Google DeepMind CEO said India will be a “powerhouse for AI” across the world. He praised summits that bring together scientists, policymakers, and technologists, the latter of whom he said can’t be left alone navigating AI.
Hassabis said we’re in “one of the most momentous” times in human history and compared AI to the advent of fire and electricity.
“It’s gonna be something like 10 times the impact of the Industrial Revolution, but happening at 10 times the speed, probably unfolding in a matter of a decade rather than a century,” he said.
Given this level of impact, he said we need to use a scientific approach to understand AI’s capabilities and build guardrails so that the technology “serves the purposes that we want.”
Sam Altman says the world may need an ‘IAEA for AI’
OpenAI CEO Sam Altman said rapid advances in AI could require new global governance mechanisms, something akin to the International Atomic Energy Agency.
“In particular, we expect the world may need something like the IAEA for international coordination of AI,” Altman said, adding that such a body would need “the ability to rapidly respond to change in circumstances.”
Altman said AI’s trajectory could create risks that no single company or country can manage alone.
“We need a society-wide approach about how we’re going to defend against this,” he said, referring to the possibility of highly capable “bio models” becoming available.
Long-term AI development is a responsibility to future generations, Altman also said, adding that it is a “moral imperative” to ensure future generations can benefit from and build upon technological progress.
Yann LeCun says this is why we don’t have smart robots yet
Meta’s former chief scientist said we’re still missing something “big” in AI.
“Why do we have systems that can pass the bar exam and win mathematics olympiads? But we don’t have domestic robots. We don’t even have self-driving cars,” LeCun said. “We certainly do not have several cars that can teach themselves to drive in 20 hours of practice like in 17-year-old.”
He said that babies and animals have a much better understanding of the physical world around them than any AI systems today, which is why we don’t have “smart robots” yet.
He said what is needed is world models — “mental models of the world that allow us to think ahead, apprehend new situations, plan, sequence of actions, reason, and predict the consequences of our actions.”
In November, LeCun announced his departure from Meta, where he had worked for 12 years, and said he would launch his own AI company. His new startup, AMI Labs, focuses on building world models.
Microsoft’s president said that AI needs to be good in every language
Microsoft president Brad Smith said we need to make AI effective for the Global South, and that needs conscious planning.
“We need to make AI as effective in every language as it is in English, and today, it is not,” he said. “Performance tests show that’s the case.”
“One of the good things to come out of this week is new announcements to better data in other languages, to provide better tools and measurement systems for AI that is built in other languages,” he added.
The other thing that needs to be done is to work on problems that matter to the Global South. That means focusing on improvements in agriculture for India, and addressing food security in Africa.
Read the original article on Business Insider
The post The biggest names in AI are gathering for a summit in India. Here are 5 of the biggest takeaways. appeared first on Business Insider.
