Oracle’s Ellison wants all the world’s data and DNA in one place to train AI

If governments want Artificial Intelligence (AI) to improve services and security for their citizens, then they need to put all their information in one place – even citizens’ genomic data – according to Larry Ellison, the Oracle database tycoon whose companies, funnily enough, specialise in putting data all in one place. But he also has a point.

Ellison shared his take on what governments need to do to succeed with AI during a discussion with his buddy former UK prime minister Tony Blair at the World Governments Summit in Dubai.

The world's fourth-most-richest man – a good friend also of the world's richest man Elon Musk – insisted AI is soon going to change everyone's lives "across the board."

If governments want in, they’ll need to gather all their data – spatial information, economic data, electronic healthcare records including genomic data, and info about infrastructure. Whatever they’ve got, basically. And put it all in one place to be analysed by algorithms. The American multi-billionaire used the US as an example, if not a goal.

"I have to tell [the] AI model as much about my country as I can," Ellison said. "We need to unify all the national data, put it into a database where it's easily consumable by the AI model, and then ask whatever question you like," he said. "That's the missing link."

He believes the payoff will include better healthcare, thanks to treatments tailored to individuals, and the ability for governments to lift food production by better predicting crop yields. Analysing land so that farmers can be advised where to apply fertilizers or increase irrigation was another scenario Ellison floated.

"As long as countries will put their data - all of it - in a single place we can use AI to help manage the care of all of the patients and the population at large," Ellison said, adding his belief that AI can handle other social services and eliminate fraud.

Of course, such a vast database system could also be the precursor to pervasive surveillance – an idea Ellison last year said he feels is desirable and would like Oracle to help facilitate.

Constant real-time surveillance of populations, analysed by Oracle-powered machine learning products, would keep everyone "on their best behavior," Ellison said at an Oracle financial analyst conference in September 2024. We're reminded of the NSA, PRISM, Snowden et cetera.

Ellison is not just a techno-optimist. He’s also a top executive and shareholder who has made big AI investments as well as a database company to feed.

He therefore told the Dubai audience that Oracle, already a big-time government and military contractor, is ready to help nations realise his above-mentioned AI visions – to put all this data into one big expensive Oracle system to learn from and process.

"Oracle is building a 2.2GW datacenter that costs between $50 and $100 billion dollars to build," Ellison said, noting it's sites like that where super-powered AI models will be trained. "Because these models are so expensive, you won't build your own as a rule. There'll be a handful of these models."

And a handful of players that can train them with Oracle’s own Zettascale AI Factory facilities likely being one. The super-corp has also joined another, the Stargate project, that plans to blow $500 billion on AI infrastructure in the US in the next four years.