(Washington) Amazon, the world’s largest cloud provider, is working to persuade utilities and non-governmental organizations to join the generative artificial intelligence (AI) revolution as the race for this cutting-edge technology heats up. intensifies with Microsoft and Google.
AWS, the cloud computing arm of the American e-commerce giant, is convinced, like its rivals, that generative AI (at work in ChatGPT) can profoundly improve areas such as health and security.
To allow interested institutions and associations to test their ideas, the company announced on Wednesday an investment of $50 million over two years.
“I’m seeing a lot of ideas, applications, projects, things that I think will really have an impact,” Dave Levy, an AWS vice president responsible for global public sector customers, told AFP.
“Public organizations really need support and help to get their ideas off the ground,” he assured before an AWS “summit” on this subject in Washington.
Generative AI, popularized by ChatGPT (OpenAI), makes it possible to produce all kinds of high-quality content with a simple query in everyday language, from mountains of data. However, there is no shortage of data and public services.
“The power of generative AI is to bring the data together in one place, and then people get creative to figure out what can be done with it,” says Levy.
The race to develop and deploy generative AI tools is closely linked to the cloud market, because this technology, like video streaming and Internet platforms, requires servers and remote computing services to function.
To remain the leader in the cloud, AWS must therefore also distinguish itself in this new AI, while Microsoft (main investor in OpenAI) and Google have taken the lead in the digital revolution.
The global government cloud market size is estimated at $35.5 billion in 2024 and is expected to more than double over the next five years, according to Mordor Intelligence, a market research firm.
While the United States accounts for the largest share of the pie, Europe’s share is growing rapidly too, and companies hope that governments around the world will increasingly rely on the cloud rather than using in-house servers. often obsolete.
But expansion is hampered by national laws that sometimes require data to remain sovereign and backed up locally, forcing U.S. giants to build secure data centers around the world.
“We have infrastructure all over the world […] and we respect all the laws imposed on us,” stressed Mr. Levy.
AWS is betting that generative AI could spark a shift to the cloud when public agencies see new tools playing a more important role in accomplishing their mission.
With the initiative presented on Wednesday, Amazon hopes to give public actors an easy way to start using generative AI despite their possible reluctance.
Because despite the immense expectations that the new technology arouses, it is also known for its “hallucinations”, when the machine, programmed to compose sentences or create images without understanding their meaning, “invents” crazy answers.
Governments and institutions generally prefer traditional, more predictable IT.
AWS claims to differentiate itself from its competitors by giving top priority to security and ensuring in advance that its AI tools are ready for deployment.
Lagging behind Microsoft and Google, in October it launched the Bedrock platform, which allows its clients to use different generative AI models, such as Claude (Anthropic), Llama (Meta), those of Mistral and Titan, an Amazon brand.
In one example cited by AWS, the Dana-Farber Cancer Institute in Boston built a new research solution using the Claude model to help doctors interpret certain laboratory results that may be particularly complicated to understand.
In the United Kingdom, Swindon Borough Council used the platform to build a generative AI tool to make complex rental contracts more understandable for everyone.
