Big tech companies are aggressively investing in and forming alliances with artificial intelligence start-ups through their cloud computing arms, raising regulatory questions over the role of both suppliers and competitors in the battle to develop “generative AI”.
Google’s recent $300 million bet on San Francisco-based Anthropic is the latest in a string of cloud-related partnerships between nascent AI groups and the world’s biggest technology companies.
Anthropic is part of a new wave of young companies developing generative AI systems, sophisticated computer programs that can parse and compose text and create art in seconds, which are being built in-house by big companies like Google and Amazon. Are.
The technology behind products including OpenAI’s ChatGPT, a chatbot that can interact with users via text, requires massive amounts of computing power – expensive infrastructure controlled by a single handful of tech giants.
,[This] That’s exactly the type of scenario the Federal Trade Commission has said they’re going to focus on,” said William Kovacic, former Republican chairman of the US Antitrust Agency and professor of antitrust law at George Washington University.
“There is growing concern about how large information services firms are limiting opportunities for a new generation of competitors to come forward,” he said, adding that they may be paying “too much attention” to these deals. . The FTC declined to comment.
These partnerships provide Cloud’s owners with insight into the talent and technology inside start-ups, while allowing smaller companies to sidestep the huge capital investments that would otherwise be required to build their own data infrastructure. AI start-ups that need to train models have no choice but to run into the arms of larger companies offering the required cloud computing at discounted rates and requiring huge amounts of capital.
“Clouds tend to lock-in, they force people into massive multi-year commitments,” said Jonathan Frankel, co-founder of MosaicML, an AI company that helps its corporate clients with AI model needs. trying to commoditize the cloud for
The two companies announced a separate cloud partnership after the Financial Times first reported that a Google-Anthropic investment gave the search giant a 10 percent stake in the company.
The arrangement echoes the $1bn cash-for-computing investment Microsoft made in OpenAI three years ago. In January, Microsoft announced a “multi-year, multi-billion dollar” investment in OpenAI estimated at $10bn.
The deal solidifies Microsoft’s position as the exclusive infrastructure provider for one of the world’s leading AI start-ups. Chief executive officer Satya Nadella claimed that Microsoft had built a supercomputer to handle OpenAI’s work, and it can now handle some AI calculations at half the cost of its competitors. Minimizing cost is important for compute-intensive development of large language models: it is estimated that the cost of running ChatGPT, assuming 10 million monthly users, is $1 million per day.
Meanwhile, Amazon’s most prominent alliance so far among AI start-ups is Sustainability AI, which in November announced AWS as its “preferred cloud partner” for building and training its media-generation models.
The partnership includes a commitment by Sustainability to use Amazon’s Trenium chips, custom-designed processors that rival Google’s tensor processing units. The deal gives Amazon, which some in the AI industry see as lagging behind Microsoft and Google in terms of AI capabilities, a major partner to showcase its cloud platform. The deal is not exclusive, according to a person familiar with the terms, freeing up Stability to potentially work with alternative cloud providers such as Google Cloud. Google also said that its cloud deal with Anthropic was non-exclusive.
However, according to AI researchers, building and deploying large language models with billions of parameters, such as GPT or Google’s PALM model, requires stable hardware, making it difficult to move between different platforms once the model has begun to train. It happens.
Historically, this type of dependency has attracted the attention of antitrust regulators in other sectors, including telecommunications, according to Kovacic. “The fact that your supplier of a vital service is also your competitor is an inherently awkward and tension-filled relationship.”
The fundamental need for a reliable cloud provider that can supply computing infrastructure at the volume and frequency a generative AI start-up needs means companies are increasingly seeking Big Tech cloud partnerships .
Google and Amazon have close ties to other well-funded AI start-ups building their own language models, including California-based Fog and the Israeli company AI21 Labs, whose co-founder Yoav Shoham The last two companies have been sold to Google.
Cloud management company Yellowdog, which helps customers switch between cloud services, says it is aware of several alliances between nascent AI companies that have not yet launched products and cloud providers. When they are willing to give up and associate themselves with a supplier stake.
“Some academics who want to go into their own start-up have their first conversation with cloud providers before they even recruit developers because they know it’s impossibly expensive. That’s the key ,” said Tom Bees, chief executive officer of Yellow Dog. He declined to name any of the companies involved because of non-disclosure agreements signed with Big Tech cloud providers.
Such deals can quickly draw regulatory scrutiny. A motion aimed at the so-called self-preferential behavior of the tech giants was advanced in the US Congress last year to prevent large online platforms from using their influence in one area to promote their other products.
“These platforms use their dominance to unfairly harm their rivals,” US Democratic Senator Amy Klobuchar said in a statement last year. “All at the expense of competition and consumers.”
Additional reporting by Tim Bradshaw in London and Richard Waters in San Francisco