Just_super | E+ | Getty Images
“AI is not where the money makes more money,” Calkins told CNBC in an interview in his London bureau on Tuesday.
Calkins was referring to high-profile deals such as Microsoft and Amazon's agreements with ambitious and fast-growing fundamental AI modelers such as OpenAI and Anthropic.
Microsoft has invested a total of $13 billion in OpenAI, a deal that would see Microsoft acquire a stake in OpenAI and later sell its GPT language models to the Redmond, Washington-based technology giant. To add to the Azure cloud computing platform.
Microsoft has struck a similar deal with Mistral, taking a 15 million euro ($16 million) stake in the French AI firm.
In OpenAI's case, Microsoft has a non-voting observer on the firm's board.
This follows a shocking series of events last year that saw OpenAI CEO Sam Altman temporarily ousted, before hundreds of OpenAI employees left Microsoft before returning. Altman threatened a coup to join.
Separately, Amazon has invested $4 billion in US AI firm Anthropic, which is behind the Cloud AI system. Amazon holds a minority stake in Anthropic but has no board seats.
Google has also committed billions in funding to Anthropic, agreeing to invest up to $2 billion last year.
British regulators are examining whether deals struck by Microsoft and Amazon with foundational AI model startups could constitute effective mergers that could substantially lessen competition.
Microsoft repudiates its contracts with OpenAI and Mistral and hires from Inflection formed integration. Amazon says its partnership with Anthropic is a limited corporate investment, not a merger.
This is a smart market. The fact that you have enough money to buy or purchase an Anthropic or a Mistral or one of those pieces is impressive. But AI may not be a 'winner-take-all' market.
Matt Calkins
CEO, Appian
For Calkins, whether or not the deals qualify as mergers that threaten competition in AI, there will be room for innovators to flourish.
“If allies had won the AI race, Google would have won by now,” he said, invoking the US tech giant's $500 million takeover of British AI lab DeepMind.
Far from it, Calkins argues — instead, he thinks Google lost early to Microsoft when it came to creative AI, which threatens to erode the fabric of Google's search business.
It follows an error that saw Google's Gemini text-to-image generator produce errors in historical images that went viral online. Google stopped photographing people to improve the tool. CEO Sundar Pichai called the defeat “unacceptable,” according to an internal memo obtained by CNBC in February.
Google was not immediately available for comment and was contacted by CNBC.
“It's a market for the smart,” Calkins said. “The fact that you have enough money to buy or buy a piece of Anthropic or Mistral or one of those is impressive. But AI can't be a 'winner take all' market.”
“There are going to be different AI algorithms for different purposes, and they're going to be more or less valuable, depending on what and how you load your data into it,” he added.
Calkins said that the only way for AI systems to become truly intelligent and effective is to be able to understand what we want them to use in our daily lives.
“The best AI will be the AI you put your data into, not the one that bought the biggest stack,” he said.
Calkins said the AI race today has become more about “how much data you can eat” than how smart the AI actually is.
Big tech companies are “doing everything they can to get as much data as possible,” Calkins said. “But that game was almost over,” he added.
That's because, with no concrete laws in place to prevent privacy violations from Big Tech's data hoarding, these companies were allowed to obtain the data they needed to train their models.
Calkins said he was frustrated by the lack of progress on AI regulation at the federal level in the US.
He told CNBC's “Squawk Box Europe” on Tuesday that Europe kind of has a “head start” on AI “because of the emerging clarity on regulation.”
“In the United States, it's not as clear, in part because the government has been a little too friendly to Big Tech,” he said.
The European Union formally adopted its AI Act in March, the first comprehensive law governing artificial intelligence.
Businesses need clarity on how they can safely use AI and guarantee things like intellectual property protection and users' personal privacy, Calkins said.
“There's a natural skepticism in Europe … here, we have regulations that push back against American big tech firms,” Calkins said.
“I would suggest that it's time again for fair use of copyrighted information. We need a clear playing field, we need to understand what data we are allowed to use.”