The writer is the founder of Sievedan FT-backed site on European start-ups
California’s latest gold rush is all about a mad rush on generative AI. Big US tech companies, like Google, Microsoft, Meta and Palantir, as well as a swarm of venture capital firms, are all maniacally digging for new veins of digital treasure. But the big, and still unanswered, questions are: who will end up spitting dust and who will get the most gold?
The obvious assumption is that the large companies that are developing these text, image, video and audio generation models will dominate the field. As Lina Khan, chairman of the Federal Trade Commission, has written, a handful of powerful companies appear to control all the necessary commodities: vast data stores, computing power and cloud services. To which, she might have added: they also have many of the world’s leading AI researchers and mountains of cash.
“The expanding adoption of artificial intelligence risks further blocking the market dominance of large incumbent tech companies,” Khan wrote recently in the New York Times. Such a worldview provides more traction to Khan’s confidence boost.
But that’s not how the world looks to those inside a few big tech companies, judging by a leaked memo from a Google executive titled We Have No Moat. The executive noted this April that Google and OpenAI, heavily supported by Microsoft, may have developed the most capable and closed generative AI models, such as Bard and GPT-4. But they were already in danger of being outpaced by nimbler competitors who were building smaller, cheaper, more customizable open-source AI models and attracting some of Google’s top researchers. “The inconvenient truth is that we are not in a position to win this arms race and neither is OpenAI,” the executive wrote. “Who would pay for a Google product with usage restrictions if there is a free, high-quality alternative without them?”
This week, Google has been trying to up its game announcing that it would incorporate generative AI into an ever-widening range of services, looking to catch up with Microsoft. But, according to Google’s note, it could be the main beneficiary of the trend towards open source models Meta, which also pivoted towards AI. After launching its own open source LLaMA model in February, Meta is now aiming to build the platform for others to play on. Just as Google has built a new ecosystem of apps around its open source Android mobile phone software, Meta could emerge as the platform on which innovation happens. “The one clear winner in all of this is Meta,” the Google executive wrote.
Meanwhile, venture investors are betting that a new wave of generative AI startups, including Anthropic, Cohere, Stability AI, Inflection and AI21 Labs, can also successfully strike for gold. Their rationale is that at least some of these startups can move faster than larger companies, dominate select market niches, and largely ignore costly security checks (something that should alarm regulators).
The future will belong to smaller, more specialized generative AI models that are cheaper to train, faster to run, and suited to a specific use case, says Yoav Shoham, co-founder of Israeli start-up AI21 Labs. “The moat is not technology. It’s the relationship with the consumer,” Shoham tells me. “I think it will be a ‘few takes more’ market, not a ‘winner takes all’ market.” Dozens of other startups, which provide only a generic service and have no traction with customers, will fail.
Established companies in most industries that can feed their proprietary data into generative AI models and fine-tune the results are also likely to thrive. The challenge for them is to re-engineer their organizational structures to take advantage of the new technology. The other certain winners will be the “spade and shovel” companies that provide the tools for this technological transformation. The big cloud computing companies—AWS, Google, and Microsoft—will benefit from the models’ voracious demand for computing power. But Nvidia also stands out, the dominant designer of graphics processing units that power most AI models. “We are at the time of the iPhone for AI”, says Jensen Huang, chief executive officer of Nvidia.
Such is the speed at which the industry is evolving, however, that today’s best guesses could turn into tomorrow’s cheap betting slips. The history of other general-purpose technologies, such as electricity, automobiles, and the Internet, suggests that it is generative TO THE it will create new markets and business models that nobody can imagine today. It may be that a company yet to be founded will mine more gold.
—————————————————-
Source link