The assumption of investors and most media is that the Cloud Czars are going to dominate the AI age.
This runs counter to all past narratives. It assumes industry leaders cross a technology chasm untouched.
Remember that in 1995, AT&T and IBM dominated the Internet. Before streaming cable companies, many owned by broadcasters, dominated video. Remember newspapers?
It has always been the case that startups are able to muscle into the Great Game when the game changes. Google and Facebook didn’t exist in 1995. Apple and Microsoft were start-ups in 1975.
There are exceptions, big companies that remain big after a tremendous change happens. IBM was dominant in the technology of typewriters, adding machines and punch cards. Yet they managed to muscle their way into mainframe computing and, with the System 360, defined it. Big radio broadcasters became big in TV, after the government handed them necessary frequencies in the 1940s.
Yet the narrative remains. Google sells for 25 times earnings, Meta (formerly Facebook) for 27. Apple sells for 29, Microsoft for 35, and Amazon for 52. Their cash flow is paying for the Nvidia upgrades needed to make AI happen.
Startup Nation
But it turns out we’re in a start-up boom. As The Economist reported recently , we’ve been averaging 400,000 new business applications each month since the COVID pandemic began easing off. Among the fastest-growing sectors for start-up energy is AI, according to the Census Bureau.
This is not what financial reporters are telling you to expect. It is what a technologist would naturally expect. Most of these start-ups will either fail or get bought out. But some are going to come up with unique tweaks. Some are going to be backed by patient capital rather than get-richer-quick capital. There will be new Googles and new Facebooks. Some of today’s giants are going to fall, as IBM and AT&T fell.
Where might next Facebook come from? Some of today’s start-ups are focused on the application of corporate data to specific use cases. They’re trying to build cross-industry systems that can be applied generally. Others are building AI solutions under contract, like system architects of 30 years ago. Some are working on interfaces that will help Apple make AI profits on its Watch or Vision Pro. Others are working to make AI more compute-efficient or bring it to new geographies like Africa and India.
Most have yet to emerge, and only a few will emerge before next year. But they will emerge. You can be certain of it.
If you’re a young tech reporter on the make, this is what you want to be covering.
