Even though Generative AI feels like the wave of the future, with an ever-growing list of potential applications, efficiency drivers, and monetization opportunities, it often seems as though a handful of players have already come to dominate the market, leaving little room for startups or new entrants. During our recent thought leadership webinar, Alex Johnston addressed this situation. A Research Analyst in the data, AI and analytics channel with 451 Research, Johnston noted the advantage that the first-mover cloud providers already have in GenAI. Here’s what he said about what the marketplace looks like and what opportunities still exist for would-be competitors.
“Obviously, the cloud providers are in a very strong place,” Johnston said of the firms that have already defined the GenAI landscape. “The way that we should see sort of hyperscalers is almost like a stack offering. And not only that, the infrastructure, obviously, and on their foundation models, and they've also wrapped these as sort of enterprise tooling, for example, for sort of model development.” As a result, he said, “A small improvement on the infrastructure side will sort of move to the rest of the stack, which is really interesting. So obviously, [the hyperscalers] are in a very strong position.”
Nonetheless, Johnston said he sees opportunities in the margins where innovation is occurring. “With our data, we do notice on the workload environment side of workload venues, organizations often start in public cloud, experimenting, and then often start to shift in production to maybe sort of private cloud, maybe we're also talking about edge applications, for example. And obviously, there's a lot of focus on how we compress language models, for example, moving out to the edge. So that's a slight caveat to that and public cloud focus.”
In many cases, of course, the major AI players are already absorbing the cutting-edge innovators. “There are some niche for quite exciting markets that the hyperscalers are involved in, really. So one, I think, tabular synthetic data or synthetic data more generally around numbers,” Johnston said. “Another one is data sets using tools about storing any of the utility of that data. And that, in turn, feeds into AI workloads. But actually, the hyperscalers are involved in that.”
In the short term, Johnston said he expects the large cloud players to continue to dominate without entirely squeezing out the startups. “The companies are involved to a small pool of startups effectively. They are a couple of years ahead of when this trend will be relevant. So there are fields that the public cloud players are playing, and there is opportunity as well, obviously, the other exciting technology providers,” Johnston said. “But yes, again, the hyperscalers are doing very well.”
To hear further insights from Johnston on the future of GenAI, access our complimentary, on-demand webinar.
For more on GenAI and tech trends, read our Big Picture outlook report.