|
Listen to this story
|
As Microsoft and Google wrestle to build generative AI chatbots and integrate them into their search engines and productivity suites, Amazon enters the ring, albeit with a bit of diversion. Instead of building AI models on their own, AWS has announced an API platform, called Amazon Bedrock, which helps customers host generative AI models built by startups like AI21 Labs, Anthropic AI and Stability AI.
Amazon’s Play in Generative AI with Bedrock
Amazon’s strategy is a smart one. It stays fairly neutral while leveraging its massive market share of 33% in the cloud segment. Everybody’s a winner in this deal – developers can pay to use the APIs to these models plus the infrastructure that supports these services. Amazon wins simply because it is the market leader in the cloud while offering startups like Stability AI some much-needed help. A recent exclusive by Semafor revealed how Emad Mostaque’s company, which was behind the popular text-to-image generator Stable Diffusion, had burned through wads of cash and was at the risk of going under.
Even as Bedrock itself was under wraps, AWS did not release any research around new AI models that set them up against the fierce contest that other big-tech companies were embroiled in. Instead, AWS went on a signing spree inking partnerships with Stability AI in November last year, and Hugging Face and NVIDIA in March.
Bedrock is already estimated to reach around USD 110 billion in value by 2030, according to data from Grand View Research.
Gen AI: An Important Piece in the Cloud Puzzle
There’s more to the generative AI push than meets the eye. Microsoft has integrated its generative AI tools like ChatGPT into Azure. Last month, Google also said it would sell access to the Pathways Language Model or PaLM to developers using the company’s cloud services. Turns out, what hyperscalers want to do in reality is sell their cloud services with the flashy sticker of generative AI put on them.
The last quarter was rough for the cloud infrastructure market as spending soared and companies searched for ways to cut back on their expenditure in an uncertain economy. Synergy Research found that growth for the biggest cloud players (AWS, Azure and Google) had all slowed down in the fourth-quarter of 2022.
Microsoft had managed to gain ground on Amazon with an increase in their market share from 21% in the previous quarter to 23% in this quarter. AWS’ share fell from 34% to 33%, while Google remained at 11%. These were the lowest numbers in growth that AWS had recorded in its history. By the first month of this year, the growth dropped even further to the mid-teens.
Being the first to break into the market, AWS has had the advantage of a long head start but the market has been catching up after years of constant growth. There are worries that AWS might be finally reaching a point of market maturity. The generative AI partnerships may be the opening that AWS was looking for.
It is no wonder that all tech companies have put generative AI front-and-center now. In their earnings call last month, Oracle chief Larry Ellison said, “There’s actually more demand for AI processing than there is available capacity. We are expanding as fast as we can.”
AWS seems to have figured out that enterprise might be the missing link to making real moolah out of generative AI and cloud. Bratin Saha, VP of AI/ML services at Amazon spoke to VentureBeat saying, “The fact that customers will be able to use foundation models with the AWS enterprise security and privacy guarantees, we think, makes it much easier for using these models at scale. Customers will be able to use Amazon Bedrock in the same environment and with the same AWS services that they’re already comfortable with using.”
As exciting as generative AI is, it needs cloud to exist. And turns out, the cloud computing market overall can get a big push from generative AI too.
In an interview with Money Control, Nasscom chairman Krishnan Ramanujan discussed how generative AI’s success was deeply dependent on how widespread its usage is. “It has the potential to completely reset the economics of coding by significantly improving productivity and efficiency. However, the true benefits, both in terms of productivity and transformation, will be realised when ChatGPT, Bard, or any of the generative AI technologies can deliver enterprise use cases,” he stated.
All of this stands true – after all, what use is a cool potent technology unless it can help companies everywhere?
