Amazon says yes commit as much as $230 million for startups creating generative applications based on artificial intelligence.
The roughly $80 million investment to fund Amazon’s second AWS Generative AI Accelerator program is intended to position AWS as an attractive cloud infrastructure selection for startups developing generative AI models to power their products, applications and services. Much of the recent tranche – including the entire portion earmarked for the accelerator program – is in the type of compute credits for AWS infrastructure, which suggests it can’t be transferred to other cloud service providers reminiscent of Google Cloud and Microsoft Azure.
To sweeten the deal, Amazon guarantees that startups in this 12 months’s Generative AI Accelerator cohort will gain access to experts and technology from Nvidia, the show’s presenting partner. They will even be invited to the Nvidia Inception program, which provides corporations with the opportunity to attach with potential investors and provide additional consulting resources.
The Generative AI Accelerator program has also grown significantly. Last 12 months’s cohort of 21 startups only received as much as $300,000 in AWS compute credits, representing a total investment of roughly $6.3 million.
“Through this new venture, we will help startups launch and scale world-class enterprises by providing the building blocks needed to unleash new AI applications that will impact all aspects of how the world learns, connects and does business,” Matt Wood, vice chairman for AI products at AWS, he said in a statement.
Amazon’s growing spending on generative AI technology, which incorporates efforts like the $100 million AWS Generative AI Innovation Center, free credits for startups using mainstream AI models and his Project Olympus the model comes as the company looks to catch up with tech giants in the thriving – and increasingly competitive – generative artificial intelligence space. Although Amazon claims that various generative AI corporations have it achieved “multi-billions” in productivity, it is widely believed that the company did not leverage generative AI.
Originally AWS planned to present its own generative model of artificial intelligence just like OpenAI’s ChatGPT codenamed Bedrock – which eventually became owned by Amazon Bedrock at its annual conference in November 2022, in accordance with The Information. However, serious errors forced the organization to postpone the launch. (Amazon PR disputes this.)
Amazon Alex’s section was stuffed with challenges also because of technical failures and political conflicts, as Fortune’s Sharon Goldman reported this week. Nine months after a spectacular press demonstration of the “next-generation” Alexa, the recent Alexa is reportedly far from ready for launch – the results of insufficient training data, insufficient access to training equipment and other obstacles.
Amazon also took early opportunities to support two leading artificial intelligence startups, Cohere and Anthropic. The company later tried to take a position in Cohere, but was rejected and needed to settle for a co-investment (albeit a large one, totaling $4 billion) in Anthropic with archrival Google.
In addition to the recent departure of Howard Wright, the head of startups at AWS who managed the organization’s startup relationships, an obstacle in Amazon’s path is growing regulatory scrutiny over Big Tech’s investments in AI startups.
The US Federal Trade Commission recently opened the site inquiry on Microsoft’s support for OpenAI, in addition to Google and Amazon’s investments in Anthropic. European policymakers do signaled they too are skeptical of such deals.