Low prices. Wide choice Hard comfort.
These have been the decisive keys to Amazon’s e-commerce dominance over the past two decades.
Now, it looks like the tech titan is duplicating this winning playbook by borrowing parts of it to fuel a new AI strategy. Amazon Web Services division this is partly due to the low prices and wide selection.
To be clear, it will be a long time before the business world judges the effectiveness and financial sustainability of Amazon’s approach. But, if successful, the game plan would go a long way toward quieting critics who argue that Amazon is getting caught up in the AI wars, while also making the future company one of the world’s most powerful and influential tech corporations. in the coming decades.
Amazon executives unveiled crucial parts of their AI strategy this week at the AWS Re:Invent conference in Las Vegas. A key element is a portfolio of built-in foundational models, known as FM or LLM, called Nova, which can handle text, image and video queries, respectively.
Amazon’s introduction of a new class of native AI models can, on the surface, be confusing because the company already has one. He invested $8 billion in Anthropicgen creator of the popular Claude family of AI models. But, like my colleague Sharon Goldman recently pointed outAmazon is betting that “there will never be one tool” or AI model to rule them all. In short, Amazon believes that companies will want to choose models, such as Amazon, Anthropic or other tech giants like Meta.
Amazon actually started pushing this idea of offering a selection of AI models through a single API to business customers when its AWS division introduced a service called Amazon Bedrock last year. With Bedrock, business customers would be able to choose from a relatively limited selection of AI models, but a selection nonetheless to train for their needs and form the basis for their own AI applications.
On Wednesday, Amazon doubled down on its strategy with the announcement of Bedrock Marketplace, which offers a total of 100 AI models. LLMs on the market come from a variety of companies, and some are designed for specialized purposes.
“Finding and evaluating these models can be difficult and expensive,” Amazon said in a blog post announcing the market. “You have to find them in different services, create abstractions to use in your applications, and create complex layers of security and governance. The Amazon Bedrock Marketplace addresses these challenges by providing a single interface to access both specialized-purpose and general-purpose (basic models).”
Looking at Amazon’s e-commerce business, a key piece is the Amazon Marketplace, where hundreds of thousands external retailers list products For sale which make up 60% of all goods sold through Amazon. Amazon supplements the selection of these marketplace sales by also selling its inventory of goods, including under its brand name when a certain product category or price point is not met by marketplace sellers or Amazon’s brand partners.
Amazon also provides businesses with an enterprise version of marketplace AI that one might imagine will only be selectively deployed in the future. (It’s also worth noting that Amazon’s flagship AWS business offers a marketplace of more than 10,000 software tools spanning categories from cybersecurity to data analytics.)
Low prices have also been another hallmark of Amazon’s retail business. Amazon aggressively matches other retailers’ prices, throwing two huge discount events that attract big spenders and new Prime customers. (The FTC has argued in its anti-Amazon case that the e-commerce giant artificially inflates prices for consumers by penalizing retailers who undersell products at other retailers, but that’s a topic for another day).
And of course, Amazon CEO Andy JassyThe slide deck introducing the new Nova AI models began with pricing: “75% more profitable” was the first feature to be called.
Simon Willison, an independent AI researcher, did a quick test and agreed, writing on social networking app Bluesky that Amazon’s “price and performance (are) competitive. google Gemini family, that means they’re _really_ cheap.”
“With this release I believe Amazon has earned a place among the top tier of model providers,” Willison added. “Maybe we need a new FAANG acronym that includes OpenAI, Anthropic, Google, Meta and Amazon. I like GAMOA”.
Amazon’s leaders don’t care what the new acronym is called as long as they earn a place there. If they do, the company’s longstanding attributes of low prices and selection are likely the main reasons.
Are you a current or former Amazon or AWS employee with thoughts or advice to share on this topic? Contact Jason Del Rey at jason.delrey@fortune.com, jasondelrey@protonmail.comor through the messaging apps Signal and WhatsApp 917-655-4267. You can also write to him It’s LinkedIn or at @delrey good X.
