Generative AI ‘FOMO’ is driving tech heavyweights to invest billions of dollars in startups

Tech titans aren’t doing much acquiring these days, owing primarily to an unfriendly regulatory climate. However, they are discovering new ways to spend billions of dollars on the next big thing.

Amazon announced a $2.75 billion investment in Anthropic, its largest venture acquisition to date. This is part of the AI gold rush that is causing major internet corporations to invest heavily.

Anthropic developed the AI model Claude, which competes with GPT from Microsoft-backed OpenAI and Google’s Gemini. Meta, Apple, and other companies are integrating generative AI into their products and features to stay competitive in a market expected to generate over $1 billion.

Within a decade.

According to PitchBook, investors invested $29.1 billion in roughly 700 generative AI investments in 2023, representing a more than 260% increase in value over the previous year.

A considerable portion of that money was strategic, meaning it came from tech businesses rather than venture capitalists or other organizations. According to Fred Havemeyer, head of U.S. AI and software research at Macquarie, fear of losing out is one of the factors motivating their decisions.

“They definitely don’t want to miss out on being part of the AI ecosystem,” Havemeyer went on to say. “I definitely think that there’s FOMO in this marketplace.”

AI models are costly to construct and train, needing thousands of specialized chips, mostly from Nvidia. Therefore, significant investments are required. Meta, which is building its own model dubbed Llama, has announced that it will spend billions of dollars on Nvidia’s graphics processing units, one of many firms that have helped the chipmaker increase year-over-year revenue by more than 250%.

Whether building or investing, only a limited number of enterprises can afford to participate in the market. In addition to producing processors, Nvidia has emerged as one of Silicon Valley’s top investors, investing in a number of budding AI businesses, in part to ensure that its technology is broadly adopted.

Similarly, Microsoft, Google, and Amazon occasionally provide cloud credits as part of their investments.

In the Amazon-Anthropic agreement unveiled on Wednesday, the two businesses stated they will collaborate closely in a variety of ways. Anthropic will use Amazon Web Services and Amazon chips to meet its computing demands. Amazon will share Anthropic’s models to its AWS clients.

Earlier this month, Anthropic introduced Claude 3, its most powerful model, which allows users to input photographs, charts, documents, and other unstructured data for analysis and replies.

Microsoft first entered the generative AI investment market in 2019, spending $1 billion in OpenAI. Since then, its investment has increased to over $13 billion. Microsoft significantly relies on OpenAI’s model and provides open-source models on its Azure cloud.

Alphabet is taking on the roles of both constructor and investor. The business has refocused much of its product development on generative AI and its newly rebranded Gemini model, adding functionality to search, documents, maps, and other areas. Last year, Google agreed to investing $2 billion in Anthropic, having earlier confirming a 10% ownership in the startup and a major cloud contract between the two companies.

According to Havemeyer, IT giants aren’t merely throwing money at the “hype cycle,” as their investments in AI startups correspond with their product roadmaps.

“I don’t think it’s frivolous,” he replied.

According to Havemeyer, collaborations with large cloud providers not only offer startups with much-needed funding, but also assist them in acquiring clients.

“Come to us, work on our platform, have native access to the latest and greatest AI models, and also use our infrastructure,” Havemeyer said. “It’s also part of a much larger ecosystem play.”


“We’re seeing a lot of alliances appearing among those hyperscalers that have substantial scale, infrastructure and very deep pockets,” he went on to say.

‘Shaping the next decade’
In recent earnings calls, tech executives underlined their commitment to generative AI, emphasizing to investors that they must spend money to produce money, whether through internal development or investing in startups.

Microsoft Chief Financial Officer Amy Hood stated last year that the corporation was adapting its “workforce toward the AI-first work we’re doing without adding a material number of people to the workforce.” She stated that Microsoft will continue to prioritize investing in AI as “the thing that’s going to shape the next decade.”

Google, Apple, and Amazon executives have also indicated to investors that they are willing to reduce costs across departments in order to redirect more funds toward their AI research.

Startups are among the beneficiaries.

Microsoft has made investments in Mistral, Figure, and Humane, in addition to OpenAI. The company invested in Inflection AI before it was basically dissolved and acquired by Microsoft earlier this month. Mistral is an open source-focused startup that uses Azure’s cloud and provides services to Azure customers.


Figure, a firm that aims to create a robot that walks like a person, raised funds from Microsoft, OpenAI, and Nvidia and was valued at $2.6 billion last month.

Anthropic is Amazon’s greatest bet, with a total investment of $4 billion so far. Hugging Face, the provider of an open source AI platform, has also received investment from the firm.

Google has invested in Essential AI, a company supported by AMD and Nvidia that develops consumer AI systems. Alphabet and Nvidia are also investors in Runway ML, a generative AI firm known for its video editing and visual effects software. Nvidia’s portfolio also includes Mistral, Perplexity, and Cohere.

Meanwhile, many Big Tech companies continue to invest internally in establishing their own models.

Microsoft has invested in many of the techniques that enable generative AI through its Microsoft Research business. Amazon apparently intends to train a larger, more data-hungry model than even OpenAI’s GPT-4.

Apple researchers recently shared details about their work on MM1, a family of modest AI models that can accept both text and visual input. Apple is in a unique position among its contemporaries because it does not sell cloud services. The internet giant is apparently seeking AI partners, including Google in the US and Baidu in China. An Apple spokeswoman declined to comment on AI partnerships.

Creativity in negotiations
According to Daniel Newman, CEO of technology analysis firm Futurum Group, tech businesses must be strategic when investing in AI.

For example, Microsoft’s investment in OpenAI included profit share in a nonprofit wing as well as credits for using Microsoft’s cloud service. Microsoft’s acquisition of Inflection AI was a pricey one, with some sources estimating a total cost of $1 billion. As part of the purchase, Microsoft engaged Mustafa Suleyman, the founder of Inflection AI, to manage Copilot AI activities.

“I think we’re starting to see some creativity and dealmaking,” Mr. Newman said. In reference to Amazon’s arrangement with Anthropic, he stated that an acquisition would be “a lot harder than investing.”

This is because regulators around the world are pushing down on Big Tech, making large-scale acquisitions more difficult. Even the investments are being scrutinized.

The Federal Trade Commission said in January that it will launch a thorough investigation into the industry’s major AI firms, including Amazon, Alphabet, Microsoft, Anthropic, and OpenAI.

The FTC Chair, Lina Khan, defined the investigation as a “market inquiry into the investments and partnerships being formed between AI developers and major cloud service providers.” The regulator has the right to require corporations to file specific reports or provide written responses to questions regarding their enterprises.

“We know regulators are becoming increasingly focused on the traditional path of closing an acquisition,” Newman said in a statement. “Right now, the game is having access to the most fundamental IP.”