When Google is willing to spend $2.4 billion for a handful of engineers and the non-exclusive rights to use a bit of artificial intelligence software, you know things have gotten out of hand.
Google’s deal to poach the CEO and select staff of AI startup Windsurf, swooping in after OpenAI’s $3 billion takeover attempt folded, isn’t unique. On Wednesday, Thinking Machines Lab, another AI startup started by former OpenAI executive Mira Murati, raised $2 billion in a funding round that valued the company at $12 billion — and it doesn’t have any products yet. Anthropic, the startup behind the Claude chatbot, is eying a funding round with a $100 billion valuation. This all came after AI-chip supplier Nvidia briefly hit the $4 trillion threshold in market capitalization.
You read that right. That’s “trillion” with a T.
Compare that to the media world. Paramount Global, a company that had roughly 18,600 employees at the end of last year and boasts “Star Trek”, “Mission: Impossible” and Stephen Colbert (for now) among its stable of pop culture touchstones, will be acquired for only $8 billion by Skydance Media. Walt Disney Co., one of the crown jewel companies in media, is worth $218.6 billion.
The stark contrast in valuation underscores how wild the frenzy is for AI — the promised next big thing for the last two and a half years — driving valuations sky high based on a niche bit of code or even the promise of the right skills from the right engineer. It also harkens back to the Dot Com boom of the 1990s, when every company with an internet presence senselessly skyrocketed in value, profit or not (and mostly not).
Indeed, Torsten Slok, chief economist for asset manager Apollo Global Management believes the situation is even worse now. “The difference between the IT bubble in the 1990s and the AI bubble today is that the top 10 companies in the S&P 500 today are more overvalued than they were in the 1990s,” Slok said in a note published last week.
But with the arms race in AI — whether it’s between tech giants like Google and Meta or, more broadly, against China — only accelerating, it doesn’t look like this bubble will burst anytime soon. The AI market is projected to hit $244.2 billion this year, but grow to $1.01 trillion by 2031, according to market data firm Statista. With that much at stake, the FOMO is real.
“AI is going to touch everything,” said Carolina Milanesi, an analyst and president of market research firm Creative Strategies. “From an impact perspective, it doesn’t matter if its consumers, enterprise or different verticals. It will disrupt everything.”
You’ve probably already experienced some of the benefits of AI yourself, whether through Google’s Gemini transcribing video conference calls, or bots scheduling your calendar or planning out a vacation. Then there’s the downside, with AI already replacing everyone from fast food workers to programmers and executives like Anthropic CEO Dario Amodei saying AI could erase half of all entry-level white collar jobs.
If you don’t live and breathe AI everyday, all of this can be jarring, and even completely head-scratching. But don’t worry; that’s what this article is for. Let’s make sense of what’s going on in the world of AI.
How we got here
The buzz around AI erupted in November 2022 with the emergence of OpenAI’s ChatGPT. While AI has been around for years in many forms – think automated customer service voice prompts – ChatGPT took things to another level with generative AI, which is smart enough to go beyond its initial programming and come up with fresh responses and solutions based on the data used to train its underlying algorithm, called a Large Language Model.
For the first time, companies and individuals alike understood the promise and potential of AI, and they embraced it at an unprecedented rate. ChatGPT reached 100 million monthly active users in just two months — a feat that took Instagram two and a half years to achieve.
Big tech companies like Google and Meta noticed, and quickly pivoted. These companies each “own” strong relationships with people, whether through YouTube, search, Instagram or Facebook, that are critical pipelines worth billions of dollars in revenue. They recognized that over the long term, ChatGPT threatened to replace many or all of those relationships if they stood still.
So all of a sudden, these big tech companies were all-in on generative AI and large language models too.
A frantic, relentless pace
A myriad of competing AI chatbots, which serve as the interface to the underlying LLM, have sprung up. Beyond ChatGPT, there’s Anthropic’s Claude and Google’s Gemini. Meta offers Llama, an LLM that’s open source and available for people to use for research and commercial purposes, allowing them to build their own AI-powered applications.
The reason why there’s such a bidding war for companies is simple: The pace of innovation has been relentless.
It takes a year for a new iPhone to arrive. Computers may get a software update every few months. Updates to AI models offering new capabilities like multimodal inputs, which allow you to speak or feed it images that it can understand, are coming out every few weeks. And they’re getting smarter, more powerful and more efficient all the time. Earlier this month, Elon Musk showed off X.ai’s Grok 4, which is now considered the most powerful AI model. It came a week after the Grok chatbot had a total antisemitic meltdown.
Beyond paying attention to the various competitors, the tech giants also have to be mindful of China, which has more aggressively fostered AI development with fewer ethical guardrails. In February, a Chinese AI model called DeepSeek came out of nowhere as a model with the same capabilities as some of the leading players, but trained at a fraction of the cost.
Then there’s the talent, which Arun Chandrasekaran, an analyst at Gartner, said is exceedingly rare. It’s why Meta CEO Mark Zuckerberg poached key AI researchers from OpenAI and Apple for his Superintelligence Labs team, and why Murati can command such a high valuation for Thinking Labs without any products or services. Venture capitalists are throwing cash with reckless abandon because they’re all looking for the next OpenAI.
If you think all of that is dizzying, imagine being one of these companies tasked with the goal of staying ahead of the curve.
“There’s a frantic pace that we’ve not seen before in technology,” Milanesi said. “There’s a real sense of urgency.”
What AI bubble?
If there is an AI bubble, experts don’t think it will burst anytime soon. The valuations will continue to soar thanks to the long-term prospects and constant gamesmanship between tech giants and the investment community.
The Economist calls it “vibe valuing,” or “the ability of venture capitalists to conjure up vast valuations for AI startups with scant regard for old-school spreadsheet measures.”
In other words, there’s no real fundamental basis for these calculations, but if everyone else is doing it, so be it.
There’s at least some explanation for the recent deals. The Windsurf talent grab was all about the AI-based programming capabilities that the startup was working on. Those capabilities — and the brains behind them — are attractive because they represent the promise of AI-built services that can be offered, allowing Google to create a new relationship with customers down the line, Chandrasekaran said.
“We’re excited to welcome some top AI coding talent from Windsurf’s team to Google DeepMind to advance our work in agentic coding,” said a Google spokesperson.
Spokespeople for Anthropic and Thinking Machines Lab couldn’t be reached for comment.
There are challenges that may dampen the unbridled enthusiasm. One is litigation, as we’ve seen with Disney and Universal suing AI company Midjourney over its ability to generate characters that appear like Elsa from “Frozen” or the Minions. The other is government regulation, with the EU likely to present the strictest limits.
It’s unclear just how extensive those guardrails will be, and when we’ll actually see them. For now, the frenzy continues.
“I don’t think anyone can make sense of these crazy numbers,” Chandrasekaran said.
