The AI vibe shift, Sebastian Thrun is the textbook example of an AI optimist
Sebastian Thrun is the textbook example of an AI optimist. An early AI leader,
Thrun created Google's self-driving car unit and was the director of the
Stanford Artificial Intelligence Laboratory. Thrun stresses he has "no doubt"
that AI will fundamentally change the way we work and live. But even he has
concerns about the current state of the AI market.
"In my adult life, I've never seen a hype cycle this extreme," Thrun said in an interview, noting that many AI startups don't seem to have a sustainable enough business model to support their sky-high valuations. "In every hype cycle, there is a moment of extreme hype and there is a moment of sobering. We might be entering that moment of sobering."
Thrun isn't the only one questioning whether there's an AI vibe shift underway. The launch of OpenAI's ChatGPT in late 2022 prompted startups and large techology companies to invest tens of billions in building large language models to power new generative AI services. But in recent weeks, there have been heightened doubts in the public and private market.
Wall Street investors are growing impatient to see clear payoffs from Big Tech's heavy spending on AI hardware, development and talent. At the same time, a series of once-high-flying AI startups have either had layoffs or been almost-but-not-quite acquired in unusual deals with large tech firms, a sign of possible consolidation in the industry.
Much of the industry has settled on a similar playbook: selling artificial intelligence software to businesses to offset the high computing costs of building AI. But the harsh reality is that it's difficult for most companies to go up against the likes of OpenAI, Google and Anthropic, which have access to top AI talent and financial resources.
Sebastian Thrun Photographer: Martina Albertazzi/Bloomberg
"If you're trying to create value in building a chatbot, you can be completely out-innovated by foundation model providers within a matter of minutes," said Navrina Singh, founder and chief executive officer of AI governance platform Credo AI. "So I can see the nervousness that some of the investors have."
Many of Credo AI's customers are large companies who want help responsibly implementing AI technologies. Singh said she's seen "case after case" of businesses getting real value from using generative AI models, particularly when they're integrating these services with their own data and doing so in a careful, risk-aware way. But not every AI startup promising to drive efficiency for their corporate customers is actually delivering.
"There's a lot of hype and excitement around Gen AI," Singh said, "but within the next 12 months or less, we will start to see how much people are lighting their money on fire versus actually getting something valuable."
Some venture investors I spoke with acknowledged the concern about whether AI companies will reap enough rewards from their enterprise offerings, but still didn't seem worried. It's just a matter of time, they say, for these AI tools to become more capable and for larger companies to fully embrace them and unlock value.
"Across our portfolio, the adoption of AI is moving faster than any tech wave I've seen before," said Menlo Ventures partner Matt Murphy, who led the firm's investment in Anthropic. "Some of the current skepticism is going to be muted as more of these use cases are launched in the market."
"I don't think we're going to go through the classic trough of disillusionment," Murphy added. "If you aren't moving quickly with AI, you are falling behind your competitors and risk being obsolete."
In the meantime, though, anxiety is growing at some startups. One founder, who spoke on condition of anonymity to protect professional relationships, called it "an extremely dangerous time." The founder pointed to the breathtaking amount of money companies like Meta and Google are spending on AI. "I don't think any startup except OpenAI can keep up."
"In my adult life, I've never seen a hype cycle this extreme," Thrun said in an interview, noting that many AI startups don't seem to have a sustainable enough business model to support their sky-high valuations. "In every hype cycle, there is a moment of extreme hype and there is a moment of sobering. We might be entering that moment of sobering."
Thrun isn't the only one questioning whether there's an AI vibe shift underway. The launch of OpenAI's ChatGPT in late 2022 prompted startups and large techology companies to invest tens of billions in building large language models to power new generative AI services. But in recent weeks, there have been heightened doubts in the public and private market.
Wall Street investors are growing impatient to see clear payoffs from Big Tech's heavy spending on AI hardware, development and talent. At the same time, a series of once-high-flying AI startups have either had layoffs or been almost-but-not-quite acquired in unusual deals with large tech firms, a sign of possible consolidation in the industry.
Much of the industry has settled on a similar playbook: selling artificial intelligence software to businesses to offset the high computing costs of building AI. But the harsh reality is that it's difficult for most companies to go up against the likes of OpenAI, Google and Anthropic, which have access to top AI talent and financial resources.
Sebastian Thrun Photographer: Martina Albertazzi/Bloomberg
"If you're trying to create value in building a chatbot, you can be completely out-innovated by foundation model providers within a matter of minutes," said Navrina Singh, founder and chief executive officer of AI governance platform Credo AI. "So I can see the nervousness that some of the investors have."
Many of Credo AI's customers are large companies who want help responsibly implementing AI technologies. Singh said she's seen "case after case" of businesses getting real value from using generative AI models, particularly when they're integrating these services with their own data and doing so in a careful, risk-aware way. But not every AI startup promising to drive efficiency for their corporate customers is actually delivering.
"There's a lot of hype and excitement around Gen AI," Singh said, "but within the next 12 months or less, we will start to see how much people are lighting their money on fire versus actually getting something valuable."
Some venture investors I spoke with acknowledged the concern about whether AI companies will reap enough rewards from their enterprise offerings, but still didn't seem worried. It's just a matter of time, they say, for these AI tools to become more capable and for larger companies to fully embrace them and unlock value.
"Across our portfolio, the adoption of AI is moving faster than any tech wave I've seen before," said Menlo Ventures partner Matt Murphy, who led the firm's investment in Anthropic. "Some of the current skepticism is going to be muted as more of these use cases are launched in the market."
"I don't think we're going to go through the classic trough of disillusionment," Murphy added. "If you aren't moving quickly with AI, you are falling behind your competitors and risk being obsolete."
In the meantime, though, anxiety is growing at some startups. One founder, who spoke on condition of anonymity to protect professional relationships, called it "an extremely dangerous time." The founder pointed to the breathtaking amount of money companies like Meta and Google are spending on AI. "I don't think any startup except OpenAI can keep up."
No comments