待翻译：AI is crushing a generation of startups built before ChatGPT

AI 服务暂时不可用，以下为来源正文，待恢复后补全翻译。

AI is crushing startup valuations for pre-ChatGPT firms Skip Navigation Nearly half of America's 857 unicorn startups haven't raised fresh funding in three years, PitchBook data shows. Startups that last raised in 2021 are worth 68% less on average, PitchBook found, while those that last raised in 2022 have seen their valuations decline 52%. More than 220 companies that once hit billion-dollar valuations are now considered "fallen unicorns" — including Glossier, Savage X Fenty, AG1 and The Farmer's Dog, according to PitchBook, which provided a list of the companies exclusively to CNBC. Largely to blame is the AI boom that has funneled more than $250 billion into OpenAI and Anthropic and reset valuations on entire classes of startups. Matthias Balk | Picture Alliance | Getty Images Five years ago, venture capitalists were pouring money into American startups selling everything from lingerie subscriptions to scheduling software, anointing them with billion-dollar valuations before most even turned a profit. It was a frothy era for startups, fueled by a combination of cheap money and pandemic-boosted demand. But even after the Federal Reserve took some froth off by starting to raise interest rates in 2022, many founders believed that they could grow into their inflated valuations, investors told CNBC. Then, an app called ChatGPT arrived. "The ChatGPT moment was when people said, 'Holy smokes, the next generation of entrepreneurs, their coding language is spoken English,'" said Samir Kaul, a partner at the venture firm Khosla Ventures, an early backer of OpenAI. "Now you're seeing 50 engineers do what it would've taken 500 engineers to do five years ago," Kaul said. "We had to completely reshuffle how we valued these companies." While the shares of public software companies like Salesforce, ServiceNow and Workday got hammered this year because of the threat from artificial intelligence, a quieter reckoning has been unfolding in the private markets. The AI boom that funneled more than $250 billion into OpenAI and Anthropic ahead of their expected mega-IPOs this year has left hundreds of startups built before ChatGPT's arrival in 2022 stranded — effectively cut off from venture funding because of their inflated valuations and outdated technology, yet not profitable enough for the public markets. There are 857 U.S. startups valued at $1 billion or more, the threshold for being deemed a "unicorn" company, according to PitchBook data. But nearly half of that group hasn't raised fresh funding in the last three years, making those valuations stale, according to the private markets data firm. Startups that last raised in 2021 are now worth 68% less on average, while those that last raised in 2022 saw a 52% decline, according to Pitchbook's own valuation estimates. As a result, more than 220 companies that had reached billion-dollar valuations in the venture boom are now fallen unicorns, according to PitchBook, which provided a list of the companies exclusively to CNBC. The estimates are based on factors including head count growth and comparisons with public companies. "A lot of those companies are pre-AI, not just in their cost structure, but also in their products," Mercury CEO Immad Akhund told CNBC. His company, which raised $200 million in funding last month, provides banking services to a third of early-stage U.S. venture-backed firms. "They're definitely in a difficult spot," he said. "All the attention's on AI, so if you're not an AI-first company, you need really strong numbers to raise." Glossier, Brooklinen, AG1 The list of fallen unicorns includes well-known brands like Glossier, The Farmer's Dog, Rothy's, Brooklinen and Savage X Fenty, the lingerie company founded by musician Rihanna. The companies were part of a wave of direct-to-consumer firms built on the hope that digital retailers could earn software-like margins. Also included are mainstays of podcast advertisements including the powder supplement maker AG1 and the robo-advisor pioneer Betterment as well as the online ticket marketplace SeatGeek. These companies came of age in an environment that rewarded growth at nose-bleed valuations based on two broad assumptions: interest rates would remain low and a startup could always be acquired for its engineering talent. But the arrival of generative AI has redrawn the venture landscape, redirecting capital toward AI-native firms while making it impossible for many older startups to justify their previous valuations. Hit hardest are enterprise software companies like scheduling startup Calendly, which represent the single largest category among the fallen unicorns. There are 75 software-as-a-service, or SaaS, firms appearing on PitchBook's list, which is double the number of fintech companies, the next-biggest group. That reflects both the enormous valuations that software startups commanded during the 2021 venture boom and the degree to which generative AI has destabilized assumptions underpinning the sector. David Zhu, an ex-DoorDash head of engineering, said that after the "ChatGPT moment" he looked across the software landscape — from startups to medium-sized firms funded with private credit to the largest public SaaS companies — and saw a seismic shift on the horizon. "The thesis I had was that all workflow-driven enterprise SaaS companies will be either disrupted or dead in the next decade," Zhu told CNBC. The SaaS model, where companies embed themselves in employee workflows and often charge by the user, is especially threatened by the rise of autonomous agents. After leaving DoorDash, where he led more than 200 engineers, Zhu founded Reevo, an AI platform that automates corporate sales and marketing teams. Companies built before generative AI are weighed down by bloated staffing models and software designed for a pre-AI world, according to Zhu, making it hard for them to transform themselves. "Unless they make a stark, 180-degree pivot to rebuild the exact same thing from scratch, they're going to slowly fail," Zhu said. "What that means is that investors would rather just bet on new entrepreneurs at lower valuations rather than double down on older startups." 'Dominoes to fall' Most of the 20 fallen unicorns highlighted by CNBC either didn't respond to multiple requests for comment or declined to comment. A spokesperson for the drone maker Skydio — estimated by PitchBook to have dropped in value from $2.5 billion to $509 million — said in a statement, "This third-party speculation is false and not based on Skydio's operations or the exponential growth we are seeing in revenue and customers." Weeks later, Skydio announced that it had raised $110 million through existing investors, raising its valuation to $4.4 billion. An AG1 spokesperson didn't provide a statement for this article, but after CNBC's inquiry, Reuters reported that the supplement maker was looking to sell part or all of the company at a $2 billion valuation. That figure would include AG1's debt, the report said. If a company hasn't raised funding since 2021 or 2022, its unlikely it'll ever do so again, say investors and founders. Without access to venture funding or a plausible initial public offering ramp, the most likely exit for many fallen unicorns is an acquisition at a fraction of their old valuation, they say. "When we see companies not raising, it's a red flag," said PitchBook analyst Andrew Akers, adding that it usually means their growth is tepid or even negative. While some startups might've avoided fundraising because they are generating robust profits, that is the exception to the rule, he said. "Underneath the surface, I think there are a lot of dominoes to fall," Akers said. Collapsing floor There have been glimmers of a reset among some startups this year. In February, Stash, the investment and savings app, was acquired by Singapore-based everything app Grab at an enterprise value of $425 million, below the roughly $660 million that investors put into the company during its lifetime. That same month, another fintech, Step, was acquired by the YouTube star MrBeast for an undisclosed amount, leading investors to speculate that the purchase price was far below the approximate $500 million the startup raised before the deal. "Many of these businesses just aren't worth that much anymore, which is why you're seeing them get acquired at steep discounts," said Ryan Falvey of Restive Ventures, which invests in fintech firms. Valuations have compressed by about sixfold from the 2021 peak of 50 times future revenues, meaning that a company with the same revenue is worth about 85% less in today's market than five years ago, Falvey told CNBC. Before the reset, a startup could often be sold to a larger technology company looking to acquire the smaller firm's engineers for roughly $2 million per coder, according to Khosla Ventures' Kaul. A firm with 100 engineers would be worth at least $200 million to $300 million, he said. But that assumption, which provided a floor under startup valuations during the boom, evaporated after AI coding tools allowed far smaller teams to build products — leaving exit opportunities few and far between. 'OpenAI, Anthropic or Google' The result is that post-GPT startups are running laps around their older competitors, according to Falvey. He called investments made over the past three years "undoubtedly the best" his firm has made. "We noticed by 2023 that the companies we invested in post-ChatGPT were already making more money than most of the companies we invested in before ChatGPT," Falvey said. Generative AI may ultimately reduce the amount of capital required to build successful software companies, challenging one of the core assumptions that fueled the venture boom of the past decade. The shakeout is probably just beginning, as the impact of AI reverberates across the business funding ecosystem, from venture to private credit to public giants. Older software firms, Kaul said, still rely on business models built around charging customers based on the number of employees using their products, an approach he believes AI will undermine as companies automate more white-collar work. Software providers will need to shift toward outcome-based pricing models and AI-native infrastructure to survive, he said. "The question I ask every time one of them presents is, why can't OpenAI, Anthropic or Google do this?" Kaul said. "For most of them, the answer is, 'They can.'" Choose CNBC as your preferred source on Google and never miss a moment from the most trusted name in business news.