Startups and investors are learning, once again, that great technology doesn’t necessarily translate into great business. After months of hype surrounding the potential of generative AI, investors and startups are ratcheting back their enthusiasm and trying to take a more measured approach to the market.
A number of startups built on a foundation of AI are confronting declining customer interest and the need for layoffs, reports The Wall Street Journal. Adding to the pressure: Investors aren’t convinced that new companies will survive as brand-names like Microsoft and Google push into the space.
In addition, the flood of products unveiled since last year is bewildering many technology customers. On the one hand, they want time to understand how AI’s capabilities can fit their needs. On the other, they’re pressuring vendors to keep up with developments. “Some of our customers, all they want to hear is that we’re thinking about AI,” said Ellen Loeshelle, director of product management – intelligence platform at Qualtrics. “Like I could say that in one sentence and get off the phone and they’d be happy.”
“There is a ton of VC propaganda” out there about investments in AI, said John Luttig, an investor at Founders Fund, which has invested in ChatGPT developer OpenAI. “It’s an unequivocal optimism without asking any of the hard questions on product, user interface, distribution or end markets.”
For example, consider chatbots. They’ve been around for a while but the advent of generative AI raised the market’s stakes as developers hyped improved accuracy and more natural interactions between machines and users. ChatGPT’s breakneck growth led investors to believe customers would adopt such products quickly. That, in turn, encouraged them to pour money into fledgling businesses that had few customers and less revenue.
Now the bloom is off the rose. Visitors to ChatGPT’s website dropped by 10% in both June and July, the Journal said. Midjourney, whose technology creates images from user-created descriptions, reported a drop in visits in May, June and July. Over the past six months Synthesia, which uses text to create videos, has seen flat or declining website user growth.
Money and Size
Big companies, with their fatter wallets and large sets of established customers, are better able to withstand the pressures of a developing market. Well-funded VCs can also afford to take the long term view. “We’ve moved from a moment of ‘How big can this be?’ to ‘How do we make it work?’ ” explained Sunil Dhaliwal, a general partner at Amplify Partners.
Such attitudes are helping insulate generative-AI startups from dropping rates of venture funding, the Journal said. Pitchbook estimates that funding for those companies increased by nearly two-thirds, to $3.3 billion, during the period.
Still, developing and maintaining generative AI models is expensive. Because they work with such heavy volumes of data, training large language models can cost billions of dollars. By some estimates, just running ChatGPT costs OpenAI $700,000 a day.
How patient investors will be remains to be seen. OpenAI reportedly lost $540 million in 2022, despite ChatGPT’s popularity. While it’s unlikely investors will back away from generative AI bets entirely, numbers like that are sure to put a brake on things.