Startups Face Reality Check As Venture Investors Begin To Demand Profits

Venture investors now find themselves victimized by their own hype about generative AI. Online visitors to ChatGPT as well as many other startup rivals have now been declining for months as potential users puzzle how to actually apply - and pay for - it. 

This comes as an overall decline in venture funding has caused investors to be more critical and hesitant about generative AI startups, especially as the dominant role being played by tech behemoths such as Google and Microsoft makes competitive opportunity for new enterprises less certain, given the expense required to train the models and launch them. But the tech and VC industries created and then actively defended this consolidated model, so there should be little surprise - or grousing - now that it has come to dominate the latest 'new, new thing.' JL

Berber Jin reports in the Wall Street Journal:

The initial surge in user growth for ChatGPT—the fastest ever for a consumer app—led investors to overestimate the rate at which consumers would use generative AI. Investors raced to back startups before they had customers or revenue. (But) monthly online visitors to ChatGPT decreased by 10% in June and July. (And) AI buzz hasn’t been enough to stem a precipitous drop in startup funding. U.S. venture funding plummeted by almost half in the second quarter, even as investment in generative AI startups increased 65%. Investors have become hesitant given the uncertain path to profitability and competition from rivals such as Google and Microsoft's OpenAI. "VC propaganda (flogs) unequivocal optimism without asking hard questions on product, user interface, distribution, or end markets.”

Founders and venture capitalists who flocked to artificial-intelligence startups are learning that turning the chatbot buzz into successful businesses is harder than it seems.

Almost a year into the boom ignited by the November launch of ChatGPT, some startups that epitomized the zeal for so-called generative AI are now navigating layoffs and reduced user interest. Investors are unsure whether the new crop of AI startups will be able to survive, especially as tech giants such as 

Microsoft

 and 

Alphabet

’s Google solidify their dominance over the technology. 

There was optimism in the venture world that commercial AI applications “would materialize at light speed,” said Mark Goldberg, a partner at Index Ventures, a longtime backer of AI startups including Cohere, which is building generative AI models for businesses. Now, “there is a shallow trough of disillusionment,” he said.

 

ChatGPT’s release reignited optimism in the startup world after months of industry malaise. Leading voices in Silicon Valley proclaimed that generative AI—the technology underpinning humanlike chatbots—would usher in a new technology era. Venture firms announced splashy new hires and boasted that they were devoting billions of dollars to back the still-unproven sector. 

Goldberg said the initial surge in user growth for ChatGPT—the fastest ever for a consumer app, UBS analysts said in February—led investors to overestimate the rate at which consumers would use tools powered by generative AI. Investors raced to back startups building these products before they had customers or revenue, leading to concern over a potentially overheated market.

Instead, usage has waned from peak levels. Monthly online visitors to ChatGPT decreased by 10% in both June and July after months of growth, according to data from analytics platform 

Similarweb

. An OpenAI spokeswoman said the Similarweb data didn’t track use of the ChatGPT app, which was launched in May.

Midjourney, an AI tool that can create images from user descriptions, saw monthly visits decline for three consecutive months ended in July, the data show. Synthesia, a text-to-video creator that raised $90 million in June, saw flat or declining website user growth for the past six months. 

Jasper, a generative AI writing tool for writers and businesses, saw declining user growth for four consecutive months ended in July. The company, which raised $125 million last fall, conducted a round of layoffs in July and cut revenue projections for this year, people familiar with the matter said.

Synthesia Chief Executive Officer Victor Riparbelli said the company recently completed its best-ever month for sales. The chief executives of Midjourney and Jasper didn’t return requests for comment. 

Much excitement for AI remains in the tech industry. Microsoft, Google and other established companies are investing heavily and rolling out new AI products, and business is booming for 

Nvidia

, whose chips are used to train AI models. Startup investors say the technology is still early and could one day mint a new generation of multibillion-dollar startups.

“We’ve moved from a moment of ‘How big can this be?’ to ‘How do we make it work?’ ” said Sunil Dhaliwal, a general partner at Amplify Partners. Dhaliwal sits on the board of Runway, which builds generative AI tools for content creators, and said about half of his firm’s recent investments have been in AI startups. 

So far, the AI buzz hasn’t been enough to stem a precipitous drop in total startup funding. Overall, U.S. venture funding plummeted by almost half in the second quarter, even as investment in generative AI startups—those focused on systems that produce humanlike text, images, and computer code—increased 65% to $3.3 billion, according to PitchBook data. 

Venture investors say they are still unsure what a winning business model looks like for a startup building new products around the technology. Many young businesses have yet to prove they can retain users and develop products that existing tech companies couldn’t easily mimic.

Training cutting-edge models can cost companies billions of dollars, thanks to the large volumes of data they need to ingest and analyze. Investors have become hesitant to bankroll such businesses, given the uncertain path to profitability and heavy competition from well-funded rivals such as Google and Microsoft-backed OpenAI. 

“We cannot sort of unleash AI and have no business model to pay for it,” said Frank Slootman, the chief executive of the data-storage company 

Snowflake

, on an August earnings call. A lot of executives, he said, “have characterized their foray into language models as experimental, exploratory, and they’re sort of trying to get their arms around how big a bread box is this.”

So far, the largest AI startups have sought much of their funding from tech giants that are able to provide computing power and other resources beneficial to training language models. Microsoft has agreed to invest more than $10 billion in OpenAI, The Wall Street Journal has reported. Its rival Anthropic, which has released a general-use chatbot, has drawn hundreds of millions of dollars in commitment from Google.

In June, Microsoft and the chip maker Nvidia helped Inflection AI raise $1.3 billion. The startup, co-founded by the LinkedIn co-founder Reid Hoffman, recently released a personal chatbot.

“There is a ton of VC propaganda” about the investment potential in AI, said John Luttig, an investor at Founders Fund, which has invested in DeepMind, an AI research lab purchased by Google in 2014, and in OpenAI. “It’s an unequivocal optimism without asking any of the hard questions on product, user interface, distribution, or end markets,” he said.

One AI startup that helped ignite the AI boom has come under scrutiny.

 

In August 2022, Stability AI released a viral text-to-image generator called Stable Diffusion and soon after raised $101 million from venture capitalists. The company held a launch party at San Francisco’s Exploratorium museum last October. 

Emad Mostaque, its chief executive, has since been accused of fraud. Mostaque has denied the allegations.

In July, Cyrus Hodes, one of Stability’s early co-founders, sued Mostaque and the company, alleging that he sold his shares in Stability for just $100 after Mostaque convinced him the company was essentially worthless. A Stability spokesperson said the suit was without merit.