Artificial intelligence startups had a wild ride in 2023. Everyone and their grandmother tried out some sort of AI tool, startups in the space raised rounds at 2021 valuations, there were high-profile shutdowns, and then to close out the year, we had all the drama surrounding Sam Altman and OpenAI — plus New York Times’ lawsuit against the company.
With so much in the rearview mirror, it’s hard to predict what will happen with AI startups in 2024. But some people, like investors, make their living from shrewd bets, so TechCrunch+ recently asked more than 40 investors what they think AI investing could look like in 2024.
Most investors told TechCrunch+ that they expect the current swell of funding to continue but were optimistic that the industry is moving past its initial hype cycle and toward more durable businesses. They also think that 2024 could see the beginning of a second wave of AI startups that are more verticalized, that are focused on specific sectors, and that move away from building layers on top of technologies from companies like OpenAI and Google.
Lisa Wu, a partner at Norwest Venture Partners, expects opportunities in verticalized AI to be particularly attractive this year. She thinks that there could be lower risk in investing in these startups, as they won’t be as likely — or easily — replicated by legacy companies like Microsoft and Google.
“These are AI applications with deep underlying knowledge of end-user workflows and access to industry-specific training data to make employees and teams more productive,” Wu said. “For example, law firms that effectively leverage AI will be able to offer their services at lower cost, higher efficiency and higher odds of favorable outcomes in litigation.”
This was echoed by several other VCs who are looking for AI startups creating sector-specific solutions. Sarah Sclarsic, founding partner at early-stage firm Voyager Ventures, said she’s already backed a few companies using AI to address climate issues and hopes that there will be more startups entering the category in 2024. Grace Isford, a partner at Lux Capital, and Tasneem Dohadwala, founding partner of Excelestar Ventures, are similarly looking forward to seeing more AI applications in sectors like healthcare and biotech.
Enterprise adoption
Several investors, like Sarah Guo, founder of early-stage venture firm Conviction, predict 2024 will be a good year for enterprise-focused AI startups.
A mix of enterprise and consumer-focused startups launched in 2022 and 2023, but consumers have largely been at the forefront of adoption, with products like DALL-E and ChatGPT. According to Andrew Van Nest, a managing partner at Exceptional Capital, now that large corporations have had some time to hash out their AI strategies, or figured out where the highest ROI (return on investment) could be, there will be more adoption on the enterprise side.
“Customer buying behavior will mature, from initial interest and strong top-down mandate from leaders to ‘incorporate AI’ in 2023, to more sophisticated evaluation and broader operationalization of AI products in 2024,” Guo said.
Sailesh Ramakrishnan, managing partner and co-founder of Rocketship.vc, feels enterprise startups will be the big driver of AI investment in 2024 and that they will raise capital at much more reasonable valuations than their consumer predecessors.
Michael Marks, founding managing partner at Celesta Capital, agreed: “The overall acceleration of AI funding and the growth of that sector will bring with it the heating up of some key AI subsectors, namely, a move away from consumer and toward proprietary enterprise applications that can better handle the higher privacy and data security needs of more critical applications in sectors like finance, insurance, and defense.”
Some VCs also predict that with these changes, 2024 will be the year when the market stops referring to companies in the space as “AI companies,” and AI will simply be embedded into the DNA of software more generally.
“What’s an AI company?” Ed Sim, the founder and managing partner at Boldstart, asked. “Our belief is that it’s just a technology, like the internet or mobile, and it will be infused into every product where it makes sense and where customers are willing to pay for it.”
Big picture
If the market really does start to shift its focus in 2024 to AI startups with good business fundamentals and clear use cases, it could be a hard year for existing startups that were able to ride the funding wave last year.
Pradeep Tagare, head of investments at National Grid Partners, expects the divide between strong startups and those that raised money simply on hype to become more obvious this year. Elena Mazhuha, an investment director at Flyer One Ventures, agreed, adding that even some of the companies that raised last year will struggle if they try to raise again in 2024.
“Many companies will probably fail, including those who raised a round in 2023,” she said. “These will be the startups building ‘wrappers’ for big players’ algorithms, and companies that use AI to create a feature — not a product they can use every day.”
Shakeouts like these are part of the growing pains of pretty much any hot new industry, though. New technology really does need to go through a period of trial and error so that startups and investors can figure out where they should be spending their time and money. So while many VCs probably want to write off and forget the bets they made on AI over the past year, they remain bullish on the future. Kevin Lalande, managing director at Santé, and Don Butler, managing director at Thomvest Ventures, both think we’ve yet to see the best of AI.
“I think 2024 will be one of the most promising years for AI investing,” Butler said. “While there has been a bubble building in this space, 2024 and 2025 will be when some of the most interesting companies of the next generation will be started. If we look back at prior shifts in technology, such as the introduction of the iPhone, we’ve seen a pattern that has repeated: The most interesting companies that take advantage of such shifts are typically started one to three years after such technologies are introduced.”