Gen AI 2024 Early Stage VC Deal Value Drops 76 Percent From Q3 2023 Peak

VC investing in early stage generative AI has dropped precipitously from its peak a year ago as market saturation and big tech dominance raise questions about viable paths to profitability for startups. 

While there will undoubtedly continue to be growth in the space, the question is where that will come from, who will benefit and whether that leaves much room for young companies hoping to catch the wave, but who may be too late unless they are truly differentiated, which is becoming increasingly difficult to achieve. JL

Jacob Robbins reports in Pitchbook:

For two consecutive quarters, generative AI dealmaking at the earliest stages has declined, dropping 76% from its peak in Q3 2023 as wary investors reassess the initial flurry of capital into the space. Deal count has declined, with only 34 deals recorded in Q1, marking the lowest deal value and count since ChatGPT’s launch kicked off the gen AI frenzy in November 2022. One reason for wariness is the release of OpenAI’s GPT store, which offers free services like code completion, copywriting and document analysis. This has made (many) startups obsolete and forced investors to exercise greater caution.  “There was this huge rush, and a lot of people got funded who shouldn’t have. There's a ton of market saturation.” Investors are now chiefly concerned about profitability.

Even with AI, what goes up must eventually come down.

For two consecutive quarters, generative AI dealmaking at the earliest stages has declined, dropping 76% from its peak in Q3 2023 as wary investors sit back and reassess following the initial flurry of capital into the space.

VC deal value for pre-seed and seed-stage deals fell in Q1 2024 to $122.9 million, a far cry from Q3’s peak of $517.7 million. Deal count has also declined with only 34 deals recorded in Q1. In all, the latest quarter marks the lowest deal value and count since ChatGPT’s launch kicked off the generative AI frenzy in November 2022.

After flooding the space with VC dollars and with how quickly the tech is advancing, investors say it’s time to hit the brakes and watch how things play out.

“There was this huge rush, and a lot of people got funded who probably shouldn’t have,” said Matt Cohen, founder and managing partner at Ripple Ventures, which invests in pre-seed enterprise software startups.

One reason for wariness is the release of OpenAI’s GPT store, which offers free services like code completion, copywriting and document analysis. Effectively, Cohen said, this store has made a bevy of startups obsolete and forced investors to exercise greater caution when fielding potential investments.

With the recent release of OpenAI’s image and speech-heavy new model GPT-4o, some are concerned that this pattern will repeat itself. Stability AI, the maker of the popular text-to-image generator Stable Diffusion, has faced layoffs and its CEO’s resignation in recent months as it grapples with increasing competition.

“I ask them (startups) now, ‘don’t bullshit me,’ just tell me what’s the differentiator and if you’re using third-party tools,” Cohen said. “Right now, there’s just a ton of market saturation.”

Investors are chiefly concerned about how questions around profitability will shake out. “The ability to make money as an investor at this point is really hard. The bets have been made,” said Richard Dulude, co-founder and general partner at pre-seed and seed investor Underscore VC.

As AI excitement has begun to slow, the bar for investing in AI has continually risen, Dulude said. Investors, he said, are keen to move on from investing in the infrastructure layer. The latest Y Combinator demo day showcased a flurry of AI application startups, with music generation services, AI therapists and virtual teaching assistants taking center stage.

Still, Dulude told PitchBook that he isn’t hopeful for how everything will shake out.

“Everyone’s now experimenting with AI, and the question is are they going to be successful with it and is the ROI justified? I think in a lot of cases, people are going to be very disappointed.”