VC explains why most consumer AI startups still lack staying power

Three years after the generative AI boom began, most AI startups still make money by selling to businesses rather than individual consumers.

While consumers quickly adopted general-purpose LLMs like ChatGPT, most specialized consumer GenAI applications have yet to catch on.

“A lot of the early AI applications around video, audio, and photography were really cool,” Chi-Hua Chien, co-founder and managing partner of Goodwater Capital, said on stage at TechCrunch’s StrictlyVC event in early December. “But then Sora and Nano Banana came along, and China open sourced the video model, and a lot of those opportunities disappeared.”

Chien compares some of these applications to a simple flashlight. Flashlight was initially popular as a third-party download after the iPhone was launched in 2008, but it was quickly integrated into iOS itself.

He argued that just as smartphone platforms took years for game-changing consumer apps to emerge, AI platforms will need a similar period of “stabilization” for permanent AI consumer products to flourish.

“I think we’re really at the point where mobile was in the 2009-2010 era,” Chien says. This was the era when large mobile-first consumer businesses like Uber and Airbnb were born.

There may be signs of stabilization as Google’s Gemini reaches technical parity with ChatGPT, Chien said.

Elizabeth Weil, founder and partner at Scribble Ventures, echoed Chien’s sentiments about GenAI’s early days, calling the current state of consumer AI applications at “an awkward midway point in the teenage years.”

What does a consumer AI startup need to grow? Maybe it’s a new device beyond smartphones.

“A device that you pick up 500 times a day and only see 3% to 5% of what you see is unlikely to end up being the one that introduces use cases that take full advantage of AI capabilities,” Chien said.

Weil agreed that smartphones may be too limiting to rethink consumer AI products because they are not primarily ambient. “I don’t think we’ll be developing for this in five years,” she said, pointing to an iPhone for the audience.

Startups and established technology companies are racing to develop new personal devices to replace smartphones.

Former OpenAI and Apple design chief Jony Ive is reportedly working on a pocket-sized device with “no screen.” Meta’s Ray-Ban smart glasses are controlled by a wristband that detects subtle gestures. Meanwhile, many startups are trying to introduce AI-powered pins, pendants, or rings in ways that differ from smartphones, with often disappointing results.

However, not all AI consumer products rely on new devices. Chien suggested that one such service could be a personal AI financial advisor customized to a user’s specific needs. Similarly, Weil foresees the proliferation of personalized, “always-on” tutors, with expert instruction delivered directly from smartphones.

Although Weil and Chien are excited about the potential of AI, they expressed skepticism that some stealth AI-based social networking startups will emerge. Chen said these companies are building networks of thousands of AI bots that interact with users’ content.

“It turns a social game into a single-player game, and I don’t know if it’s going to work,” he says. “The reason people enjoy social networking is because they know there’s a real human being on the other side.”