a16z’s Olivia Moore: Consumer AI Is Still Early, With ChatGPT Dominant and Big Categories Untouched

Moore covers consumer AI as a partner at Andreessen Horowitz. Her report examined the top 100 consumer AI apps, where ChatGPT is still the biggest player by a mile, according to TechCrunch. Suno and ElevenLabs, two smaller companies in music and voice, are demonstrating real staying power. The report also details half a dozen consumer categories that appear completely untouched by AI, including social apps, dating apps, marketplaces, retail, travel, finance and health. Moore said there are no entrants from those categories in the top 100 list, a result she called surprising and one she wants to see change over the next six months.

The interview comes at an uncertain moment for consumer AI. OpenAI has moved back toward enterprise this year, feeding pessimism about the overall revenue picture for consumer AI. Moore said she does not see that as a pivot because OpenAI is still launching a lot in consumer; she described it as an expansion. She said almost all AI revenue so far has come from subscriptions and token usage, which is much more concentrated on the enterprise and prosumer side.

Moore pointed to a State of Markets report showing that just 2.2% of U.S. households are paying for AI. She said she is not sure she wants that user number to rise. She is more interested in returning to a world where consumers are monetized in ways other than paying subscription dollars out of their own pockets. In Silicon Valley, she said, it is easy for high-income people with corporate cards to say they would rather pay for a product, but most people would prefer free access with ads and then decide whether to subscribe to remove the ads.

The marginal cost of AI services remains much higher than that of classic internet services such as Facebook or Google Search. Moore said there are signs of improvement. ChatGPT now has a Go plan at $8 a month, which she assumes runs on cheaper models. She said not every task requires frontier intelligence, which could mean cheaper models for some consumer use cases. More consumer companies are also building on top of open-source models, a shift she said founders have been discussing recently.

The users driving revenue today are doing coding and other technical automation, where they probably do need frontier intelligence, Moore said. As more people build for consumers, and the model is not the product, she expects lower-cost models to be used more often.

Moore also sees the lines between consumer and enterprise blurring. She wrote an article about a year ago called The Great Expansion. Before AI, she said, companies such as Canva were consumer-first and often took six or seven years to add teams or enterprise plans. Now companies including Gamma, ElevenLabs and Cursor start as consumer businesses and within 18 months become majority-enterprise businesses.

Almost all of what is thought of as consumer AI is really prosumer AI, Moore argued. On the revenue list, she said, basically three categories of power users are spending: product-building apps such as Lovable, Replit and Fal; product marketing and AI ad generators such as Higgsfield and HeyGen; and general work management tools such as Manus, Fireflies AI and Granola. Those may be paid for by consumers at the start, but she argued they are not consumer in the pre-AI sense.

Moore identified social apps, dating apps, marketplaces, retail, travel, finance and health as whitespace categories. There are no entrants from those areas in the top 100 list, and she said that is what she thinks the market needs to see over the next six months. Asked whether true consumer AI has yet appeared, she said it is definitely very early.