Elad Gil where AI markets have winners – and which are still wide open

– said the extraordinary individual VC investor Elad Gil on stage at TechCrunch Disrupt that artificial intelligence is one of the least predictable technology booms ever seen.

Gil is ranked among virtually every successful company of the last decade, including many of today’s leading artificial intelligence firms.

However, he believes that over the past yr, some AI markets appear to have been almost completely sewn together by the market leaders. There is a huge amount of AI available outside of those areas.

“I started investing in generative AI in 2021… at the time, not many people were paying that much attention to it,” Gil said. But he saw a huge jump in capabilities between GPT 2, launched in 2019and GPT 3 have been launched in 2021. “The step between 2 and 3 was so large that if you just extrapolated the scaling laws, i.e. the curve, you could really assume that it would be extremely important,” he said.

This convinced him to begin supporting early-stage startups creating products based on large language models. His bets ranged from core modelers reminiscent of OpenAI and Mistral to application firms reminiscent of Perplexity, Harvey, Character.ai, Decagon and Abridge. But throughout 2024 and most of 2025, the capabilities of the core models increased with each release, turning AI the wrong way up every few months.

“I was saying then that AI is the one market where the more I learn, the less I know. Typically, the more you learn about something, the better you know it, the easier you can predict the future, etc. But AI was just unclear. There’s just too much uncertainty. And I think there are still markets like that in AI,” he said.

However, he now also sees markets with clear winners. The most blatant example is the base models themselves. While tons of of models exist, and some countries reminiscent of South Korea are still working to develop sovereignty models through local firms, leaders have emerged. “Google, Anthropic, OpenAI, maybe xAI, maybe Meta, maybe Mistral – that’s a handful,” he predicts the winners.

He believes that after models, AI-powered coding has clear winners, which will make it difficult for latest entrants to catch up. Not only have major modelers moved into the company (Anthropic from Claude Code, OpenAI from Codex), but startup leaders like Cursor from Anysphere and Devin from Cognition (which acquired Windsurf) can be hard to beat. And hot on its tail are well-funded startups like Magic (which Gil called a possible “outlier”) and Poolside.

He believes that medical transcription is cornered, with Abridge leading the way and a few others like Ambiance being “important.”

According to him, customer support – which has been an early goal of each traditional AI and the next generation of AI agent startups – has hard-to-catch market leaders like its portfolio company Decagon. (Uplifted $131 million at a valuation of $1.5 billion in June.) OpenAI CEO Bret Taylor’s startup, Sierra, competes in this space. This is also an area where incumbents – Salesforce, Hubspot and many others – are adding AI offerings.

So which markets seem wide open? Gil says that financial tools (fintech), accounting, artificial intelligence security and “other markets that we know are fundamentally very interesting. We just don’t know who will do it.”

Ironically, rapid growth is not the signal it once was that a company would grow to be a hit. “The CEOs of every large company are basically telling their teams, hey, we have an edict. We need to develop our AI strategy,” Gil said. “These giant enterprises are willing to try things that they would never have tried two years ago, and that’s only because of AI.”

So latest AI markets can quickly capture big revenues from established enterprise customers, “but that doesn’t mean they’ll stick,” Gil notes.

Only once the market has passed the trial phase of the boom cycle will the startup and investors have the ability to see whether these revenues will sustain and grow. “There’s a false signal, and then there’s something that just goes wrong,” Gil said. He calls Harvey, a legal AI startup, one of the market leaders that “just works.” It raised three massive rounds in 2025, jumping from a $3 billion valuation to $5 billion 8 billion dollars, in just a few months.