According to TechCrunch, solo VC investor Elad Gil stated at TechCrunch Disrupt that AI has been one of the least predictable tech booms he’s ever seen, despite backing virtually every major AI company including OpenAI, Mistral, Perplexity, Harvey, Character.ai, Decagon, and Abridge. Gil began investing in generative AI in 2021 after observing the massive capability leap between GPT-2 and GPT-3, correctly predicting the technology’s importance. He now sees foundational models, AI-assisted coding, medical transcription, and customer support as markets with clear winners, while financial tooling, accounting, and AI security remain wide open. Gil specifically noted that Decagon raised $131 million at a $1.5 billion valuation in June, while Harvey leaped from $3 billion to $8 billion valuation in just a few months through three massive 2025 rounds. This creates a fascinating landscape of consolidation versus opportunity.
Sponsored content — provided for informational and promotional purposes.
The Consolidation Patterns Emerging in AI
What Gil is observing represents a classic technology adoption pattern playing out at hyperspeed. The markets that are consolidating fastest share several characteristics: they solve immediate, high-value business problems with clear ROI, integrate easily into existing workflows, and benefit from network effects. Foundational models represent the ultimate winner-take-all market due to the immense computational resources, data advantages, and talent concentration required. The leap from GPT-2 to GPT-3 that Gil referenced wasn’t just incremental improvement—it represented a phase change in capability that created insurmountable barriers for newcomers without similar resources.
The Enterprise Adoption Reality Check
Gil’s observation about enterprises being willing to try AI solutions they wouldn’t have considered two years ago reveals a critical dynamic. Large companies are in an AI arms race, driven by FOMO and board-level mandates to implement AI strategies. This creates what Gil calls “false signals”—rapid revenue growth that may not indicate sustainable business models. Enterprises are essentially running parallel experiments across multiple AI vendors, creating the appearance of market traction that may evaporate once they consolidate their vendor relationships. This phenomenon particularly affects horizontal AI solutions that lack deep domain expertise or proprietary data advantages.
Who Survives the Coming Shakeout
The markets Gil identifies as “wide open”—financial tooling, accounting, AI security—share an important characteristic: they require deep domain expertise that foundational model providers lack. AI security, for instance, isn’t just about detecting malicious prompts but understanding enterprise security postures, compliance requirements, and threat landscapes. Similarly, accounting AI requires understanding of tax codes, audit requirements, and financial regulations that evolve constantly. These domains represent the next wave of AI specialization where startups can build defensible moats through vertical expertise rather than just model performance. Companies like Decagon that have secured substantial funding in consolidating markets will need to demonstrate they can maintain their lead as incumbents like Salesforce and HubSpot accelerate their own AI offerings.
The Geographic Implications of AI Market Structure
Gil’s mention of South Korea developing sovereign models highlights an underappreciated aspect of AI market dynamics: geographic fragmentation. While foundational models may consolidate among a handful of U.S. and European players, we’re likely to see regional champions emerge in markets with strong data sovereignty requirements, language specificity, or regulatory barriers. Countries with large domestic markets and technological capabilities—China, India, Brazil, South Korea—will likely develop their own AI ecosystems. This creates opportunities for startups that can navigate these regional dynamics while global players focus on English-language and Western markets.
The Investment Strategy Shift Required
For investors and entrepreneurs, Gil’s analysis suggests a fundamental shift in approach. The era of betting on AI broadly is ending, replaced by the need for deep domain expertise and timing market entry correctly. Markets experiencing rapid consolidation require either first-mover advantage with substantial funding or a fundamentally different approach that incumbents can’t easily replicate. Meanwhile, the “wide open” markets require patience and specialization—they may not experience the explosive growth of early AI applications but offer more sustainable competitive advantages. The most successful AI companies will be those that understand whether they’re playing in a winner-take-most market requiring massive scale or a specialized domain where expertise trumps brute force.
