38:112024: The Year the GPT Wrapper Myth Proved Wrong
From Y Combinator · Published Jul 20, 2025 · Watch on YouTube
TL;DR
2024 proved the "GPT wrapper myth" wrong: the ChatGPT store was a nothing burger, and many applications (Perplexity, Glean, Case Text, Harvey, Photo Room) built on top of foundation models are generating real revenue. Open-source models (Llama) reached top benchmarks, giving startups choice and breaking the idea that all value accrues to foundation model companies.
Key insights
- A year ago, enterprise pilots/POCs were viewed with cynicism; by 2024 they have turned into real revenue, and YC startups are selling into enterprises faster than ever.
- The consensus that all value would go to OpenAI (the "GPT wrapper" view) was proven false; breakout applications in consumer, enterprise, legal, and prosumer spaces are built by startups not by OpenAI
- Open source was the key driver: Meta's Llama (initially seen as 18 months behind) became the top foundation model in summer 2024, forcing commodity pricing and making product, sales, and user feedback
Want the full analysis - every claim cited to the second it was said?
This page only shows a teaser. Sign up to chat with the complete, cited breakdown of "2024: The Year the GPT Wrapper Myth Proved Wrong".