That day, when we heard ChatGPT was launching an agent product, I might have been the happiest person on earth. Many investors asked the same question: “What if ChatGPT builds the same thing?”
Honestly, we couldn’t really answer—because by the time people ask that, they already believe they know the answer: model companies are superior by nature, and anything they build must be better than what application companies build.
But when the product actually launched, everyone realized they aren’t the best by default. With stronger agentic architecture and better context engineering, application companies can outperform frontier labs, too. Now we finally have the chance to prove it. What a happy day.
더 보기
