The First Year of FundaAI

FundaAI started writing reports in late November 2024 and began formally covering Previews/Reviews on January 1 this year, marking exactly one year.

Over the past year, we’ve gone from being nobody to establishing an initial sphere of influence. If not for Michael Burry’s sudden appearance on Substack, I believe we would have been the fastest-growing Substack this year.

We’re especially grateful to the subscribers who discovered and supported us over the past year. Many subscribers have provided valuable suggestions through chat and DM, which we deeply cherish.

We especially thank Gavin Baker, DaRazor, and TMTBreakout, who discovered us early in our journey. Your reposts and recommendations have been tremendously helpful to us. We're also grateful to Ben Pouladian, HHao, Simeon McMillan, and many other friends who have supported us along the way. Many friends have also privately recommended us to their friends, for which we are deeply grateful. We're also grateful to our sell-side friends for summarizing our report content in their Daily Wraps. This is extremely helpful to us, and we also hope to have more exchanges and collaborations.

FundaAI’s reporting style has remained highly consistent. We insist on extensive primary research and work hard to connect the entire flow from Taiwan Supply Chain → Frontier Model Progress → AI Application/Software. For instance, we discovered early on that mid-training was driving storage demand, and how each model iteration and developments like O1/O3 and TPU impact the optical industry. We’ve also helped readers understand specific AI implementation scenarios through extensive coverage of AI software names.