2025 was a year of transition. Many things were set in motion that will bear fruit in the years to come.
At the behest of my son François, I found a surrogate to attempt to get him a baby brother. To accommodate a larger family, I listed my New York apartment and began looking for a larger place in Tribeca. Compelled by Alpha School pitch, I enrolled François there for kindergarten starting September 2026. I listed my house in Turks & Caicos and started the process of moving to Antigua. I also allocated more time to Midas, FJ’s latest incubation. I am sure not all of these will play out as expected, but I am excited for the new chapters to come.
The epic adventure of the year did not happen. I was supposed to train in Finse, Norway, ahead of an expedition to cross Greenland by snow kite. Sadly, my tennis elbow worsened to the point where I tore 80% of the tendon sidelining me from adventure travel and sports altogether.
The first few doctors I saw told me I needed immediate surgery and would most likely never play tennis again. As you can imagine, I did not accept this answer and went down the regenerative medicine route. Over the course of the year, I did cyto-rich PRP, exosomes with matrix, BPC-157 & TB-500 peptides, endless isometric rehab exercises, hyperbaric chamber sessions, and red-light therapy. My elbow fully healed and I was able to start playing padel again towards the end of the year. Regardless of what ails you, get alternative opinions and explore multiple avenues. Try to avoid surgery if possible.
While sidelined from adventure travel, I split my time between my three wonderful homes: Revelstoke, Turks & Caicos, and New York. I spent the first few months in Revelstoke, where we skied and snowmobiled up a storm.
It was incredibly special to properly ski with Fafa for the first time, until he promptly broke his leg.
I also loved my time in Turks where I even managed to kite and eFoil with Fafa and Angel.
New York continued to serve as my base for urban adventures. It remains the perfect haven for intellectual, professional, social, and artistic pursuits.
I continue to find it an exceptional place for the kids to grow up.
I finally went to visit my brother Olivier in Austin and discovered the city and its charms. Fafa was particularly impressed by Waymo’s robotaxis.
We also went to Olivier and Cristina’s super cute wedding in Rio.
I then visited family in Nice before renting a house in Saint-Tropez for two weeks in July to try it out. It was amazing. It’s interesting how a place so reputed for its parties can also be laid-back and remarkably kid-friendly. The Grindaverse, writ large, had a blast. It accommodated everyone’s needs and desires, especially since the house came with a padel court.
I skipped my traditional two weeks in Turks this past summer. Instead, I headed to Revelstoke for a longer five-week stint. I initially worried it would feel too long, but it was perfect. We had the most epic summer adventures there: car racing, jet skiing, hiking, camping, and more.
I then made my annual pilgrimage to Burning Man. This year I brought many members of the Grindaverse. We were extremely lucky with the weather and avoided getting stuck in the rain-related closures. In fact, the rained-out moments when we stayed inside felt healing and connecting. For a variety of reasons, the journeys veered into life-lesson territory rather than being purely enjoyable, and I left with very specific insights about what to do differently next year.
Fall in New York continued to be extraordinary. September and October remain among the city’s most vibrant months.
I also did a wonderful fall father–son bonding trip to Revelstoke, which allowed me to appreciate the town in the off-season for the first time. I worried it might be too rainy and miserable during shoulder season, but it turned out to be perfect, and a great complement to East Coast fall foliage.
As usual, I spent Christmas in Turks & Caicos and New Year’s in Revelstoke with the family. Amélie even skied for the first time and she’s only 22 months old!
Professionally, 2025 continued to be extraordinarily busy. I spent a significant amount of time helping scale Midas. We completed the first close of FJ Labs IV. We remained contrarian. We avoided over-hyped and overpriced AI companies and instead focused on network-effect businesses that use AI to become more efficient.
I spent considerable time reflecting on the impact of AI on marketplaces. I will present our updated perspective in an upcoming Playing with Unicorns episode.
Overall, FJ Labs continued to rock. The team remained at 19 people. We deployed $49 million. We made 174 start-up investments, 98 first-time investments and 76 follow-on investments.
Despite the macro environment, we were fortunate to achieve several successful exits, including the IPO of Klarna, the acquisition of Momence by Club Essentials, and the secondary sale of AgVend.
Since Jose and I started angel investing 25 years ago, we invested in 1,268 unique companies, had 425 exits (including partial exits), and currently hold 892 active unique company investments. We achieved realized returns of 26% IRR with an average multiple of 2.5x. In total, we deployed $735M.
We also hosted our bi-annual retreat in Saint-Tropez for the first time, and it was a huge success.
On the content front, I continued developing and refining Fabrice AI. I just launched an alpha version of Pitch Fabrice, where founders can pitch their startups and receive feedback. It is trained on many of the pitches we evaluate at FJ Labs, but remains a work in progress. With more at-bats, I hope it will provide actionable feedback to founders we do not have the opportunity to speak with directly, and perhaps even surface one or two exceptional companies for us to invest in.
I finally took the time to put my spiritual learnings into writing:
I appeared on multiple podcasts. I gave my most in depth interview ever on Open VC. Towards the end of the year I spent more time discussing the impact of AI on marketplaces on VC10X and Startupable.
2026 is going to be interesting. It’s hard to tell where we are in the AI bubble. It feels like we are in the later innings, but no one really knows. In Internet bubble terms, are we in 1995, 1998, or January 2000? I hope it lasts quite a bit longer. Like previous bubbles that helped lay down railroads or fiber across the United States, this one is building the infrastructure for the AI revolution to come. It could lay the foundation for a productivity boom that might allow us to grow out of a looming debt crisis driven by unsustainable government deficits and rising debt-to-GDP ratios.
As with most technological revolutions, I suspect it will take longer to impact the broader economy than optimists expect. In the startup world, we are early adopters, so we naturally assume the rest of the world will move just as quickly. But startups account for only a small fraction of GDP. Governments, large enterprises, and SMBs, who make up the bulk of the economy, move far more slowly. Broad adoption and deep implementation will take time. People tend to massively overestimate the short-term impact of new technologies while massively underestimating their long-term effects.
More pragmatically, I want the AI bubble to persist because I worry that if it implodes all startup funding will be impacted. Non-AI startups would find it even harder to raise than they do now. History suggests that when bubbles burst, the baby often gets thrown out with the bathwater. I suspect that our prudence and discipline in these bubbly times would not be rewarded beyond not having to take massive write offs and write downs on AI startups.
Beyond that, what excites me most is seeing how the many personal, professional, and geographic threads I set in motion this year begin to intertwine. 2025 was about positioning and planting seeds. 2026 will reveal which of them take root.
Happy New Year!
