The year in review with Harry Stebbings, Jason Lemkin, and Rory O’Driscoll
The year that changed everything is almost over. In this special year-end episode, Harry Stebbings, Jason Lemkin, and Rory O’Driscoll break down the founders, funds, and companies that defined 2025—and make bold predictions for what’s coming in 2026.
From Dario Amodei earning “Founder of the Year” to why SpaceX will IPO before OpenAI, from the talent wars that “threw all conventions out the window” to why unemployment data could trigger a massive tech lash—this is the definitive look at where we’ve been and where we’re going.
Spoiler: If you’re a B2B company that didn’t reaccelerate with AI in 2025, you got an F minus. And if unemployment ticks up even 2-3 points next year for any reason? “Society will be terrified of AI.”
Top Takeaways
Founder of the Year: Dario Amodei and Anthropic’s “Sensible Play”
The unanimous choice wasn’t about predictions of mass unemployment (which the panel largely dismissed). It was about execution. Claude 3.5 and 4 changed everything—enabling vibe coding, powering Cursor, making Lovable and Replit actually work. “Without this, we have no vibe coding,” Jason said. “Many of these products existed for years. They did not work until Claude.”
But it’s not just product. Rory highlighted Anthropic’s business execution: “Growth rate faster than OpenAI, valuation convergence even after OpenAI’s potential new round at $800B… he’s cranked out the ‘I’m going to get profitable, I’m going to be sensible.’ If you own those stocks at the start of the year, this is the stock you’d feel most excited about.”
Honorable mentions went to Gwyn Shotwell at SpaceX (“navigating geopolitically challenging times with Elon’s brand this year… exceptional”), Alex Wang at Scale AI (“the best of the pseudo-hire playbook—moment in time, playing your cards right”), and Vlad at Robinhood (“Nine products doing over $100 million in revenue. JFC.”).
Fund of the Year: Index, Neo, and the Power of Sticking to Your Knitting. And a Big Nod to Hummingbird
Index wins on pure execution: Wiz (announced this year, $32B), Figma (seed), and Revolut ($75B). “Competent execution across multiple different exits,” Rory said. “Exits are the coin of the realm.”
But Jason gave his vote to Neo on aesthetics and hustle: first money into Cursor and Cognition. “24 months ago, he seemed to be having weird arguments with Gary Tan on Twitter. Now he’s literally hustling his way into cursor, hanging out at MIT, giving programming tests himself.”
- At seed, Hummingbird gets the nod—$100M fund returning $800M on one deal, including first investor in Lovable.
- At Series A, Benchmark: “Manus, Sierra, Firework, Lora, Cerebrus, LangChain. My word, that is one killer fund.”
- At growth, Thrive: OpenAI central position, Cursor, Databricks, Carvana, Revolut.
And the wildcard? Google—owning ~10% of SpaceX, ~14% of Anthropic, and the majority of Waymo. “Those three with double-digit ownerships, even if done in a circular way… corporate investor of the year.”
The Biggest Surprise: Talent Wars Threw Out Every Convention
Meta paying $100M for individual researchers. The $14B deal to acquire a company’s team while letting investors keep the husk. Nat and Daniel getting their venture fund acquired too. “The entire convention on why people buy companies, how employees get treated, how much a human can be paid to do a job got thrown out of the window in the space of 6 months,” Rory said.
In retrospect, the logic makes sense: “If you’re spending $73 billion on capex, spending $5 billion to make sure the people using the capex know what they’re doing probably makes sense.”
The other massive surprise? There’s no ceiling on venture outcomes anymore. Lovable going from $2B to $8B valuation in weeks. Anthropic at $100M being “the deal of the century from a risk-reward outcome.” “That there is no ceiling on venture… changes all the math and calculations.”
Breakout Company: Databricks Rode the AI Wave Like Nobody Else
“Every founder, every VC—our job was to ride this torrent of AI,” Jason explained. “If you’re not an LLM, your job this year was to ride the greatest wave of our lifetimes.”
Databricks did exactly that. In 2021, Databricks was presenting about managing cloud compute and data—nothing about AI. Now they’re at $5B revenue growing 55%, accelerating, while Snowflake’s AI revenue remains small. “If we had to have canon of the year for who rode the AI wave and utterly changed the trajectory of the company, it would be Databricks.”
Other breakouts: Eleven Labs (scaled to $400M ARR, standing off OpenAI competition—”$30 in 10 minutes versus $2,000 in two weeks”), and Open Evidence (capturing 500K of the 1M US doctors in one year by perfectly targeting medical search).
2026 IPO Predictions: SpaceX First, OpenAI Last
Jason’s order: SpaceX goes first (probably summer), Canva second (“the AI story is not perfect, but the numbers are there—do it now before looking obsolete”), Databricks in the back half (“it’s just the Series M”), Anthropic at year-end (“the simpler way to solve their capital needs”).
OpenAI? Probably mid-2027. “It’s burning too much.”
The challenge: taking a company public at $1 trillion is an unsolved banking problem. “What do you do with $950 billion in stock when many owners have been in for 10-15 years? Could be tricky.”
Best Performing Tech Stock 2026: The AI Tailwinds Will Accelerate
The top B2B public stocks (Palantir, Cloudflare, Monday, Shopify, CrowdStrike, Snowflake) have all figured out their AI tailwinds. Jason’s betting at least three stay in the top six next year: “I’m not a hedge fund guy, but the trends aren’t going to change much.”
But the contrarian bet? Salesforce. Trading at all-time low revenue multiples (5-5.5x), but when Agent Force actually works at scale: “Every single Salesforce customer is going to want that product. Would you like this agent to automatically go after all the customers your team didn’t follow up with? Sign me the f— up.”
The warning: Adobe is “the worst offender” for claiming $5B of “AI-influenced revenue” without net new bookings. “If you’re going to hide in your tower and say ‘we grew 8% but half was AI-influenced’—that’s table stakes.”
The Co-Pilot Failed. Here’s What Actually Works.
“2025’s biggest failure was the co-pilot,” Jason said. “We’re not going to put it in our base product. We’re not going to make the core product better. But we’re not going to provide so much value that you want to pay for it. That was the cynical whiteboard of January 2025 that failed up and down the B2B board.”
The counterexample? Notion. If they really hit 50% growth at $600M, the only explanation is customers paying $20/month instead of $10 for their AI offering. “I can’t use Notion without AI as a 10-year customer. It’s worth the 2x.” Very few B2B companies have earned that premium.
Buy One, Short One: Nvidia vs. Amazon
Rory’s pick: Buy Google, short Nvidia. “If the capex cycle continues and we’re all still spending, Nvidia does fine. But unless they need continued acceleration… Google has upside if the capex tide goes out.”
Jason’s counter: “You’re a fool to invest against Nvidia for 2026. There’s not enough time for competing GPUs and TPUs to get into data centers that don’t exist. 2027? I don’t know. But 2026—the bet’s still on.”
Amazon is the underperformer candidate—only up 2% this year, and “there’s just not enough time” for their AI strategy to materialize.
The Coming Tech Lash: “Society Will Be Terrified”
Here’s the prediction that should worry everyone: if unemployment ticks up 2-3 points for any reason—tariffs, business cycle, bank events—AI will take the blame.
“All the AI executives talk about there will be an impact on unemployment,” Rory noted. “Politics doesn’t work by looking at BLS data by subgroup. All they’ll know is: unemployment’s ticked up, and those guys building AI are admitting the crime. They’re like, ‘We did it. We put you all out of a job.'”
Jason agreed: “When there’s hard numbers at the top of the New York Times… it will become every dinner table conversation. Society will be terrified of AI.”
The dark humor? “If the robots are going to take over society, I want to be sure that I own the robots.”
Quotable Moments
Harry Stebbings
“What was the biggest surprise of 2025? The three days when Windsurf was being bought by Cognition and I interviewed Varun the day before—are they being bought? Are they not? OpenAI was buying them and then they’re not—that was a bizarre moment.”
“Nvidia investing $100 billion into OpenAI and the start of the circular deals… all of us giving up on caring about things like circular deals. We’ve all stopped caring. If it marks up my fund, great. It’s just the way it is now.”
“Which companies go public in 2026? SpaceX, Canva, Databricks, Anthropic—four IPOs backloaded next year. OpenAI probably should have gone first, but it’s burning too much.”
Jason Lemkin
“The product of the year is Claude 3.5 or 3.7. Without this, we have no vibe coding. Without this, we have no Lovable, no Replit, no Cursor that really works. Many of these products existed for years. They did not work until Claude.”
“What agents can do—we just started. Outside of parts of coding, we just started. We haven’t missed the boat as investors. We haven’t missed the boat.”
“If the robots are going to take over society, I want to be sure that I own the robots.”
Rory O’Driscoll
“The entire convention on why people buy companies, how employees get treated, how much a human can be paid to do a job got thrown out of the window in the space of 6 months. Blew me away.”
“Exits are the coin of the realm. The only fair way to measure things is: did you end up owning a lot and having big ownership in big exits?”
“If the AI kind of lift starts to recede, they’ll all come down because they’re way over their long-term valuation. Pennies in front of steam wheels. Someone’s going to look up and say, ‘Oh my god, 70x… I should panic.'”
This post is part of the ongoing 20VC x SaaStr collaboration. Happy holidays from the team—see you in 2026.
