As of today, April 2nd, I’ve officially left the Manifold team. I’m stepping down from my executive role, returning my unvested equity, and, yes, giving up my yearly salary of 10 million mana.
This is really, really hard to say. When I work on a thing, I can’t help but imbue it with a piece of my soul — so walking away now feels like tearing myself in two. But I think it’s the right choice, for me and for Manifold.
This market resolves to my subjective feeling in 2 years.
Why I’m leaving
Manifold feels trapped in local optima
The Manifold product is pretty stable now - we’ve succeeded at our original mission of “prediction markets that anyone can create.” Iterating on it now feels like pushing pixels around for minor gain. We spend our cycles tweaking onboarding flows and our algorithmic feed.
The community feels less and less like “my people”. Less of the weird mechanism geeks, more random speculation I don’t enjoy. (It’s a bit how I imagine Vitalik feels about the degen gamblers who now characterize crypto.)
… and I’m not excited for the next steps
For the last year or so, our priorities have been “grow DAU”, or “get to Series A”. These seem important for Manifold as a business, but are also deeply uninspiring to me.
Especially insidious is when our work captures people’s attention and gets them addicted.
Moreover, Manifold doesn’t feel on my path for personal growth. I thrive on zero-to-one innovation, and have flexed that muscle again and again over the course of Manifold — but I see much less of this in the future.
James, Stephen, and the rest of the team have a big pivot in store; it seems like a great bet, but not something that I personally feel drawn towards enacting. And so, now feels like a good time to step out.
“Prediction markets” feel insufficiently powerful
I got into this game entranced by dreams of futarchy, of superpowered social coordination. By the vision of building out an objectively correct way to make decisions. But even just within Manifold, we’ve failed to dogfood our markets; they do not serve as a meaningful guide our actions.
At the core: binary prediction markets do not provide enough bits of information to steer by. Operationalizing questions well is tricky & labor-intensive, and boils off much of the interesting nuance in a question.
Manifold markets today serve the niche of “a fun game to bet on the news”. That’s a good niche - but not where my heart is.
AI is coming
We started Manifold in December 2021. My intended playbook was something like: “build the next FAANG, get rich, give it all to charity.” My idols were founders like Bill Gates and SBF. It felt like Manifold had a reasonable shot at this.
Then, ChatGPT dropped. Since then, I’ve seen the tech landscape shift; my most talented friends jump ship. But our conception of Manifold has not updated in response.
Text, image, and code generation have all been breathtaking, but our day-to-day work on the site feels orthogonal to all this.
Even for prediction markets: I think it’s likely that betting and forecasting will be mostly up to AIs, not people.
Broadly: there’s not enough timeline left to try for the network effects of social media. Scaling up Manifold to hit the adoption of a Reddit or a Twitter, feels like fighting the last war.
My eternal gratitude
First and foremost: to the Manifold core team, thanks for spending your lives with me over the last couple years.
From a ragtag bunch of misfits, to mad scientist barons: I’m extremely proud of this team we’ve built. People often comment on the how fast we ship, how fun the site is, and how great is the community; this is all downstream of your work.
And it’s just been a blast, working together. From hackathons to offsites to just the daily grind, it doesn’t quite feel fair that work should be so much fun.
I’ll still be around, eg providing advice and guidance during our “board meetings”; but where Manifold goes from here is up to you.
To our funders: thanks for betting on us!
Every dollar feels like a gift, a validation of our work and ourselves. I’m sorry to not be seeing it through all the way myself; but Manifold remains in excellent hands.
(TBH, I’m most worried about what my family will think. It’s one thing to declare this to investors, another to our charitable donors, but to my aunts, uncles, cousins and parents…? We’ll see how that convo goes!)
To our users: thank you for loving Manifold as much as I have. The amount of passion we’ve seen from y’all is frankly, insane.
You’ve spent hundreds of thousands of dollars and millions of hours on our site. Writing questions from serious to inane; insider trading up the wazoo; building and sharing tools; hosting meetups and dating shows; making friends with one another.
I was floored the first time we got a physical gift: Melanie mailed us some stickers, a handmade thank-you and a folded crane. I’ve worked on over a dozen projects, but this was a first; it felt like a sign that we were doing something right. And then it happened again, and again, and again…!
And thanks for sharing Manifold with others! Sending a market to your friends is a precious extension of trust. It’s the primary way Manifold grows. Every time I spot a Manifold link in the wild, I feel a surge of warmth.
What’s next for me
Manifest
Manifest is still on! It was a ridiculously fun last year, and we’re hoping to do even better this time around. Come hang out with us~
In the longer term, I’ll be focusing on Manifund. Funding for public goods remains broken, a bajillion dollar bill left on the sidewalk. We’re going to pick it up.
How, exactly? …Still kind of tbd. It could be one of the things we’ve been piloting (regranting, impact markets, prize rounds), or something totally new. We have a bunch of ideas here; stay tuned, and reach out if you’d like to be involved.
A baby!
1,000🏅 Top traders
| # | Trader | Total profit |
|---|---|---|
| 1 | Ṁ96 | |
| 2 | Ṁ85 | |
| 3 | Ṁ75 | |
| 4 | Ṁ62 | |
| 5 | Ṁ60 |
People are also trading
Resolving this market to 5% regret.
First, taking stock of Manifold today:
Manifold the product feels pretty similar to how it did 2 years ago. There was a big push for sweepstakes, which didn't get the adoption needed to become a big success. (I feel a bit vindicated about leaving before that, though of course I'd have preferred to be wrong and rich).
Manifold the business has been left far behind the likes of Kalshi and Polymarket. In 2022 and 2023, we were neck-and-neck -- in pageviews and credibility, if never in revenue -- but now Manifold is a footnote, in comparison. This does make me feel that we were some combination of unlucky and just wrong.
Manifold the community, I'm much less in tune with. This seems natural, but I'm also nostalgic for the early spirit and users of Manifold. I'm hoping to catch some of this with Manifest 2026!
Manifold the team has almost wholly disintegrated
I was actually quite surprised when James left to start Codebuff; I'd never have guessed that he'd be the next to go. Fwiw if I'd been insider trading then, this market would have spiked - when I left, I'd felt better with two great founders at the helm
Then Sinclair, Inga, and David all departed for different reasons, very much changing what the team looked like.
Stephen and Ian are still, nobly, keeping the lights on (alongside a small part-time team!); but with MNX coming up, it's not clear how much of their attention they'll be able to spare on Manifold proper
So all in all, Manifold is in a fairly fragile state, in a much worse position than if I'd stayed on. Do I regret this?
Well, a bit. But my writeup from 2 years ago still feels pretty accurate. Prediction markets as a whole have scaled up to do billions in monthly volume, received all kinds of press, and become household names. But they still aren't doing much to help us make better decisions and navigate the world, the goal that would have been worth spending more of my life on.
In the two years since, I've scaled up Manifund, started Mox, helped with more Manifests, raised Ada, and done a bunch of other side quests. I'm probably still best known for my work on Manifold, which I remain extremely proud of; and at times I've flirted with the idea of coming back! But I'm fundamentally happy with the choice I made, knowing what I did then, and seeing where the world is at now.
@Austin Thanks for the writeup! Incredible how time passes.
Agree with most everything. I also have a similar level of regret for leaving.
And I was also just as surprised as you were that I left 😂 . I'd been planning to stick it out with Manifold, but got too into AI coding.
Manifold does seem pretty stagnant, though I'd give some credit for the website being snappier, with better UX, and vastly better/cheaper infra, than two years ago.
We never figured out the thing that would blow Manifold up, while the other prediction markets zoomed ahead, faster and farther than I expected.
Most of the reason for stagnation is that we're play money, which was the bet I wanted to make.
Manifold grew to ~100 daily active users in its first few months. We added streaks, among other features and saw retention skyrocket, growing 10x in a few more months. One year in, I assumed that we were just one product change away from another 10x growth, and be on the path of an app like Reddit, but it never came.
But Manifold is such a great test ground for fun mechanics, there could yet be such an idea that gets it back on the curve!
Most of all, I'm thankful to those who have kept Manifold running all this time. Thank you for your service. 🫡.
@Austin I bought 1000 mana of No on this market on the very first day showing great confidence in my assessment. How are you feeling after 1 year and 364 days?
@Eliza (FWIW it seems like the market description was a perfect description of the following 2 years of Manifold....)
@Eliza nope, if it was a perfect description, each piece would've been a tractable market 😂
Wow, what to say. When I first saw this market notification, I was sad, but as I read your bullet points, I found myself nodding along. Manifold is an amazing piece of infrastructure that I still appreciate checking every day, but at least for me, it’s less nonstop exciting, and many of my friends have moved on a bit. And I agree that the current implementation of futarchy has ran into some structural challenges. Maybe just regression to the mean, though. Spending time on Manifold circa 2022 was really fucking fun, and it’s still fun. If I were you, though, I don’t think I’d want to spend more time increasing DAUs.
I want to thank you for being so open and kind to me and Proof School. You didn’t need to care about what I had to say about anything but you listened, over and over, and supported my dreams of running super memorable events - the two official Proof/Manifold ones, the Market Manipulation Podcast episode, Manifest, and of course the amazing wedding bonfire. You inspired and excited a ton of people.
Manifold has been made far better by your compassion, hyperfriendly attitude, and big dreams. I’m so grateful, and excited for whatever you build next!
@Conflux haha thanks, I do think 2022 Manifold had a certain mood to it, "really fucking fun" sounds about right.
IDK how much credit I can accept; it always felt like the proofniks were the ridiculously kind and awesome ones (and you in particular, Conflux!) It's every creator's dream to find folks as passionate about their work as you and the rest of your school felt about Manifold, I never thought I'd feel like a rock star, coming onto a campus of math geeks. The wedding bonfire is the stuff of legend, thank you for that (among the many, many other hijinks).
@Austin Oh, I forgot to respond to this! I actually think you deserve quite a lot of credit. You’ve had a large role in shaping the culture of Manifold as a really open, welcoming, friendly place where cool things happen; I’m not sure all that would’ve happened with different management.
Wishing you poignant nostalgia and joy as you see Manifold pivot and grow. More importantly, new challenges, passion and growth to nourish you in the Manifund endeavor. I myself will regret not having delved deeper into Manifold before your departure— as your insightful and loving energy are manifest in this commentary farewell.
@jacksonpolack yeah "Will I return to working on Manifold full-time" is the kind of interesting important question to know about here! Off the cuff there's maybe like, a 20% chance over the next 3 years? Though some versions of this are weird (eg if Manifund takes off and then merges with/acquires Manifold, what does that count as?)