Conversation

2) NOT INVESTMENT ADVICE some names have been altered to preserve anonymity some facts have been altered to improve the heartiness of the laughs and the deepness of the sighs readers feel.

3) First, Sir Buys-A-Lot. S B-A-L heard about the FTT presale and was one of the first to buy. Something really struck me about that: they didn't seem hesitant, they didn't demand concessions, they didn't demand attention. They were excited.

4) I learned: (a) they're someone I always want supporting a project: they'll help when they can, and be as efficient as possible. (b) I used to teach interns to recognize the signs that you're getting picked of in a trade. This checked a lot of those boxes.

5) Sometimes, when someone invests, it's a partnership. This had that feeling! But sometimes it's a trade, and the other side is going to flip ASAP. SBAL wouldn't, but others would. And selling FTT was a trade. Maybe, in fact, a _bad_ trade.

6) I thought I believed in FTX more than anyone else. But apparently SBAL believed in it more than I, given that I sold and they bought! Maybe it was time for me to update my projections for FTX. Fair had increased (in fact we had made it increase!).

12:15 PM · Mar 6, 2021Twitter Web App

7) Second, The Right Arrow. TRA was the first employee to tell me they thought that we should take FTX and run with it, and make it the best we could. It hadn't launched yet, and everyone else was still uncertain and thought maybe we should just try to get _anything_ for it.

Replying to

8) Third, The Two Silhouettes. It takes an army to build an exchange; most companies have 100-1,000 developers. We had two. Sometimes, 2 > 1000. And that says a lot about the 2. (And, I mean, about the 1,000.) And it was more than that.

113

9) I didn't write up a complete spec for exactly how FTX would work. I was like, "hey wanna make a futures exchange?" And they were like "yeah sure". And then a few weeks later: "Hey Sam, here's an alpha version. How should positional margin requirements scale with size?"

10) Any good team of devs can implement a perfectly specified protocol. But taking "futures exchange" and inferring "we should have a risk engine, it should liquidate at some point, bigger positions have larger impact but it might not be linear, maybe constant*sqrt(size)"...

11) Fourth, The Late-Night Dancers. T L-N D were smart and knowledgeable and quick, which is great! And valuable. But they were more. They understood something: that businesses can succeed and fail based on whether they do whatever they have to do.

12) Crypto is 24/7, and sometimes things happen at 2am. Life can be brutal that way. You can either hope the night is never important, or you can sleep next to your phone, ready to sprint to the office on a moment's notice. Not everyone needs to! But someone does.

116

13) Fifth, The Relay Circle. You hear about the products and the tech and the posts. You even hear about back-end dev work. You don't hear about back-end corporate work.

14) It's easy to abstract it away. "Oh yeah, then you use your company [which apparently you have and it's the right one in the right place]" "Our KYC process [build and managed by people]" "Then you wire the money [from the bank account you got]..."

15) TRC is possibly the most crucial piece of the puuzzle. It's also brutal. For a while, TRC spent 6 hours per day in a physical bank branch, watching tellers manually type 20-digit memos and watching for typos.

16) TRC also has to deal with customers when they're wrong. And with lawyers, when they're wrong. (If you think lawyers are never wrong -- what happens when two lawyers disagree with each other?) And with service providers when they break. And employees, when they feel broken

17) Which is why it's a relay. It's hard and grinding and thankless and crucial. And so from time to time we pass the baton to each other. Some people thrive on it. Some feel its weight heavily. Sometimes those people are one and the same.

18) And, finally, sixth: Gottlieb Heileman, who needs no introduction.