This Week in Startups
00:42:50 · Lon Harris, Alex Wilhelm
[00:42:50] S0: Yeah. I feel like it's it's probably like a pendulum swing. Like, right now, things are just in this incredibly... Like, we figured out what a few of the key AI bottlenecks are, and everybody is just pouring into those niches. Like, okay, we need, like, faster, more efficient inference. Like, everybody try to do that. We know that's a problem. You know? So I think we'll probably wait a little while, and then we'll get back to, like, we have to do a little bit more legwork to figure out exactly what these bottlenecks and obstacles are to the next iteration of our AI agents because I'm assuming it's all gonna be AI from now on. [00:43:25] S3: I do just wanna add a caveat, though. I'm not trying to say that today's companies that are growing quickly are not valuable, aren't doing something cool. Of course. I just appreciate that there was a time when you could do the equivalent of messing about in the garage longer. And I think that's something that that we don't want to lose in the broader startup conversation.
00:43:25 · Alex Wilhelm, Lon Harris
[00:43:25] S3: I do just wanna add a caveat, though. I'm not trying to say that today's companies that are growing quickly are not valuable, aren't doing something cool. Of course. I just appreciate that there was a time when you could do the equivalent of messing about in the garage longer. And I think that's something that that we don't want to lose in the broader startup conversation. [00:43:42] S0: I'm reminded of, Clue, q l o o, who've been on This Week in AI a few times recently. Alex, Elias, a great founder. They predated AI. They were already trying to figure out recommendation engines and algorithms, and then AI came along and it's like, oh, this technology is even better. Like, let's just adopt that. And, like, that's been a fascinating process to sort of follow is, like, becoming this AI company because, obviously, AI is good at recommendation engines and put in a lot of data about what you like, and it spits out. Here's what you'll... Else you'll like. And, you know, it just reinvented that business, but he's been building that business all along. [00:44:20] S3: Yeah. No. I love it. Now my last clip from this interview comes when we're talking about the darkest moment in Stripe's history, and the story is not gonna go the way you think it is. So listen closely to what Patrick has to say here.
00:44:20 · Alex Wilhelm, Jason Calacanis
[00:44:20] S3: Yeah. No. I love it. Now my last clip from this interview comes when we're talking about the darkest moment in Stripe's history, and the story is not gonna go the way you think it is. So listen closely to what Patrick has to say here. [00:44:35] S2: Here's a common founder scenario. You're looking over your sales numbers at the end of the quarter, and, hey, your team, they're at risk of meeting their quota. This kind of thing happens all the time. You can't figure out what the specific problem is without taking a closer look at all of the data. Not long ago, that meant pulling it all in from across the entire organization, a time consuming ordeal. But now today, there's Rippling AI. Rippling is already built on your real time people and business data. That means Rippling AI pulls metrics from both their data and your CRM into one dashboard, and then you get to see how close you are to hitting quota. And you can figure out the root of the problem of why you're not getting there. Maybe it's the marketing. Maybe it's the targets they're going after. You can break down all of that revenue by region, by customer size. All of that important information is gonna be at your fingertips in seconds with Rippling AI. When you're running a company, you have important questions that need immediate answers. That means ASAP, as in right now, as in yesterday. So if you're ready to rule your business, head to rippling.ai/twist to get the only AI built to give you full visibility and take complex actions across your startup. What was the darkest moment?
00:44:35 · Jason Calacanis, Patrick Collison
[00:44:35] S2: Here's a common founder scenario. You're looking over your sales numbers at the end of the quarter, and, hey, your team, they're at risk of meeting their quota. This kind of thing happens all the time. You can't figure out what the specific problem is without taking a closer look at all of the data. Not long ago, that meant pulling it all in from across the entire organization, a time consuming ordeal. But now today, there's Rippling AI. Rippling is already built on your real time people and business data. That means Rippling AI pulls metrics from both their data and your CRM into one dashboard, and then you get to see how close you are to hitting quota. And you can figure out the root of the problem of why you're not getting there. Maybe it's the marketing. Maybe it's the targets they're going after. You can break down all of that revenue by region, by customer size. All of that important information is gonna be at your fingertips in seconds with Rippling AI. When you're running a company, you have important questions that need immediate answers. That means ASAP, as in right now, as in yesterday. So if you're ready to rule your business, head to rippling.ai/twist to get the only AI built to give you full visibility and take complex actions across your startup. What was the darkest moment? [00:45:48] S4: The moment where you're like, [00:45:50] S2: this is
00:45:50 · Patrick Collison
[00:45:50] S4: Yeah, not going to so there was a moment very early on, you know, now, sort of Stripe is replicated across multiple data centers, and invested kind of a huge amount in redundancy and availability and failover capacity and everything else, right? That was not the case, you know, four months in, call it. Maybe it was a little bit more, maybe it was six months in. Six And months in, we still ourselves, of course, took it very seriously, and the fact that our customers were relying upon us for the revenue was a big deal for us, right? And so my pager went off at 4AM one night, and the API was down. And again, this is not like a social media app or something, right? This is the conduit for our customers' revenue, so it was a big deal. And so I remember kind of immediately heading to the office, we lived very close to the office, I was at the office within five minutes, know, four or 5AM, and started to kind of try to figure out, diagnose what was going wrong. And it turned out that sort of an upstream switch at our data center had failed, nominally they had redundancy and multiple Internet connections, had been misconfigured, and so it was all down. And so we had multiple servers, but the underlying uplink connection was broken, and so upshot being API was down, and we couldn't do anything about it. And it was amazingly painful, right? Because the thing is down, you sort of want to be furiously typing, fixing, getting it back up, whatever, and you just couldn't have to wait on hold, we're getting these kind of forlorn updates in the data center every fifteen minutes. That's kind of their progress in fixing it. Eventually, and you know, this felt like one of the longest eight hour periods of my life. It was almost literally eight hours. They sort of got a new switch and they replaced it, the whole thing came back up. But that's not why it was the darkest moment. The darkest moment, of course, was because I was bracing myself for the deluge in complaints and dissatisfied customers and anger and vitriol and everything else. And, you know, 1PM rolled around, and as far as I could tell, nobody had cared or noticed.
00:47:52 · Jason Calacanis, Patrick Collison, Alex Wilhelm
[00:47:52] S2: Wow. Dodged a butt. [00:47:55] S4: Anyway, Kevan, some of that was good news, but, you know, as I reflected on it, I realized that it was actually kind of existentially bad news. [00:48:02] S3: Oh my gosh. What a what a brutal story. So a couple of learnings from that. One, early on, super hands on founders, very close to the office. They were not on a jet. They were not at a conference. They were building. I think there's a little... A lot of side quest founders can do these days. At that time, far fewer. But the the lack of of being yelled at as a very dark moment, thought, was a really good encapsulation of the founder journey because either way, he was gonna feel bad. Either he was gonna get yelled at and feel bad, but it turns out there's something that's even worse, Lawn, which is being ignored because no one cared.