1 00:00:00,080 --> 00:00:03,840 In today's episode, Frank and Andy sit down with special guest 2 00:00:03,840 --> 00:00:07,600 Max Sklar to delve into the world of artificial intelligence, data 3 00:00:07,600 --> 00:00:11,184 science, and data engineering. Max, a 4 00:00:11,184 --> 00:00:14,625 trailblazer in location data and machine learning, shares 5 00:00:14,625 --> 00:00:18,420 insights from his extensive experience at Foursquare, including his 6 00:00:18,420 --> 00:00:22,260 work on local search and bias correction. Get ready 7 00:00:22,260 --> 00:00:26,085 for a thought provoking discussion about groundbreaking projects, tech 8 00:00:26,085 --> 00:00:28,904 drama, and the ever evolving landscape of technology. 9 00:00:30,005 --> 00:00:33,190 So sit back, relax, and prepare to be amazed. 10 00:00:38,550 --> 00:00:42,225 Hello, and welcome to Data Driven, the Podcast, we explore the 11 00:00:42,225 --> 00:00:45,845 emergent fields of artificial intelligence, data science, data engineering, 12 00:00:46,625 --> 00:00:49,570 and all that good stuff. With me on this ever present, 13 00:00:51,649 --> 00:00:55,110 journey down the information superhighway is Andy Leonard. How's it going, Andy? 14 00:00:55,329 --> 00:00:59,025 Good, Frank. How are you doing? I'm doing alright. It's been a wild 24 15 00:00:59,025 --> 00:01:02,625 hours in, Maison Levin, or or 16 00:01:02,625 --> 00:01:05,690 Maison, Lavinia, depending on your, how you wanna pronounce it. 17 00:01:07,690 --> 00:01:11,130 At one point, we will share those crazy 18 00:01:11,130 --> 00:01:14,515 details, but it's been good most of the part. I I am, 19 00:01:14,895 --> 00:01:17,555 recovering from a, a bout of COVID, 20 00:01:19,455 --> 00:01:23,075 that hit the entire house. I think we all picked it up on the cruise. 21 00:01:23,720 --> 00:01:27,259 Again, there are worse places to pick up COVID, and there's worse 22 00:01:27,399 --> 00:01:31,159 things that could happen. I've I've sneezed quite a 23 00:01:31,159 --> 00:01:34,475 bit, I've coughed quite a bit, but The thing that's bugging me the most is 24 00:01:34,475 --> 00:01:36,795 this headache I had now for 48 straight hours. 25 00:01:38,314 --> 00:01:41,435 But it's okay, like I'm kind of living, learning to live with it, and I've 26 00:01:41,435 --> 00:01:45,159 actually given it a name, I call it Charlie. So, you 27 00:01:45,159 --> 00:01:48,759 know, Charlie is is is gonna be on the show today as 28 00:01:48,759 --> 00:01:52,585 well. I see. Yeah. You do sound a little different, but not much. 29 00:01:53,145 --> 00:01:56,985 Nah. That's why I need to clone my voice, and, maybe how they 30 00:01:56,985 --> 00:02:00,025 do the avatar stays. As I was saying in the virtual dream room, this is 31 00:02:00,025 --> 00:02:03,740 the 1st time I've I've felt fit for camera In about 32 00:02:03,740 --> 00:02:07,580 a week. Perhaps you can use some of that good 33 00:02:07,580 --> 00:02:11,340 AI voice modulation. There you go. We are 34 00:02:11,340 --> 00:02:14,985 pleased to present, with us, today is Max Sklar. 35 00:02:14,985 --> 00:02:18,825 Max is a, not only a fellow, data 36 00:02:18,825 --> 00:02:22,080 guy, but also a Fellow podcaster. 37 00:02:22,380 --> 00:02:26,140 Welcome to the show, Max. Thank you so much, Frank and Andy, 38 00:02:26,140 --> 00:02:28,860 for, for having me on. I'm looking forward to it. And, 39 00:02:30,220 --> 00:02:34,045 yeah. Thank you. Cool. So you you've been at you were at a 40 00:02:34,045 --> 00:02:37,245 company for a number of years. You were talking about this in the virtual dream 41 00:02:37,245 --> 00:02:41,050 room. You were at Foursquare for, oh, quite some time. Right. 42 00:02:41,050 --> 00:02:44,730 That's unusual, actually. 10 years at a at a company as as, you 43 00:02:44,730 --> 00:02:48,415 know, building software is pretty unusual. But I was really there For, I 44 00:02:48,415 --> 00:02:52,175 think, like, 3 different phases, so I kind of break it up into 3 45 00:02:52,175 --> 00:02:55,535 jobs. Interesting. So, you know, 46 00:02:55,535 --> 00:02:57,715 Foursquare is one of those companies that 47 00:03:00,490 --> 00:03:04,090 On when they broke up into 2 different parts, it kind of I didn't 48 00:03:04,090 --> 00:03:07,275 understand it. Like, this is just user. I'm not I'm not I know 1 I 49 00:03:07,275 --> 00:03:10,735 know that sounds terrible, but despite being my headache, Charlie talking, but, like, 50 00:03:13,125 --> 00:03:16,900 what, you know, obviously, that was a decision 51 00:03:16,900 --> 00:03:19,780 that was made in the marketing department, and I don't think people probably thought that 52 00:03:19,780 --> 00:03:23,115 through. But I love that location. Like, I I think, you know, like, I was 53 00:03:23,115 --> 00:03:26,895 the mayor of, like, 5 6 different places. It was such a cool 54 00:03:26,955 --> 00:03:30,620 creative, concept that if you go there long 55 00:03:30,620 --> 00:03:34,459 enough, you're the mayor, and some places would offer the mayor, like, a free coffee 56 00:03:34,459 --> 00:03:38,080 and a donut or something like that. Like, it was really clever. 57 00:03:38,605 --> 00:03:41,405 I know. I used to I mean, that was one of the things that got 58 00:03:41,405 --> 00:03:43,965 me really excited about it back in the day. And this is already 10 years 59 00:03:43,965 --> 00:03:47,680 ago. I assume you're talking about the, The app split back in 60 00:03:47,680 --> 00:03:51,520 2014, which I guess is That sounds about right. I I guess 61 00:03:51,520 --> 00:03:55,095 almost 10 years. I guess it's It it's actually been 10 years since we started 62 00:03:55,095 --> 00:03:58,614 working on that, which was also how it worked. But, 63 00:03:58,935 --> 00:04:02,780 no. I I mean, I actually remember a lot of what was going on 64 00:04:02,780 --> 00:04:06,620 at the time, and, we could talk about that a little more in a 65 00:04:06,620 --> 00:04:10,295 second. I I I I I'm happy to talk about that. 66 00:04:10,775 --> 00:04:14,295 I I think the the whole gamification thing was really 67 00:04:14,295 --> 00:04:18,055 exciting. And I really loved the idea that I can go into 68 00:04:18,055 --> 00:04:20,569 a place and be like, hey. I was here for 5 times. Oh, we're gonna 69 00:04:20,569 --> 00:04:24,169 give you some free, you know, free, you know, dessert or 70 00:04:24,169 --> 00:04:27,535 free chicken nuggets or whatever. Maybe these days, you know, kind of trying to watch 71 00:04:27,535 --> 00:04:30,595 what I eat a little bit more, it might not be as exciting. I I 72 00:04:30,815 --> 00:04:34,575 it was really sad that that didn't scale as a business. I think what 73 00:04:34,575 --> 00:04:38,240 would happen was, and and this I both experienced 74 00:04:38,240 --> 00:04:42,000 this personally. And also from talking to the leadership there, it was clear 75 00:04:42,000 --> 00:04:45,435 that they felt it it, You know, it it it wasn't a 76 00:04:45,435 --> 00:04:49,275 sustainable thing because you'd often go into a place and say, hey, you know, 77 00:04:49,275 --> 00:04:52,720 I checked in here all these times. I got all these rewards. You know, 78 00:04:53,039 --> 00:04:56,479 I'm supposed to get, like, a dollar off, or I'm supposed to get some 79 00:04:56,479 --> 00:05:00,285 some free dessert or something, which is It's always exciting 80 00:05:00,285 --> 00:05:02,925 to get a little reward, even if it's not, you know, even if it's just 81 00:05:02,925 --> 00:05:06,445 a dollar or whatever. It's always but I think what would end up happening is 82 00:05:06,445 --> 00:05:10,230 like the the man, the The management who worked there or the or the 83 00:05:10,230 --> 00:05:14,010 bartender or whatever would know about it. And, you know, 84 00:05:14,390 --> 00:05:17,270 once a few times that, you know, you got that thing. Oh, oh, let me 85 00:05:17,270 --> 00:05:19,735 go in the back and check. And then they come back in, like, 20 minutes, 86 00:05:19,735 --> 00:05:22,775 and it's like a whole big deal. They're like, maybe I don't wanna do this 87 00:05:22,775 --> 00:05:26,535 again. That's interesting, because, like, from my point of view, 88 00:05:26,535 --> 00:05:30,199 it was always very It was very fun, like, so I was mayor of, 89 00:05:30,440 --> 00:05:34,280 there used to be a ferry service between, in the 90 00:05:34,280 --> 00:05:37,705 suburbs of DC, between Poolesville, Maryland and 91 00:05:37,705 --> 00:05:41,465 Leesburg, Virginia. And, I would take that ferry a lot so 92 00:05:41,465 --> 00:05:45,130 much. So I was actually the mayor of the ferry. It didn't get 93 00:05:45,130 --> 00:05:48,410 me anything, like, in that case, but it was this kind of cool, like I 94 00:05:48,410 --> 00:05:52,215 I love to I I always imagined, like, there would be, You know, like 95 00:05:52,215 --> 00:05:55,255 the ferry could have some kind of a a little like TV screen that could 96 00:05:55,255 --> 00:05:59,015 like show who the mayor was. And then you can, you know, and and and 97 00:05:59,015 --> 00:06:02,669 then then then we can like kind of scale up those fights, But alas, 98 00:06:03,449 --> 00:06:06,430 society didn't go in that direction. Interestingly, 99 00:06:07,370 --> 00:06:11,115 Foursquare, what, like, I have been interested in that this like local 100 00:06:11,115 --> 00:06:14,555 local search space since since well before forest grabbing even my, you 101 00:06:14,555 --> 00:06:18,340 know, my my senior project as an undergrad, back, 102 00:06:18,340 --> 00:06:22,180 though, starting in 2005, was this, like, you know, this this website 103 00:06:22,180 --> 00:06:25,765 called sticky map, where people would post little, Icons all 104 00:06:25,765 --> 00:06:29,385 over Google Map. It's kind of inspired by Wikipedia. You can add messages. 105 00:06:29,445 --> 00:06:32,725 And I just thought it was pretty cool. People started marking up the, the campus, 106 00:06:32,725 --> 00:06:36,180 and then people started, you know, marking up. It was like, well, it's based on 107 00:06:36,180 --> 00:06:39,480 Google Maps API, so you could just mark up anything you want. And 108 00:06:40,660 --> 00:06:44,195 I think in that first project, I noticed All the problems 109 00:06:44,574 --> 00:06:48,255 that still exist with that kind of data today. Okay. What happens when you 110 00:06:48,255 --> 00:06:51,970 add duplicates? It's the first thing that happened. As, you know, As an undergrad, I 111 00:06:51,970 --> 00:06:54,530 was like, oh, I'm so excited. Let me show something like this. They're like, okay, 112 00:06:54,530 --> 00:06:58,210 let me create a marker. And I'm like, don't create that one. It's already been 113 00:06:58,210 --> 00:07:01,985 created. And then it's like, okay, now we have duplicates Right off the bat. And 114 00:07:01,985 --> 00:07:05,824 that is still something that, you know, Foursquare 115 00:07:05,824 --> 00:07:09,425 deals with and I'm sure Google deals with and Apple Maps deals with and then 116 00:07:09,425 --> 00:07:12,150 they they all deal with it. So, 117 00:07:14,290 --> 00:07:17,970 Yeah. It it it it I I think when I 118 00:07:17,970 --> 00:07:20,790 discovered Foursquare, you know, several years later, 119 00:07:22,115 --> 00:07:25,875 it was the it was innovative in several ways. It was, first of 120 00:07:25,875 --> 00:07:29,395 all, it was based on on mobile apps actually being at the place when you're 121 00:07:29,395 --> 00:07:33,210 commenting on it, which is exciting. It was the gamification of it. And the 122 00:07:33,210 --> 00:07:36,970 fact And and for those listening Yeah. For those listening, that was new. Like, 123 00:07:36,970 --> 00:07:40,015 that was brand new. So sorry. I didn't mean to cut you off, but, like, 124 00:07:40,015 --> 00:07:43,855 the context is important because All I had was was something on a on 125 00:07:43,855 --> 00:07:47,535 a website. You know, I I wasn't thinking of the I the iPhone didn't 126 00:07:47,535 --> 00:07:50,570 exist, for actually, pre Foursquare, 127 00:07:51,750 --> 00:07:55,590 there there was something called dodgeball, which was kind of the the 128 00:07:55,590 --> 00:07:59,384 predecessor to Foursquare, Which I wasn't involved with, but it but, it 129 00:07:59,384 --> 00:08:02,985 it was based on, like, you know, SMS kind of text messages where people were 130 00:08:02,985 --> 00:08:06,669 messaging on. If you remember, You know, those, you know, 131 00:08:06,669 --> 00:08:10,190 the the the what was it? The t nine texting where people were Oh, 132 00:08:10,190 --> 00:08:13,855 God. Yeah. Yeah. Yeah. But people would use that and and Foursquare and, 133 00:08:14,895 --> 00:08:18,035 Google bought that. And then the the the founders, 134 00:08:19,790 --> 00:08:23,630 Dennis Nautz went it went in and started Foursquare after that. 135 00:08:23,630 --> 00:08:27,150 So, very, there there's a very 136 00:08:27,150 --> 00:08:30,825 interesting history of kind of, like, Local search, city guide, 137 00:08:31,205 --> 00:08:35,044 and basically sort of social kinda local applications. We're 138 00:08:35,044 --> 00:08:38,830 we're very big at the time. Nowadays, I think we need to find a new 139 00:08:38,830 --> 00:08:42,429 take on it, but, that when I joined Forescord 140 00:08:42,429 --> 00:08:46,005 in 2011, it was very exciting. I'm glad you mentioned that because there was 141 00:08:46,005 --> 00:08:49,845 a a real so I started my career in New York City. Right? 142 00:08:49,845 --> 00:08:52,940 So I worked at Barnes and Noble com. So I was there in the fairly 143 00:08:52,940 --> 00:08:56,700 early days of Silicon Alley, and that was a huge thing. It was 144 00:08:56,700 --> 00:08:59,360 Microsoft. I think it was Microsoft had something called Sidewalk. 145 00:09:01,675 --> 00:09:05,355 And then there was there was maybe it was maybe it wasn't Microsoft. Maybe 146 00:09:05,355 --> 00:09:09,089 somebody else had something called sidewalk AOL had something, going and the fact 147 00:09:09,089 --> 00:09:12,290 you can't remember it, I think says it all. Right? Like, it was like in 148 00:09:12,290 --> 00:09:16,130 in, you know, anybody that could register a .com could spell and could 149 00:09:16,130 --> 00:09:19,824 spell HTML To get funding back in those days, a 150 00:09:19,824 --> 00:09:23,185 bit like the way the AI startup ecosystem is kinda 151 00:09:23,185 --> 00:09:26,860 today. But, no, there was I mean, 152 00:09:26,860 --> 00:09:30,400 people there was a time, young children 153 00:09:30,620 --> 00:09:34,305 out there, That when, you know, people saw the online 154 00:09:34,305 --> 00:09:37,925 world is slightly different than the real world, and they saw this as an opportunity. 155 00:09:38,545 --> 00:09:42,350 Right? But no, I'd like That takes me back. As soon as 156 00:09:42,350 --> 00:09:45,230 you said to, like, the local kind of connection guides, I was like, wow, it 157 00:09:45,230 --> 00:09:48,975 takes me back. Yeah. Yeah. Coming back to 158 00:09:48,975 --> 00:09:52,815 the app split, and I wasn't expecting to talk about this today. I'm sorry if 159 00:09:52,815 --> 00:09:56,390 it brings up If for what it's worth, I'm a former Windows For what it's 160 00:09:56,390 --> 00:09:59,990 worth, I'm a former Windows phone developer, and I wrote a book on silver light. 161 00:09:59,990 --> 00:10:03,130 So I understand the pain of working on 162 00:10:03,855 --> 00:10:07,535 It'll fade and I worked at barnes and noble.com. Right? So there's my trifecta of 163 00:10:07,535 --> 00:10:11,100 ill fated technology projects. Yeah. I I think it's 164 00:10:12,540 --> 00:10:16,060 A lot of technology companies, in order to become successful, actually have 165 00:10:16,060 --> 00:10:19,735 to go through big changes where people yell at them. 166 00:10:20,055 --> 00:10:23,095 And so it's like, how do you know whether you're breaking things or whether you're 167 00:10:23,095 --> 00:10:26,454 actually doing what you're supposed to do? And so that's kind of a Right. That's 168 00:10:26,454 --> 00:10:30,110 that's kind of a tough decision. I I think For Foursquare 169 00:10:30,250 --> 00:10:32,830 at the time, there was always, like, kind of a design, 170 00:10:34,410 --> 00:10:38,010 and product, like, tension between the people who 171 00:10:38,010 --> 00:10:41,764 wanted to be there As essentially like a Yelp replacement, kind of 172 00:10:41,764 --> 00:10:45,045 like a a local search city guide. And then there and then versus the people 173 00:10:45,045 --> 00:10:48,404 who were there for the the life logging, the check ins, the game. And I 174 00:10:48,404 --> 00:10:52,170 think, I think the separation could have been 175 00:10:52,170 --> 00:10:55,985 done. I mean, my personal thing is I think there could have been a 176 00:10:56,065 --> 00:10:59,825 Separation, I could think, could have been executed a little bit better. I 177 00:10:59,825 --> 00:11:03,665 think technically we did a good job. I think the apps that 178 00:11:03,665 --> 00:11:07,480 we ended up with were well designed, but I think, I think 179 00:11:07,480 --> 00:11:10,920 we needed to do is take into account how the how people were using the 180 00:11:10,920 --> 00:11:14,394 apps, at the time and not just, like, Kick all the 181 00:11:14,394 --> 00:11:18,074 people who are checking in, which was which was Foursquare's kind of bread and butter 182 00:11:18,074 --> 00:11:20,394 and just, like, kick them to the side with this other app. And then it's 183 00:11:20,394 --> 00:11:23,970 like, well, what is this? I'm calling this something different. Sorta. That was that was, 184 00:11:23,970 --> 00:11:26,949 I think, too much. But, again, 185 00:11:27,649 --> 00:11:31,185 there there were people saying it at the time. But 186 00:11:31,345 --> 00:11:35,025 The problem is, I I guess, you know, there whenever you make a 187 00:11:35,025 --> 00:11:38,805 change, there's always a great many people saying a great many things. So 188 00:11:40,730 --> 00:11:44,490 We could wax this we could wax nostalgic about Foursquare because I used to 189 00:11:44,650 --> 00:11:47,130 like, when I I was travelling a lot at the time when I worked I 190 00:11:47,130 --> 00:11:50,965 worked at Microsoft about 10 years ago. But I remember Sometimes I would actually 191 00:11:50,965 --> 00:11:54,645 choose different connecting airports, so I could get, like, the the 192 00:11:54,645 --> 00:11:58,404 jet that was it, the the jet set tag, like level up in my achievement 193 00:11:58,404 --> 00:12:02,180 there. Right. Which is kinda sad, but, we could 194 00:12:02,180 --> 00:12:05,380 we could wax nostalgic about that all day, and I would love to. But I 195 00:12:05,380 --> 00:12:09,105 think what What was the role of AI and ML in that 196 00:12:09,105 --> 00:12:12,704 space? Right? Because you're obviously collecting a lot of data. No. Like, I'm just curious, 197 00:12:12,704 --> 00:12:16,220 like, because how how was that being used? How was that, 198 00:12:17,579 --> 00:12:20,300 leverage. In in in a lot of ways, 199 00:12:20,939 --> 00:12:24,720 and, you know, many of which I I worked on over all those years. 200 00:12:25,464 --> 00:12:28,985 You know, one of them was, I mean, just, you know, search 201 00:12:28,985 --> 00:12:32,685 ranking in general, which, you know, Foursquare had a lot of ex Google engineers. 202 00:12:32,910 --> 00:12:36,050 So I learned directly from them so they they knew what to do. 203 00:12:36,510 --> 00:12:40,350 But search ranking, search ranking was a was a was a big project. 204 00:12:40,350 --> 00:12:44,185 This is kind of more of a statistical Problem where you were kinda trying 205 00:12:44,185 --> 00:12:47,945 to weight different attributes, like, is this related to the search the 206 00:12:47,945 --> 00:12:51,690 person put in? Is this related to how much do we wanna, You know, 207 00:12:52,410 --> 00:12:56,170 score things that are, you know, maybe someone's 208 00:12:56,170 --> 00:12:59,710 friends went to. So something like that. I think that 209 00:13:00,645 --> 00:13:03,765 The biggest well, I'll talk about the one that I think is the biggest deal 210 00:13:03,765 --> 00:13:06,245 and then the one that that I worked on the most. The one I think 211 00:13:06,245 --> 00:13:10,050 was the biggest deal for for Foursquare, which I did work on a little 212 00:13:10,050 --> 00:13:13,490 bit, is basically trying to figure out where 213 00:13:13,490 --> 00:13:17,170 someone is given. So we know where someone is given there that long 214 00:13:17,170 --> 00:13:20,665 from their phone. But it's like, what are they actually in a particular 215 00:13:20,665 --> 00:13:24,345 store? Like, are they in the Starbucks? Are they in the, you know, are they 216 00:13:24,345 --> 00:13:28,079 in the office, over there? Are they Are they just walking down 217 00:13:28,079 --> 00:13:31,920 the street? And so using the fact that people were were 218 00:13:31,920 --> 00:13:35,759 giving us training data, essentially, which was a big theme there, which is, you know, 219 00:13:35,759 --> 00:13:37,935 I think, something that, 220 00:13:40,015 --> 00:13:43,775 data scientists and data entrepreneurs need to need to look in closely, which is 221 00:13:43,775 --> 00:13:46,839 like, How can you get people to give you training data? Because it is really 222 00:13:46,839 --> 00:13:50,680 useful. So if you have people giving you where they are and 223 00:13:50,680 --> 00:13:53,320 then you could see the information from their phone, not just a lot long, but 224 00:13:53,320 --> 00:13:56,964 like what, You know, things like what Wi Fi's can you see? What, you know, 225 00:13:56,964 --> 00:14:00,404 other sensors from your phone, can you figure out where they are? And then there's 226 00:14:00,404 --> 00:14:04,220 the whole stop detection, problem. And so, Yep. 227 00:14:04,279 --> 00:14:08,120 Foursquare essentially can kinda figure out, you know, where you went day 228 00:14:08,120 --> 00:14:11,555 to day, and it's actually pretty good. Like, you know, if I Don't tell Foursquare 229 00:14:11,615 --> 00:14:15,215 where I went. Even today, I still look at it and, you know, it tells 230 00:14:15,215 --> 00:14:19,040 me what, what actual stores I was in. Now maybe there's a question of, 231 00:14:19,040 --> 00:14:22,720 you know, whether whether our apps are knowing too much about us, but 232 00:14:22,720 --> 00:14:25,139 that's that's a whole another question. But that was a very important, 233 00:14:27,454 --> 00:14:30,654 a resource for the business. And the one that I worked on the most, that 234 00:14:30,654 --> 00:14:34,415 was the most exciting though for me was the natural language 235 00:14:34,415 --> 00:14:38,230 processing Pipeline. And, of course, you know, text text 236 00:14:38,230 --> 00:14:42,010 data today is is is having such a a resurgence 237 00:14:42,070 --> 00:14:45,885 with, You know, I don't need to tell your audience with AI and all that. 238 00:14:45,885 --> 00:14:49,485 But, you know, it it it back then it was like, well, people were giving 239 00:14:49,485 --> 00:14:53,320 us, you know, several sentences called tips on Foursquare Venues, 240 00:14:53,320 --> 00:14:56,920 which would often be like, here's, you know, here's what you should do here. Here's 241 00:14:56,920 --> 00:15:00,475 what you should try. Here's a little review, something like that. People were 242 00:15:00,475 --> 00:15:03,995 leaving text with their check-in. So there's a bunch of texts, there's 243 00:15:03,995 --> 00:15:07,355 menus, things like that. So there's a bunch of texts in the system. And so 244 00:15:07,355 --> 00:15:11,100 it's like, what do we do with all of that? And, one of the things 245 00:15:11,100 --> 00:15:14,800 that we did was we pulled out key terms, you know, noun phrase detection. 246 00:15:14,940 --> 00:15:18,780 This is all kind of standard natural language 247 00:15:18,780 --> 00:15:21,454 processing. You know, not 248 00:15:22,875 --> 00:15:26,610 you know, you know, people often ask, oh, you know, I I think 249 00:15:26,769 --> 00:15:30,450 Nowadays, I'm often thinking everyone's thinking, oh, you were probably using, like, 250 00:15:30,450 --> 00:15:34,209 generative AI or something. No. It was just kind of standard NLP that had 251 00:15:34,209 --> 00:15:37,545 been developed over the last, you know, several decades. But, 252 00:15:38,245 --> 00:15:41,765 we did sentiment analysis and we used that to come up with the ratings for 253 00:15:41,765 --> 00:15:45,449 the venues, which which are used today. So you could tell 254 00:15:45,449 --> 00:15:48,329 how good something is. And, 255 00:15:49,610 --> 00:15:53,384 you know, I did some things that were a little bit 256 00:15:53,384 --> 00:15:57,144 more interesting that, you know, maybe get overlooked, but they're 257 00:15:57,144 --> 00:16:00,500 they're kind of unique to To to what we did there, which was sort of 258 00:16:00,500 --> 00:16:04,260 like timeliness and seasonality, which is so, like, if you check into 259 00:16:04,260 --> 00:16:08,005 a diner in the morning versus in the afternoon, It'll statistically 260 00:16:08,225 --> 00:16:11,905 give you different suggestions based on how timely it thinks each 261 00:16:11,905 --> 00:16:15,640 each suggestion is. Because with every Check-in where someone is doing 262 00:16:15,640 --> 00:16:19,080 something in real time. We have the timestamp. We know what time of day. We 263 00:16:19,080 --> 00:16:22,175 know what time of week. We know what time of year. And so it's kind 264 00:16:22,175 --> 00:16:26,014 of cool to to put that all together. And some of 265 00:16:26,014 --> 00:16:29,454 the some of the, some of the 266 00:16:29,454 --> 00:16:32,060 models, got pretty, 267 00:16:34,920 --> 00:16:37,320 you know, it was it was pretty neat how it all turned out. I think 268 00:16:37,320 --> 00:16:39,915 that one I you know, I still talk about that one is one of my 269 00:16:39,915 --> 00:16:42,954 That's my favorite one after being in the industry so long, even though it was 270 00:16:42,954 --> 00:16:46,475 like 10 years ago, because it was like, okay, we had training 271 00:16:46,475 --> 00:16:50,240 data again from on these tips where, You know, we 272 00:16:50,240 --> 00:16:54,000 could tell if the person liked the venue or disliked the venue and because 273 00:16:54,000 --> 00:16:56,320 they they told us, and they also left the tip. There were a lot of 274 00:16:56,320 --> 00:16:59,845 people who did that. So that just gave us training data for sentiment analysis. 275 00:17:00,145 --> 00:17:03,185 And at the time, I'm sure the tools now are much more sophisticated at the 276 00:17:03,185 --> 00:17:06,839 time when we use pretrained sentiment analysis tools, Didn't really 277 00:17:06,839 --> 00:17:09,880 work well on our data because it's just it was just a different kind of 278 00:17:09,880 --> 00:17:13,720 text. People wrote on Foursquare differently than they did on Twitter, for 279 00:17:13,720 --> 00:17:17,505 example. So, so that gave us training data. Give 280 00:17:17,505 --> 00:17:20,385 us training data for every language. And so that was nice. We got kind of, 281 00:17:20,385 --> 00:17:24,065 like, you know, 90 languages for free just by just by 282 00:17:24,065 --> 00:17:27,720 using that Strategy of Oh, wow. Using the data that people gave us. 283 00:17:27,720 --> 00:17:31,240 Probably not probably didn't work very well in all 90, but certainly worked 284 00:17:31,240 --> 00:17:35,045 well. Well, the beauty of it is It ends up working 285 00:17:35,045 --> 00:17:38,505 well so long as we have good language detection, it ends up working well 286 00:17:38,885 --> 00:17:42,430 in, any language that has any Particular, 287 00:17:43,610 --> 00:17:47,390 you know, any particular popularity in Foursquare. 288 00:17:47,770 --> 00:17:51,565 So for example, if, the Turkish was very popular. Okay. Well, that 289 00:17:51,565 --> 00:17:54,945 means we have a lot of Turkish training data. That means that the 290 00:17:55,645 --> 00:17:59,120 the model, which trains monthly, is Is going to use all that training data. That 291 00:17:59,120 --> 00:18:02,960 means it's going to work very well. And so, and 292 00:18:02,960 --> 00:18:06,660 so that the fact that the models were always regenerating 293 00:18:07,475 --> 00:18:10,294 And they were always regenerating based on the latest data 294 00:18:11,235 --> 00:18:14,934 was was really cool because oftentimes you think these think about 295 00:18:16,680 --> 00:18:20,520 ML teams kind of building a model, and then they kind of throw 296 00:18:20,520 --> 00:18:24,335 it over a wall. They they productionize it. And then you have to 297 00:18:24,335 --> 00:18:27,054 work on the next one, but you have to you have to do some work. 298 00:18:27,054 --> 00:18:30,174 It's not automated, you know. So it's like, well, this is this is gonna start 299 00:18:30,174 --> 00:18:33,620 going downhill If we don't, if we don't interact. And the fact that we were 300 00:18:33,620 --> 00:18:37,300 able to set it up where it was just constantly getting smarter was, 301 00:18:37,540 --> 00:18:41,285 was pretty neat. So MLOps and pipelines before they were called 302 00:18:41,285 --> 00:18:45,125 MLOps. Well, they might have been called pipelines, but yeah. Interesting. Yeah. 303 00:18:45,125 --> 00:18:48,880 Pipeline was a big Big big key phrase. So what 304 00:18:48,880 --> 00:18:51,520 what did the data what did the back it because like one of the one 305 00:18:51,520 --> 00:18:54,399 of the jokes that we have, and in fact, it's a domain name that I 306 00:18:54,399 --> 00:18:58,225 registered. 1st, you get the data, is a phrase that 307 00:18:58,225 --> 00:19:01,585 a lot of data scientists will often use, much to the chagrin of a lot 308 00:19:01,585 --> 00:19:04,929 of data engineers, because a lot of data, 309 00:19:05,470 --> 00:19:08,350 you have to get the data in a certain way to to format it and 310 00:19:08,350 --> 00:19:12,085 and and to get it trained. And if you go to first you get 311 00:19:12,085 --> 00:19:14,985 the data.com, it should redirect you to our website, 312 00:19:15,685 --> 00:19:19,480 hopefully. God only knows if it works. 1st, I think 313 00:19:19,800 --> 00:19:21,980 yeah. I'm gonna I'm gonna try to get the data.com. 314 00:19:23,640 --> 00:19:27,385 I'm shattered to think that okay. Good. It does work. Okay. DNS 315 00:19:27,385 --> 00:19:30,284 and me have a long history. Yes. It's going back. 316 00:19:30,985 --> 00:19:33,865 Good. I I it's always good to start off a week with a win with 317 00:19:33,865 --> 00:19:37,710 DNS. What did it what did the 318 00:19:37,790 --> 00:19:41,630 because I'm curious, like, Foursquare was one of those early, kind 319 00:19:41,630 --> 00:19:45,475 of mobile first, kind of success stories. I'm 320 00:19:45,475 --> 00:19:49,315 always curious, what did the back end data platform look like? Right? 321 00:19:49,315 --> 00:19:52,755 Because, and again, going back 10 years, I 322 00:19:52,755 --> 00:19:55,930 mean, I mean, did you use what was the name of the, 323 00:19:57,690 --> 00:20:00,810 gosh. Can't think of the name of the platform, but what sorts of technologies did 324 00:20:00,810 --> 00:20:04,535 you did you guys use? Yeah. I'm I mean, I'm sure 325 00:20:04,535 --> 00:20:08,375 it works similar today in in at at Foursquare. Mhmm. 326 00:20:08,375 --> 00:20:11,980 Well, we were using data pipe I assume. But, yeah, 327 00:20:11,980 --> 00:20:15,780 if I remember correctly, we had, you know, our transactional database, our Mongo 328 00:20:15,780 --> 00:20:19,595 database that was sort of like, Every once in a 329 00:20:19,595 --> 00:20:23,115 while. And so that was kinda like the baseline. And then there'd be a series 330 00:20:23,115 --> 00:20:26,875 of jobs that, like, you know, built it up, that that 331 00:20:26,875 --> 00:20:30,530 that kind of Calculated things off of that, and that 332 00:20:30,530 --> 00:20:34,130 would, in the at the end of that pipeline, you know, release 333 00:20:34,130 --> 00:20:37,750 a, a dataset that would then be kind of, 334 00:20:38,515 --> 00:20:42,295 Automatically, deployed and then read by, 335 00:20:42,995 --> 00:20:46,759 read by the server in real time. So, if I can think of, like, the 336 00:20:46,759 --> 00:20:50,120 technologies, I think the, the pipeline technology, the 337 00:20:50,120 --> 00:20:53,804 pipeline, what was it? It was 338 00:20:53,804 --> 00:20:57,085 like Luigi. It was written in pipe. Python. I don't know if that's too interesting. 339 00:20:57,085 --> 00:20:59,585 There's a lot of different ones you could use these days. 340 00:21:02,730 --> 00:21:05,610 It's an interesting question of, like, you know, which one do you use? I, 341 00:21:07,210 --> 00:21:09,815 it's It's probably, 342 00:21:10,995 --> 00:21:14,835 you know, from from my point of view, it's always like, well, the company kinda 343 00:21:14,835 --> 00:21:17,095 chooses it. You don't really have much of a say. 344 00:21:19,020 --> 00:21:22,860 And then then it's like, well, well, how do I know how to compare? But 345 00:21:22,860 --> 00:21:26,144 let's see. Like, you know, we were using MapReduce jobs. We're using Hadoop At the 346 00:21:26,144 --> 00:21:29,825 time, I think scalding, was was one that's that's maybe kind of out of fashion 347 00:21:29,825 --> 00:21:33,585 now. That was a, a scala based framework for for 348 00:21:33,585 --> 00:21:37,190 some of these jobs which were, which which 349 00:21:37,190 --> 00:21:40,950 was based on abstract algebra. So it's actually pretty cool. I wish it 350 00:21:40,950 --> 00:21:44,565 was. It was kinda hard to to reason about sometimes if you 351 00:21:45,445 --> 00:21:49,144 It kind of went too far, to the side of, okay, 352 00:21:49,205 --> 00:21:52,904 you know, I love abstract algebra, but I don't want everybody 353 00:21:53,205 --> 00:21:56,510 who I don't I don't want that to be a barrier to entry for people 354 00:21:56,510 --> 00:22:00,350 who are we're working on this. But, I'm 355 00:22:00,350 --> 00:22:03,465 just trying to remember, like, some of the, You know, some of the some of 356 00:22:03,465 --> 00:22:06,825 the tech bud buzzwords. But if you have any specific questions, maybe they'll jog jog 357 00:22:06,825 --> 00:22:09,945 my memory. I don't know. Like, one of the things that was popular about that 358 00:22:09,945 --> 00:22:13,340 time was, HBase. Oh, 359 00:22:13,340 --> 00:22:16,240 yes. I, were we using interest? I think we were using, 360 00:22:18,140 --> 00:22:21,495 Yeah, I remember that Term, but I I know. I know. I was as you 361 00:22:21,495 --> 00:22:24,455 were talking, I'm like using it or if we wanted to use it. It was 362 00:22:24,455 --> 00:22:27,815 one of those 2. Now from that if memory serves, I think Facebook is the 363 00:22:27,815 --> 00:22:31,310 one who pioneered h base because it was really it was a right once read 364 00:22:31,310 --> 00:22:34,530 many thing, and basically, the last one in when, 365 00:22:34,910 --> 00:22:38,555 last last I can't talk, sorry. Last one wins. 366 00:22:38,555 --> 00:22:42,155 Let Andy help might help me out if with the 367 00:22:42,155 --> 00:22:45,835 technical term for that last one. Last one wins. What's the 368 00:22:46,155 --> 00:22:49,980 Oh, yeah. So right. I remember they were called h files, so it must have 369 00:22:49,980 --> 00:22:53,820 been yes. It must have been that. Yeah. Yeah. Yeah. That was one 370 00:22:53,820 --> 00:22:57,335 of those sorry, Andy. Go ahead. That's okay. You were I was thinking, 371 00:22:57,875 --> 00:23:01,715 you know, ChipLogic last in first out. Yeah. 372 00:23:01,715 --> 00:23:04,995 Yeah. Something like that. Yeah. Know if that's what you were after or not. Last 373 00:23:04,995 --> 00:23:08,730 one wins. Last one That was their concurrency strategy. 374 00:23:08,870 --> 00:23:12,650 That's, I know there's a better term for that, but again, 375 00:23:13,830 --> 00:23:17,674 it's a Monday and, I have 376 00:23:17,674 --> 00:23:21,355 a headache. But no, it's it's it's it's 377 00:23:21,355 --> 00:23:24,909 fascinating to kind of Almost like technology 378 00:23:25,049 --> 00:23:28,730 archaeology. Like what worth the big projects 379 00:23:28,730 --> 00:23:32,330 that were popular at the time. Right? You know, and it's 380 00:23:32,330 --> 00:23:36,065 just, And it's scary to think that, you know, we're talking 10 years 381 00:23:36,065 --> 00:23:39,825 ago. I mean, I mean, you Not even though. A lot of this stuff was 382 00:23:39,985 --> 00:23:43,360 I mean, a lot of this stuff is probably still in place at Foursquare today. 383 00:23:43,980 --> 00:23:47,820 Yeah. I mean, what's interesting to me is you you mentioned a lot of 384 00:23:47,820 --> 00:23:51,605 the NLP, techniques that, You know, for lack 385 00:23:51,605 --> 00:23:54,965 of better term, people would consider legacy now, right? Because they're pre 386 00:23:54,965 --> 00:23:58,325 transformers, right? They're pre GPT, right? Sentiment 387 00:23:58,325 --> 00:24:01,690 analysis, a lot of, you know, I I speak with a lot of people with 388 00:24:01,690 --> 00:24:05,529 varying degrees of technology skills, and they 389 00:24:05,529 --> 00:24:09,230 assume that this field of research didn't exist prior to 390 00:24:09,289 --> 00:24:12,875 last year. And, very 391 00:24:12,875 --> 00:24:16,414 much not the case. It's just that radically changed about a year ago. 392 00:24:17,034 --> 00:24:19,950 Right. I mean and and this is something that I'm trying to figure out how 393 00:24:19,950 --> 00:24:23,790 to do, which, I I might not be alone. It's like, okay. I did 394 00:24:23,790 --> 00:24:26,895 all these things. How do I reinvent myself now in this new world? 395 00:24:27,295 --> 00:24:31,055 And, you know, once you realize it could be exciting thing, then 396 00:24:31,055 --> 00:24:34,255 it's maybe not so much of a drag, you know, because there's there's so many 397 00:24:34,255 --> 00:24:38,029 opportunities out there. But it's like, but I can't be 398 00:24:38,029 --> 00:24:41,870 the only one out there who's struggling with this being like, okay, 399 00:24:41,870 --> 00:24:44,475 wow, I've got a, You know, I I've gotta, 400 00:24:45,495 --> 00:24:49,015 you know, work or at least do projects for companies that 401 00:24:49,015 --> 00:24:52,649 are at the cutting edge here in order to, In order to be, 402 00:24:53,610 --> 00:24:57,370 you know, in order to be at the forefront. Yeah. It's funny, like, you miss, 403 00:24:57,370 --> 00:25:01,025 like, I was, You know, offline for I tried to be offline, but for the 404 00:25:01,025 --> 00:25:04,645 better part of a week and for vacation. 405 00:25:04,945 --> 00:25:08,679 And like During that week, AMD announces that they are 406 00:25:08,679 --> 00:25:11,820 producing their own, GPU LLM 407 00:25:12,039 --> 00:25:15,794 type hardware. Gemini comes out and all 408 00:25:15,794 --> 00:25:19,315 these other innovations that come out, and I'm like, I feel, like, hopelessly behind 409 00:25:19,315 --> 00:25:23,030 now. I'm being offline for a week. Yeah. 410 00:25:23,030 --> 00:25:26,790 Yeah. It's I mean, I I guess the 411 00:25:26,790 --> 00:25:30,095 only, consolation there is everyone's dealing with that. 412 00:25:30,335 --> 00:25:33,715 Right. You know? Yeah. Right. And Kinda like impostor 413 00:25:33,775 --> 00:25:37,455 syndrome. Right? Yeah. Yeah. I think I think the 414 00:25:37,455 --> 00:25:41,070 question is, especially in this new world of generative 415 00:25:41,070 --> 00:25:44,670 AI. And and the question I'm asking I don't necessarily have answers. But it's 416 00:25:44,670 --> 00:25:48,435 like, how do you so You wanna jump in the stream 417 00:25:48,435 --> 00:25:52,215 and get all the latest stuff, but you also want to leverage your experience 418 00:25:52,515 --> 00:25:56,160 and understanding. Cannot be leveraged. And 419 00:25:56,160 --> 00:25:59,960 so what's the best way to to, you know, what 420 00:26:00,080 --> 00:26:03,040 what's the best way to balance that? I think that's something that I would like 421 00:26:03,040 --> 00:26:06,495 to see more people asking. And I would like guidance on this. I know I'm 422 00:26:06,495 --> 00:26:10,095 the guest. I'm supposed to say what I know, but No. But try now. You 423 00:26:10,095 --> 00:26:13,180 know, Mads, the, the thing is a lot of the stuff that's 424 00:26:13,660 --> 00:26:17,280 That's new. I'm doing the air quotes here for people who are listening. 425 00:26:17,580 --> 00:26:21,265 A lot of things that are new are really coming out of tech That was 426 00:26:21,265 --> 00:26:24,784 developed. The math was developed, for instance, in the late sixties, seventies, 427 00:26:24,784 --> 00:26:28,404 eighties, nineties. So a lot of that is just being reapplied 428 00:26:29,090 --> 00:26:32,790 Back when the math was developed and the theorems and and such, 429 00:26:32,930 --> 00:26:36,290 we didn't have machines fast enough to do it or at least do it 430 00:26:36,290 --> 00:26:39,835 usefully. So I wouldn't feel bad at all about, 431 00:26:40,135 --> 00:26:43,975 you know, having a bunch of, a bunch of experience that seems dated 432 00:26:43,975 --> 00:26:47,750 right now because A couple of weeks to a couple of months. That might 433 00:26:47,750 --> 00:26:51,510 be the new shiny. Right. That's true. When I went back to 434 00:26:51,510 --> 00:26:55,165 a a university computer science program, You know, they're still studying 435 00:26:55,165 --> 00:26:58,625 data structures and algorithms. It's still very relevant. 436 00:26:58,765 --> 00:27:02,525 And, you know, I think a lot of outsiders think, oh, everything's 437 00:27:02,525 --> 00:27:06,269 gonna Turnover in, in a year and a lot of things 438 00:27:06,269 --> 00:27:09,490 do. But there are also a lot of kind of like universal, 439 00:27:10,830 --> 00:27:14,215 kind of, there's a lot of universal Theory that's, good to know 440 00:27:14,215 --> 00:27:17,755 about. Sure. The fundamentals don't change that often. 441 00:27:18,294 --> 00:27:21,840 Nope. And it's a lot of reapplications. I see a lot of people 442 00:27:21,840 --> 00:27:25,600 reapplying stuff 2 or 3 times. I mean, I've been I've been 443 00:27:25,600 --> 00:27:29,425 around computing since 1975. So I've seen kinda like these meta 444 00:27:29,425 --> 00:27:33,205 patterns flow, you know, through several generations, 445 00:27:33,265 --> 00:27:36,950 and they kinda keep just resurfacing. One of one of the 446 00:27:36,950 --> 00:27:40,010 interesting ones is, like, the, well, 447 00:27:40,790 --> 00:27:44,410 both the chatbot and the text based interface versus the, 448 00:27:45,205 --> 00:27:48,905 Graphical interface seems like we keep going back and forth. You know, 449 00:27:48,965 --> 00:27:52,720 I I remember chatbots back in the, you know, AOL days. 450 00:27:52,720 --> 00:27:55,700 AIM days probably way before that too. 451 00:27:56,559 --> 00:28:00,385 And then, you know, and then There was kind 452 00:28:00,385 --> 00:28:03,985 of a a a chatbot resurgence in, you 453 00:28:03,985 --> 00:28:07,745 know, 2016, 2015, whenever when every company wanted 454 00:28:07,745 --> 00:28:11,490 a chatbot and we're excited about that. Yeah. It didn't quite work. It 455 00:28:11,490 --> 00:28:15,330 seemed to fizzle out. Then, you know, the 456 00:28:15,330 --> 00:28:18,524 the then nowadays, we have So many chat interfaces, 457 00:28:18,904 --> 00:28:22,664 chat GPT and and generative AI seems to be resurgent again. 458 00:28:22,664 --> 00:28:26,024 So there are these weird sine waves, these weird 459 00:28:26,024 --> 00:28:29,340 cycles, and I almost think of it as a coil where, you know, you're starting 460 00:28:29,340 --> 00:28:32,800 at the bottom and you're cycling, but you're also moving up at the same time. 461 00:28:33,020 --> 00:28:36,794 And so How do you how do you surf the wave? That's, that's, 462 00:28:37,255 --> 00:28:40,934 something that's once you kind of, understand the 463 00:28:40,934 --> 00:28:43,900 fact that that's what you're doing, then then then you can be excited about it. 464 00:28:44,220 --> 00:28:47,980 I I think it's fair. Well, we're at that point in the show 465 00:28:47,980 --> 00:28:51,820 where we transition to our, questions. And, we 466 00:28:51,820 --> 00:28:55,325 dropped them into the chat here for you. Our very first one is how did 467 00:28:55,325 --> 00:28:59,085 you find your way into data? Did data find you or did you find 468 00:28:59,085 --> 00:29:02,789 data, Max? Interesting. Well, I 469 00:29:02,789 --> 00:29:06,630 guess I was always interested in math and computer 470 00:29:06,630 --> 00:29:10,070 science. You know, going back to undergrad, you 471 00:29:10,070 --> 00:29:13,415 know, it was like there was a lot of different areas I could choose. I 472 00:29:13,875 --> 00:29:17,715 had a hard time going into a field that, you know, where I 473 00:29:17,715 --> 00:29:21,380 wasn't, using all different parts of my brain and 474 00:29:21,840 --> 00:29:25,280 computer science department, it was it was not just the 475 00:29:25,280 --> 00:29:28,845 mathematics. It was, you know, there was, you know, 476 00:29:29,705 --> 00:29:33,225 there was a bunch of creativity in it as well. There was human computer interface. 477 00:29:33,225 --> 00:29:37,010 There was it. So, So I was kind of, I gravitated to that field 478 00:29:37,010 --> 00:29:40,610 as an undergrad. When I graduated, I I joined a company 479 00:29:40,610 --> 00:29:43,990 called wireless generation, which, today is called Amplify. 480 00:29:45,235 --> 00:29:48,915 And that's it was an education tech company. And I was 481 00:29:48,915 --> 00:29:52,435 doing, you know, some simple kind of software engineering work. Actually, back 482 00:29:52,435 --> 00:29:56,210 then, It was, which sounds really dated now, but, you know, they 483 00:29:56,210 --> 00:29:59,890 were probably doing this up to, like, 2010, which was, you know, 484 00:29:59,890 --> 00:30:03,595 writing c plus plus for the palm pilot. You know, we yeah. 485 00:30:03,595 --> 00:30:07,435 Because it was they were assessing students and then it would sync to to 486 00:30:07,435 --> 00:30:10,815 the web and all that. And Sure. It was a lot of, like, taking 487 00:30:10,875 --> 00:30:14,610 stuff, Taking that information out of databases and putting it into a a 488 00:30:14,610 --> 00:30:18,370 dashboard. And it was it was you know, I I felt like there 489 00:30:18,370 --> 00:30:21,375 could be something more interesting I was doing even though I love kind of the 490 00:30:21,375 --> 00:30:24,655 mission of that company there. So I ended up in grad school. I ended up 491 00:30:24,655 --> 00:30:28,220 at NYU and I went there from I guess 492 00:30:28,220 --> 00:30:30,559 2009 to 2011 really discovered, 493 00:30:32,539 --> 00:30:36,139 you know, data mining, was the 1st related 494 00:30:36,139 --> 00:30:39,845 class. Then I took, You know, machine learning, natural language processing. Actually, 495 00:30:39,845 --> 00:30:43,304 the the machine learning class was with, Jan Lacun, who is, 496 00:30:43,924 --> 00:30:47,150 a very well known machine learning researcher. He's Like The Lani. 497 00:30:47,370 --> 00:30:49,210 The the the the the the the the the the the the the the the 498 00:30:49,210 --> 00:30:51,610 the the the the the the the the the the the the the the the 499 00:30:51,610 --> 00:30:53,068 the the the the the the the the the the the the the the the 500 00:30:53,068 --> 00:30:53,114 the the the the the the the the the the the the the the the 501 00:30:53,114 --> 00:30:53,160 the the the the the the the the the the the the the the the 502 00:30:53,160 --> 00:30:53,207 the the the the the the the the the the the the the the the 503 00:30:53,207 --> 00:30:55,085 the the the the the the the the the the the You know, all the 504 00:30:55,085 --> 00:30:58,865 stuff that exists today. Like, even this was 2010. He would show us a camera 505 00:30:59,085 --> 00:31:02,225 where he would point to different objects. He'd be like key, wallet, 506 00:31:02,800 --> 00:31:06,260 chair, and it would like, the the the text would appear 507 00:31:06,800 --> 00:31:10,400 on the the screen based on what he pointed at. So they knew how to 508 00:31:10,400 --> 00:31:14,155 do all this stuff, that that you think of as as kind of 509 00:31:14,475 --> 00:31:18,015 it it's it almost seems crazy that that was not, like, 510 00:31:19,350 --> 00:31:22,710 and and turned into a product that anyone could use back then that it almost 511 00:31:22,710 --> 00:31:26,070 seems crazy that it took you know so long to do it but they and 512 00:31:26,070 --> 00:31:29,865 actually it it may have been Used by someone. 513 00:31:29,865 --> 00:31:33,565 It's, sure. Maybe we just don't know about it. 514 00:31:35,750 --> 00:31:39,510 Sorry. My paranoia. No. No. You're right. I'm sure it was used quite 515 00:31:39,510 --> 00:31:43,110 a bit, but it it it's just like what it was that kind 516 00:31:43,110 --> 00:31:46,725 of Sitting on his laptop was so much more sophisticated than anything that 517 00:31:46,725 --> 00:31:49,365 that I I saw a year later. But, 518 00:31:50,485 --> 00:31:54,120 Yeah. So it was That was kind of inspiring. And so it was 519 00:31:54,120 --> 00:31:57,740 like, you know, it was 520 00:31:58,360 --> 00:32:01,705 to me, it seemed like a much more interesting problem. Well, how do you How 521 00:32:01,705 --> 00:32:05,145 does the machine learn? You know? How do you, you know, I don't I don't 522 00:32:05,145 --> 00:32:08,840 wanna sit around writing code that's just dead. I want it to To 523 00:32:08,840 --> 00:32:12,440 be alive, I wanted to to learn from experience. And so when you dive into 524 00:32:12,440 --> 00:32:16,265 that question, well, then you get into machine learning, which is actually Pretty well 525 00:32:16,265 --> 00:32:19,865 named. And then and, you figure, okay, well, you need 526 00:32:19,865 --> 00:32:23,385 data to learn from, and then that that ends up being a statistical model 527 00:32:23,385 --> 00:32:26,830 and so on and so forth. So, you know, when I 528 00:32:27,070 --> 00:32:30,910 so Foursquare, was a company that that essentially came 529 00:32:30,910 --> 00:32:34,655 out of NYU And, you know, it kind of intersected. So 530 00:32:34,735 --> 00:32:38,335 and and they wanted to, to learn from from 531 00:32:38,335 --> 00:32:42,150 their data. They wanted to kind of, sort of a 532 00:32:42,230 --> 00:32:45,910 to build a data science team. And so I had already been 533 00:32:45,910 --> 00:32:49,605 working on that sticky map project, And I was into local search. I 534 00:32:49,605 --> 00:32:52,665 loved the the product aspect. I didn't have my new 535 00:32:53,045 --> 00:32:56,679 interest in machine learning and LP in there. So it all kinda came together. And 536 00:32:56,679 --> 00:33:00,139 so that's why I think that was such a good fit for me and probably, 537 00:33:01,559 --> 00:33:04,940 probably would be very difficult to find such a fit again. 538 00:33:08,935 --> 00:33:12,695 Our next question is, what's your favorite part of your current 539 00:33:12,695 --> 00:33:16,160 gig? And that was, in The virtual green room, you said you 540 00:33:16,160 --> 00:33:18,580 kinda had a good story about that. 541 00:33:19,760 --> 00:33:23,415 Right. So I don't. Well, I don't exactly have a a 542 00:33:23,415 --> 00:33:26,535 current gig right now. I have a bunch of different projects that I'm working on. 543 00:33:27,175 --> 00:33:30,960 It was you know, I think It it was on one hand, it was 544 00:33:30,960 --> 00:33:34,400 nice in Foursquare to be able to focus on one thing, and I'm gonna come 545 00:33:34,400 --> 00:33:37,520 back to that. But I feel like you need these periods, almost like the same 546 00:33:37,520 --> 00:33:41,355 as the grad school period That I had, back in 2010 where it was 547 00:33:41,355 --> 00:33:44,955 like, well, you're working on a few different side projects, but let's see. 548 00:33:44,955 --> 00:33:48,095 Hopefully, like, eventually it'll coalesce into something, 549 00:33:48,950 --> 00:33:52,310 you know, something a little bit more long term and permanent. So I'm working on 550 00:33:52,310 --> 00:33:55,910 several projects. One is with with the Foursquare founder, Dennis 551 00:33:55,910 --> 00:33:59,545 Crowley. And we are Working on a new product, a new 552 00:33:59,545 --> 00:34:03,385 kind of city guide where you walk around the city with your headphones in, 553 00:34:03,385 --> 00:34:07,179 with your AirPods in or whatever. And We kinda know what you're passing, 554 00:34:07,580 --> 00:34:11,339 by. We sort of are are using some of the Foursquare 555 00:34:11,339 --> 00:34:15,025 tools that are publicly available that we know about, but also, You know, we're 556 00:34:15,025 --> 00:34:17,984 kind of rigging up our our own thing because we've just done it so many 557 00:34:17,984 --> 00:34:21,590 times. You know how to do it. We're okay. We know what 558 00:34:21,670 --> 00:34:24,870 stores and stuff you're walking past. So what kind of sounds can we play? Right 559 00:34:24,870 --> 00:34:28,390 now, it's a bunch of text to speech. Essentially, the way I've rigged it up, 560 00:34:28,390 --> 00:34:32,085 the the old version 0, the alpha version is, you know, 561 00:34:32,085 --> 00:34:35,925 we asked chat g p t or OpenAI API what to say. So it's 562 00:34:35,925 --> 00:34:39,125 basically like you're you're walking down the street hearing, 563 00:34:39,525 --> 00:34:42,989 content From OpenAI. Interestingly, OpenAI 564 00:34:43,210 --> 00:34:45,710 seems to the the GPT seems to know, 565 00:34:46,890 --> 00:34:50,304 stuff about Every place along the way, like, you don't 566 00:34:50,304 --> 00:34:54,085 have to go into, like, location based database. 567 00:34:54,145 --> 00:34:57,900 It seems to seems to know quite a bit. There is a question of the 568 00:34:57,900 --> 00:35:01,740 all the content is there's some interesting content in there, but it all ends up 569 00:35:01,740 --> 00:35:04,300 being kind of mediocre. So it's like, okay, well, how do we turn this into 570 00:35:04,300 --> 00:35:07,685 something really cool? I think, you know, in the end, having, 571 00:35:09,345 --> 00:35:13,185 you know, you know, maybe music and and and speeches and an art 572 00:35:13,185 --> 00:35:16,710 project somehow in there, based on where you walk is an interesting 573 00:35:16,710 --> 00:35:20,069 idea. So if I could That'd be cool. Yeah. I could be like a 574 00:35:20,069 --> 00:35:23,589 platform that people can use, like a cultural version of 575 00:35:23,589 --> 00:35:27,415 Foursquare. Yeah. Yeah. And or maybe it's just 576 00:35:27,415 --> 00:35:31,095 like an enhancement of the the sounds of the city. Or maybe 577 00:35:31,095 --> 00:35:34,690 it's, You know, I mean, a lot of people think, okay, maybe maybe a tour 578 00:35:34,690 --> 00:35:38,150 guide. I I don't know. But, you know, it it's it feels like, 579 00:35:40,015 --> 00:35:43,855 It feels like there needs to be, a variety 580 00:35:43,855 --> 00:35:47,455 of use cases tried because there's there's a lot you could do with it. And 581 00:35:47,455 --> 00:35:51,040 and Maybe, you know, if if you put this in the hands of more 582 00:35:51,040 --> 00:35:54,720 creative or of of additional creative people, they would, 583 00:35:54,880 --> 00:35:58,555 ultimately find something interesting. I'm also working 584 00:35:58,715 --> 00:36:02,395 yeah. Oh, I could answer questions about that. But then my other project is my 585 00:36:02,395 --> 00:36:05,849 other 2 projects are are kind of interesting as well. Well, I have the 586 00:36:05,849 --> 00:36:09,609 podcast, The Local Maximum. So still doing that every week and, you know, 587 00:36:09,609 --> 00:36:12,830 interviewing people. Talking about, 588 00:36:14,775 --> 00:36:18,535 talking about data, talking about AI, you know, few episodes on the 589 00:36:18,535 --> 00:36:21,994 whole. You know, all the drama around OpenAI recently. 590 00:36:24,260 --> 00:36:27,059 I I never wanted to become kind of the the the, 591 00:36:27,940 --> 00:36:31,000 the the the tech drama, you know, what's it called? 592 00:36:31,465 --> 00:36:35,305 TMZ of technology? Yeah. Yeah. But but that's something that happened because 593 00:36:35,305 --> 00:36:38,985 I remember, like, last year, a couple years ago, there was all this craziness coming 594 00:36:38,985 --> 00:36:42,690 out of Google with, You know, there was 1 guy at Google who said, 595 00:36:42,690 --> 00:36:46,370 you know, he thought that the LLM has come to life. And Oh, yeah. And 596 00:36:46,370 --> 00:36:50,075 then and then there was a there was A whole 597 00:36:50,075 --> 00:36:53,435 bunch of stuff with, like, the the AI safety, you 598 00:36:53,435 --> 00:36:56,815 know, seemingly staffed 599 00:36:56,875 --> 00:37:00,650 by, people who are a little nutty. And 600 00:37:00,650 --> 00:37:04,170 so, it was a And they fired a bunch of 601 00:37:04,170 --> 00:37:07,944 people From that team too. So, like, there's 602 00:37:08,145 --> 00:37:11,885 definitely, it was something weird some weird mojo 603 00:37:11,885 --> 00:37:15,670 was going around. That's for sure. Yeah. And when when I cover that, I 604 00:37:15,670 --> 00:37:18,550 mean, it's hard to, you know, it's hard to hide the fact where it's like, 605 00:37:18,550 --> 00:37:21,935 wow, everyone in this story seems kinda nutty. But I also try to, you know, 606 00:37:22,095 --> 00:37:25,155 I try to take a step back and say, okay, this is what we know. 607 00:37:25,375 --> 00:37:28,575 These are a few things that could be happening internally, but we don't know everything. 608 00:37:28,575 --> 00:37:31,235 I'm not gonna jump to conclusions. But, 609 00:37:32,190 --> 00:37:35,790 I I I I try when I'm covering a story in a local maximum to 610 00:37:35,790 --> 00:37:39,470 give, like, a a balanced, a balanced version of 611 00:37:39,470 --> 00:37:42,885 of whatever story I come across. You know, maybe it's my show as I try 612 00:37:42,885 --> 00:37:46,185 to give my opinion. But, yeah, I I 613 00:37:46,725 --> 00:37:50,265 I my attempts, which, you know, some people have have, 614 00:37:51,370 --> 00:37:54,970 said I I I've captured that. But my my attempt is to sort 615 00:37:54,970 --> 00:37:58,730 of, try to try to 616 00:37:58,730 --> 00:38:02,425 approach each Story with a little bit of humility and try 617 00:38:02,425 --> 00:38:05,725 to help people understand what's going on without the 618 00:38:05,945 --> 00:38:09,560 raw emotion that you get often on on Twitter. Gotcha. That's a good 619 00:38:09,560 --> 00:38:13,240 point. Yeah. So we have, go ahead. I'm 620 00:38:13,240 --> 00:38:16,635 sorry. Oh, no. No. It's okay. Go ahead. Okay. So we got, 621 00:38:17,035 --> 00:38:20,875 3 complete dishonest. And, the first is when 622 00:38:20,875 --> 00:38:24,420 I'm not working, I enjoy blank. Right. 623 00:38:24,420 --> 00:38:27,940 So now that I've moved to Connecticut, I feel like I 624 00:38:27,940 --> 00:38:31,684 am such a a Connecticut stereotype where I kinda, like, Drive 625 00:38:31,684 --> 00:38:34,984 around, going for walks in the woods and into 626 00:38:36,005 --> 00:38:39,720 various malls and stuff. So it's like it's like it's When the 627 00:38:39,720 --> 00:38:42,360 weather's good, you go into the woods. When the weather's bad, you go into the 628 00:38:42,360 --> 00:38:46,040 mall. Yeah. So I I actually like enjoy doing 629 00:38:46,040 --> 00:38:48,380 that. I enjoy listening to podcasts. 630 00:38:49,845 --> 00:38:53,365 I, honestly, enjoy hanging out with friends. You know, 631 00:38:53,365 --> 00:38:57,130 after, I used to live in New York City. I enjoyed 632 00:38:57,130 --> 00:39:00,910 it a lot, and I sorta had this, situation where 633 00:39:01,369 --> 00:39:05,155 I had this be careful what you wish for because, at the end of 2019, 634 00:39:05,155 --> 00:39:08,435 I was like, oh my god. I'm going to, like, events every single day. It's 635 00:39:08,435 --> 00:39:12,190 just it's just too much. How can we, like, how can we, you know, 636 00:39:12,190 --> 00:39:16,030 cut back on that. And then COVID came. And then to me, it was just 637 00:39:16,030 --> 00:39:19,470 it was the worst thing because it was like, okay, you stay in your apartment 638 00:39:19,470 --> 00:39:23,145 in New York City all day and you don't go and and talk to anyone. 639 00:39:23,605 --> 00:39:27,365 And it was just like it it it was just awful. It just 640 00:39:27,365 --> 00:39:31,200 felt like a a prison. So I I 641 00:39:31,200 --> 00:39:34,880 moved to New Hampshire for a couple years, then I came back. But, you 642 00:39:34,880 --> 00:39:38,320 know, nowadays, when I get a chance to hang out with with friends and and 643 00:39:38,320 --> 00:39:42,115 family, I just I try to do it, whenever I can because I'm 644 00:39:42,115 --> 00:39:45,875 not like, you know, it's not like when I was living in New York in 645 00:39:45,875 --> 00:39:48,135 the 2010s and got kind of overload on that. 646 00:39:50,130 --> 00:39:53,970 Right. Right. So, yeah. That's my answer there. And we have another complete this 647 00:39:53,970 --> 00:39:57,350 sentence. I think the coolest thing in technology is blank. 648 00:40:03,255 --> 00:40:07,080 The the way I've been putting it recently Is this, 649 00:40:07,480 --> 00:40:11,020 where, you know. It. 650 00:40:13,160 --> 00:40:16,995 You know, back maybe 10 years ago, the story we 651 00:40:16,995 --> 00:40:20,675 were getting that the hopeful story we were getting was that, okay, if you're an 652 00:40:20,675 --> 00:40:24,440 engineer, you could Build anything you want at 653 00:40:24,440 --> 00:40:28,120 a very low cost or if you're not an engineer for anyone because we 654 00:40:28,120 --> 00:40:31,505 have access to social media. You know, you can, 655 00:40:31,905 --> 00:40:35,505 you can put anything out there into the world that you want 656 00:40:35,505 --> 00:40:39,200 and and have people read it if if if they want to, or have people 657 00:40:39,200 --> 00:40:42,880 look at it if they want to. And so that was kind of the new 658 00:40:42,880 --> 00:40:46,664 exciting world. I think today, The new exciting 659 00:40:46,664 --> 00:40:50,365 world goes well beyond that, which is going to be like, 660 00:40:51,865 --> 00:40:55,589 you you can create worlds. Any Any world that you wanna 661 00:40:55,589 --> 00:40:59,349 build, any scenario that you can imagine, you 662 00:40:59,349 --> 00:41:03,175 can just have a machine fill in all the gaps for you and, You 663 00:41:03,175 --> 00:41:06,775 know, write the write the story, make the 664 00:41:06,775 --> 00:41:10,615 image and maybe, like, you know, make the make the video, make the whole 665 00:41:10,615 --> 00:41:14,340 world. So I think, I I I think the 666 00:41:14,340 --> 00:41:18,100 idea with generative AI that I want people thinking about more that that I 667 00:41:18,100 --> 00:41:21,505 I also wanna think about more is, like, Okay. If you could create any world 668 00:41:21,505 --> 00:41:25,265 you want, to explore, to live in, just 669 00:41:25,265 --> 00:41:28,545 to, you know, maybe it's something to to teach us about something. Maybe it maybe 670 00:41:28,545 --> 00:41:30,740 it's just an artistic adventure 671 00:41:32,080 --> 00:41:35,680 venture, you know, what kind of world do you want to create because that's 672 00:41:35,680 --> 00:41:39,405 that's going to become very cheap very quickly. Yeah. I could 673 00:41:39,405 --> 00:41:42,765 see that. So I'm gonna skip to, 674 00:41:43,164 --> 00:41:46,840 share something different about yourself. But we remind our guests 675 00:41:46,840 --> 00:41:49,020 to remember it's a family podcast. 676 00:41:52,765 --> 00:41:56,045 Okay. And I'm, you know, I'm I'm trying to, 677 00:41:58,285 --> 00:42:00,944 I'm I'm trying to think of an answer here, and it's not because, 678 00:42:02,599 --> 00:42:06,280 it's it's not because of the, of the of the caveat there, 679 00:42:06,280 --> 00:42:09,960 but it No. No. I get it. Well, you've already covered a lot 680 00:42:09,960 --> 00:42:13,595 that's It's different. I did. It is good stuff. Yeah. I mean, I 681 00:42:13,595 --> 00:42:17,390 think, I think one thing that, It's, 682 00:42:17,630 --> 00:42:21,470 I I enjoy doing that that that I forgot to mention, because 683 00:42:21,470 --> 00:42:24,750 I'm actually doing it again for the 1st time in in in 6 years was, 684 00:42:25,070 --> 00:42:28,845 I was a member of the Yale Alumni Service Corps. It's not 685 00:42:28,845 --> 00:42:32,245 a member. It's like you can do a a it was essentially we were doing 686 00:42:32,245 --> 00:42:35,620 trips to underserved Communities around the world and, 687 00:42:35,940 --> 00:42:39,780 you know, doing little, like, kinda, kind of service 688 00:42:39,780 --> 00:42:43,434 trips where you'd ever Either build a structure or work with small business 689 00:42:43,434 --> 00:42:47,115 owners or go, you know, teach in a school. And so I've been 690 00:42:47,115 --> 00:42:50,839 to, Nicaragua and Ghana And I actually 691 00:42:50,839 --> 00:42:54,599 got to lead one of their trips in 2017 and that was to the Fort 692 00:42:54,599 --> 00:42:58,359 Mohave Indian reservation. Very different kind of a trip because 693 00:42:58,359 --> 00:43:02,035 it was Within the United States here. And 694 00:43:02,035 --> 00:43:05,875 so it was honestly a lot easier. Because that's very cool 695 00:43:05,875 --> 00:43:09,720 flying to Vegas. But yeah. But we're actually going back there, in in a few 696 00:43:09,720 --> 00:43:13,240 months after 6 years. And so I was so even though it was less 697 00:43:13,240 --> 00:43:17,035 convenient this time around, I'm I'm very excited Do that. And so 698 00:43:17,255 --> 00:43:20,395 I I don't know. I really like learning about different cultures, 699 00:43:21,415 --> 00:43:24,900 different philosophies, different religions. I think A lot of people might 700 00:43:24,900 --> 00:43:28,660 assume given the you know given the tenor of my 701 00:43:28,660 --> 00:43:32,260 podcast that I'm very like you know rationalist and I talk about Bayesian 702 00:43:32,260 --> 00:43:36,055 inference a lot. But, I I 703 00:43:36,055 --> 00:43:39,835 sort of venture out of that a lot. I don't think that, 704 00:43:40,430 --> 00:43:43,869 That raw math can, can explain everything in life. And I also love like the 705 00:43:43,869 --> 00:43:47,170 diversity of, of cultures and stuff. So No. It's cool. 706 00:43:47,675 --> 00:43:50,795 That so that's maybe a positive thing, so I I don't know. It's very positive. 707 00:43:50,795 --> 00:43:54,635 Something different. No. It definitely is. It definitely is. So where can 708 00:43:54,635 --> 00:43:57,609 folks find out more about you and what Sure. So you mentioned you have a 709 00:43:57,609 --> 00:44:01,150 podcast, which I love the name, The Local Maximum. 710 00:44:01,930 --> 00:44:05,705 Right. Yeah. Local maximum's triple entendre, because it's got my name, Max. 711 00:44:05,785 --> 00:44:09,565 It's a local maximum is, of course, you know, in machine learning, 712 00:44:10,025 --> 00:44:13,865 when you're when you're trying to find the, well, sometimes it's often the local minimum 713 00:44:13,865 --> 00:44:17,400 if you're trying to, Minimize the the loss function, but the 714 00:44:17,600 --> 00:44:21,440 in basing inference, if you're trying to maximize the probability, whatever, you're you you get 715 00:44:21,440 --> 00:44:24,895 stuck stuck in one of your next which was Your name is Max. So Yeah. 716 00:44:24,895 --> 00:44:28,255 Exactly. Right. Right. That's the first one. That's the second one. And then, you know, 717 00:44:28,255 --> 00:44:30,975 I I worked on location data a lot, so it's kind of a a triple 718 00:44:30,975 --> 00:44:34,620 meaning. And so, I've 719 00:44:34,620 --> 00:44:38,220 been doing that for for quite a while. You can go back into kind of 720 00:44:38,220 --> 00:44:41,915 a a really extensive library there. And, I have 721 00:44:41,915 --> 00:44:45,695 the website local max radio.com. I have. 722 00:44:47,410 --> 00:44:51,250 If you go, I have local maximum labs. If you go to local maximum local 723 00:44:51,250 --> 00:44:54,849 max radio.com/labs, I have a 724 00:44:54,849 --> 00:44:58,215 bunch of papers And, you know, kind 725 00:44:58,215 --> 00:45:02,055 of works that I've done, which, you know, includes some 726 00:45:02,055 --> 00:45:05,835 discussion of machine learning, like kind of the math mathematics behind bias correction, 727 00:45:05,975 --> 00:45:09,720 but But also something kind of fun that I did, like, with the podcast last 728 00:45:09,720 --> 00:45:13,480 year, which is, like, I just rewrote the US constitution, fixed a bunch of 729 00:45:13,480 --> 00:45:17,065 things just because I I felt that was fun. I was Taken aback by how 730 00:45:17,065 --> 00:45:19,945 mad people get when you when you do that, it's not like I was actually 731 00:45:19,945 --> 00:45:23,545 trying to, you know, run a political campaign for it. I just thought it was 732 00:45:23,545 --> 00:45:27,290 a a fun project, and I learned a lot. But Some once you venture into 733 00:45:27,290 --> 00:45:30,910 the political, people start treating things different. People get angry pretty quick. 734 00:45:30,970 --> 00:45:34,415 Yes. That's true. Yeah. Yeah. I I I The 735 00:45:34,895 --> 00:45:38,255 I I I love to hear criticisms on it. I wanna hear what what what 736 00:45:38,255 --> 00:45:42,099 people think. But The one criticism 737 00:45:42,160 --> 00:45:44,960 that I get a lot, which I really hate is, like, how dare you spend 738 00:45:44,960 --> 00:45:48,320 your free time on this, which I I just don't get at all. 739 00:45:48,320 --> 00:45:52,155 Yeah. But, which, you know, whenever I put 740 00:45:52,155 --> 00:45:55,915 out some kind of math paper, even if it's like and there there is one 741 00:45:55,915 --> 00:45:59,660 called relative probability, which is, you know, sort 742 00:45:59,660 --> 00:46:03,119 of an abstract paper where it's like, okay, a reimagined probability 743 00:46:03,260 --> 00:46:06,945 theory as, okay, let's say you can't talk about The probability of something 744 00:46:06,945 --> 00:46:10,785 happening. Let's say you can only talk about 1 probability relative 745 00:46:10,785 --> 00:46:14,465 to another. What, what does that look like? And I just stated some basic facts 746 00:46:14,465 --> 00:46:18,210 and, you know, Not that many people gonna use it. Maybe people 747 00:46:18,210 --> 00:46:21,730 won't use it for for a while. I I feel like it's an interesting idea, 748 00:46:21,730 --> 00:46:24,869 and I feel like it will have uses eventually. But, 749 00:46:25,455 --> 00:46:28,495 You know, nobody criticized me for that. For like, how dare you spend your free 750 00:46:28,495 --> 00:46:32,335 time on that? Exactly. They they pick on you for the other 751 00:46:32,335 --> 00:46:35,240 stuff. Yeah. I mean, I I look at people. I mean, you spend your free 752 00:46:35,240 --> 00:46:38,920 time yelling at people on Twitter. I mean, what's the difference? I was gonna say 753 00:46:38,920 --> 00:46:42,665 you can you can look at TikTok, and you can find far more destructive uses 754 00:46:42,785 --> 00:46:46,505 is of Exactly. Exactly. So that's so that's that's my 755 00:46:46,505 --> 00:46:50,185 main thing. I I think maybe with the, with the Constitution, I think people have 756 00:46:50,185 --> 00:46:54,010 their sort of ideal society in mind. And if If your thing doesn't wind up 757 00:46:54,010 --> 00:46:57,309 with that, they they kind of perceive you as a threat. Like you're trying to, 758 00:46:57,770 --> 00:47:01,289 like I was trying to revitalize democracy, but some people are saying, no, you're 759 00:47:01,289 --> 00:47:05,115 backsliding on democracy. Alright. Like, let's talk about it. But, yeah, it's it's 760 00:47:05,115 --> 00:47:08,875 people get you know, people get different. We need to have you back and talk 761 00:47:08,875 --> 00:47:12,640 about that more. Yeah. For sure. For sure. Talk about that. Absolutely. 762 00:47:12,640 --> 00:47:16,160 We'd love having you. Both Andy and I, however, do have a hard stop, and 763 00:47:16,160 --> 00:47:18,985 I would love this This covers you to go out for a couple hours, and 764 00:47:18,985 --> 00:47:22,825 we'll talk to you more. And I just had a a conversation last night with 765 00:47:22,825 --> 00:47:26,445 my cohost that went a couple hours. I know how it goes. Yeah. Yeah. Yeah. 766 00:47:26,790 --> 00:47:29,369 We ended at 1 AM, and I was like, oh my god. 767 00:47:31,270 --> 00:47:35,050 Well, those 1 AM conversations. I know what you mean. You got it. So 768 00:47:35,375 --> 00:47:39,135 With that, we'll definitely make sure. Send us all your links, and 769 00:47:39,135 --> 00:47:42,495 we'll make sure we get them in the show notes, and we'll let Bailey, our 770 00:47:42,495 --> 00:47:45,980 semi extension AI host, Co host, 3rd 771 00:47:45,980 --> 00:47:49,420 host, wrap up the show. And thank you, dear 772 00:47:49,420 --> 00:47:53,005 listener, for subscribing to our podcast. You 773 00:47:53,005 --> 00:47:56,684 have subscribed to us, haven't you? Once you do, 774 00:47:56,684 --> 00:48:00,510 please be sure to rate and review our podcast on iTunes, Stitcher, 775 00:48:00,650 --> 00:48:04,410 or wherever you subscribe to us. Having high ratings 776 00:48:04,410 --> 00:48:08,010 and reviews helps us improve the quality of our show and rank us more 777 00:48:08,010 --> 00:48:11,595 favorably with the search algorithms. That means more 778 00:48:11,595 --> 00:48:15,275 people listen to us, spreading the joy. And, 779 00:48:15,275 --> 00:48:17,935 can't the world use a little more joy these days? 780 00:48:19,050 --> 00:48:22,570 So, go do your part to make the world just a little better and be 781 00:48:22,570 --> 00:48:24,350 sure to rate and review the show.