Listen to Carly Taylor, an expert in gaming and machine learning. She shared insights on the rise of mobile gaming, player engagement, and how machine learning improves gaming experiences. We also discussed Carly’s experience with ChatGPT, finding it engaging and informative. The episode provides a look at how gaming and machine learning intersect, and the potential for chatbots like ChatGPT in various industries.
Meet our Guest
Machine Learning for Call of Duty, Activision
Carly Taylor is a Machine Learning Engineer at Call of Duty, Activision. She is a thought leader in the data space. She obtained her M.S. in chemistry from the University of Colorado focusing on computational quantum dynamics. She has authored multiple peer-reviewed publications and holds two non-provisional machine learning patents. When she isn't writing about herself in the third person, building mechanical keyboards, or neglecting the oxford comma, she works as a security strategist for Call of Duty at Activision Publishing.
You have a couple ingestion points for machine learning. You can do machine learning real time from those events that do get sent in real time to the data lake, you can just grab it before or as it’s being written depending on how quick you need to do something.
The first time I used chat, GPT I work in ML and I was like, holy expletive. Yeah. This is crazy. Yeah. I was taken aback by it. And as someone who kind of understands the state of the industry, I wouldn’t say I’m an expert by any stretch of the word, but like, you know, cursory overview, not blown away by things too often anymore.
Dave Mariani: Hi everyone, and welcome to another data driven podcast, and today’s special guest is Carly Taylor. Hi, Carly.
Carly Taylor: Hey. So happy to be here. Thanks for having me.
Dave Mariani: Thank you. So, thank you for joining us. So, Carly is the, senior manager for Security strategy at Activision, and she’s an expert in machine learning and really you’re just a, a, a very well known data and analytics influencer . So, so happy to, to be able to talk to you today. Carly.
Carly Taylor: Well, what a kind intro. Thank you. .
Dave Mariani: , you have a, you have a, you have an amazing background. could you like to, just to, let’s just start with just how you got to where you are today. What was your path to, to, to, to, to being Carly Taylor at Activision
Carly Taylor: I feel like this goes along with what we were just talking about. I can say I was a chemist and then yada, yada, yada. No, I’m very , but a lot happened in that, yada yada yada. yeah, I, so I started my career as a chemist. It’s been a wild ride, you know. Wow. I, when I graduated, actually after undergrad was when I started looking for lab jobs, trying to figure out, you know, where am I gonna live What am I gonna do and I quickly realized that there were not many jobs, , despite what I had been promised throughout my education. Surprise, surprise. And I started to realize, you know, in order to even be competitive in a chemistry back, like in a chemistry job or chemistry lab, I would need some sort of graduate degree. And at that point I was already thinking like, I don’t know if this is gonna be for me long term.
Carly Taylor: and so I had an opportunity to do computational chemistry as part of my master’s program. And I realized, okay, this is kind of gonna be the perfect stepping stone for what I eventually want to do, which is kind of get out of chemistry and into tech mm-hmm. . And instead of going back to school for a comp sci degree and doing all of that all over again, you know, I’m old enough that this was before like data science was a degree . Like you had to kind of knee your own, like take the math classes. Yeah. Take all the comp psych classes. I, I decided to do this computational chemistry cuz it was kind of like enough of what I had already done with some new things that I’d be learning and try to see if that wouldn’t be a good entry point. It was actually still really hard cuz nobody knew what a computational chemist was. No,
Dave Mariani: I don’t know what that is.
Carly Taylor: Getting that first job was super difficult after I got my master’s degree. you know, I think it took me eight months, almost a year of applying, getting rejected, feeling really bad about myself. I was walking dogs for money, , it was a, it was actually not that bad. The dogs at least were fun. .
Dave Mariani: So what, so what, so who gave you Carly who gave you that chance Like who, who gave you the chance to get
Carly Taylor: Started It was a marketing analytics company, surprisingly. Uhhuh . I didn’t know anything about marketing, Uhhuh, . I knew about analytics but not pertaining to marketing. and so it was definitely a culture shock. I didn’t have any like business clothes cause I’d always worked in labs. Right. . So I was like, what do I do with myself It was so awkward. .
Dave Mariani: Oh my gosh. That’s hilarious. And, and so, and that was, and that was your first job Was it as a data science, scientist or it more, it was
Carly Taylor: More like a data analyst. Okay. So I think the title was like marketing science Analyst or something really like long-winded. Okay. But it was basically data analysis. And the thing I didn’t love about that job, but it was a really good entry point for me was that it was almost all in Excel. Mm-hmm. . And so I had really wanted to start using like my Python skills. I had studied r at that point too mm-hmm. . And I was thinking like, I could do a lot of this stuff in r I could automate a lot of my workflows, right Mm-hmm. , like, I was having the realization that a lot of people have when they’re doing the same thing over and over, thinking like, there has to be a better way mm-hmm. to not do this repetitive kind of like calculation work mm-hmm. and, you know, it was in trying to discover if I could do that there, I had an opportunity to go to a startup and was able to kind of investigate that a little bit more, I think, than at the other company.
Dave Mariani: You know, sometimes it’s like, it’s good to be lazy, right Because if you’re lazy, you don’t like to do repetitive things and so you’ll actually invest time into Yes. Actually, you know, taking away the work, yes. By automating it. Yes. So
Carly Taylor: I liked looking at it that way. That’s a really way
Dave Mariani: About it. Okay. So then, so now you’re in the gaming industry and you know, everybody, you know, especially everybody coming out up through tech would love to work for the gaming industry or in the gaming industry. But tell us like, what does it, what does it really look like Carly to, to work in gaming
Carly Taylor: It’s as fun as you would think, but it’s a lot harder than you would think. Hmm. And so it’s kind of this dichotomy of the challenges we face are really big. There’s a lot of, you know, data at scale that we handle. And since, especially for someone with my personality, I’m very passionate about our game. I love our game. I played Call of Duty long before I ever worked at Activision mm-hmm. . And so that passion translates into like not being good at setting boundaries and having a good work-life balance, right Mm-hmm. . And that’s something that I personally struggle with because I could try to be on all the time mm-hmm. . and there’s certainly problems to solve all the time mm-hmm. . And so, you know, I think when you marry a passion with your day job, you just have to be cognizant that you don’t let yourself burn out because it can ruin how much you loved, you know, what you were passionate about before. Completely. And so this year for me is the year of trying to set better boundaries, take my actual PTL , which has been difficult for me to do in the past. And I think that, you know, I’m not quite burned out yet, but I could see myself getting, like, going down that path of yeah, I love doing this so much and this is not only my hobby, but my work and just not really having anything outside of it.
Dave Mariani: So how do you, so besides taking PTO , how do you, how do you define those boundaries What, what, what’s some of the techniques that you
Carly Taylor: Use Well, it’s a work in progress, but I actually asked people on LinkedIn the other day about this because I was wondering what everyone else did, you know Mm-hmm. , if, especially working from home, which I love, working from home. I don’t think I’ll ever go back to an office, but it’s easy to not have like a boundary of like, when I walk in the door, I’m home. Mm-hmm. , right Because my office is also in the room where I play video games. Like this is my battle station as well. And so , you know, I, from what people were saying, it’s something I definitely need to do is like either getting slack off of your phone or silencing notifications at the end of the day. not that’s a good one. Having email on your phone, which I’m like, okay, I think I might actually die. That’s like my lifeline. So I don’t know if I’m ready for that one yet, but setting like, focus times on your phone, you know mm-hmm. so that it’ll silence everything. and then really just being, I think mindful is probably the first step, you know, like when you realize, oh, it’s 8:00 PM and I’m trying to like open my laptop and do some work mm-hmm. say like, no, not tonight, , this can wait. but I’m still, you know, I’m still trying to learn how to figure that out. There’s, there are things where I do it
Dave Mariani: , it’s very true that like, you know, because of Slack, I mean, you’re kind of always on. Yeah. I have to slack up my phone too and, you know, I just try not to look at my phone after when I’m, when I’m off. But, you know, sometimes you can’t help it.
Carly Taylor: But then how are you gonna scroll through social media
Dave Mariani: Oh,
Carly Taylor: Like the dog videos on Instagram need
Dave Mariani: My attention. Okay. So yeah, just, just a little bit a hint to you and, and, and the listeners. I really hate social media. Well, that’s good. So here I am doing a podcast and promoting it on social media. It’s like, I like, it’s really, it’s, it’s really not. It’s, it’s really something that doesn’t come natural for me. Okay. I, that’s something I don’t like as social media
Carly Taylor: Really. That’s interesting. Yeah. You seem like a natural at it. I mean, you’re always having to be out here and be the face of things.
Dave Mariani: So Yeah. You could
Carly Taylor: Probably the obsession on that.
Dave Mariani: Well, hey, I, I love to, I love to talk to people like yourself. but yeah. When it comes to, you know, posting Okay. Very unnatural for me. Yeah. even though I worked at Clout, do you remember clout, near Clout score Do you remember that in the back in the day Okay. Well, , so clout with the k, we, we would score people based on, how, influential they were on social media. Oh, interesting. And you’d have a clout score. and so I worked at the company that produced a clout score, which oh my gosh, you know, gosh, of course. I had terrible clout and I had to like, you know,
Carly Taylor: were they like this person with the least amount of clout has
Dave Mariani: To, it was the guy, you know, owning the, owning the big data that that powers the, the algorithm that calculates your, your influence score, you know Anyway. Okay. So anyway, that’s enough about that. But, but in terms of like, you know, you know, people probably don’t, I don’t realize this. I know, my, my, my kids love Call of Duty and I’m always watching them play it . I’m not, I’m not, I’m not fast enough or, you know, I don’t have enough, you know, dexterity to actually play. It’s hard. It looks really hard. I’ve
Carly Taylor: Noticed, I noticed the older I get, the worse I get. I’m
Dave Mariani: Just like, just, you know, I’ve stopped it. I think at, at we, we Sports is like the last time I , but, but, but what’s, what’s involved in making an online game work Carly, can you, can you tell us a little bit about that
Carly Taylor: So this is a really good question. I’m gonna do kind of like the TLDR high level because there’s a lot that goes into making online game work. I’d say first you have to pray to the online engineering gods that everything is gonna go smooth. , you have to make a sacrifice at the altar of networking. no, but al jokes aside, so for most online games nowadays, I’d say, we’ve moved to a client server architecture mm-hmm. wherein, you have a server, it sends the game state. So it, it’s the authority of what’s happening in the game world. And then you have all of your cli what we call clients. So that’s like your PS five or your pc mm-hmm. are connected to that server. Mm-hmm. , the server is sending game state to every connected client and the client, like your PS five is just a mirror of the server state.
Carly Taylor: Yeah. and then in return, the clients are sending the user inputs from what your character is doing back to the server. Mm-hmm. . and that’s another yada yada, yada, yada yada . Mm-hmm. , there’s a lot of like interpolation that goes along with this to make sure you don’t feel that lag. The fact that you’re mirroring a server time that’s actually in the future from where you are. Right. Yeah. And there’s a lot of like, connectivity and back and forth and interpolation to make it feel like it’s actually real time and not like a few milliseconds of lag in between. Cuz you also have, you know, the time round trip it takes from your client to the server. Yeah. Depending on where you’re at in the world can be super fast. You know, I’m in LA so it can just be like super quick if there’s a bunch of people around me playing and we have a server close by, but let’s say I’m somewhere rural, right Mm-hmm. , like I’m, I’m playing with some people who might be a thousand miles away from me and the server might be at a data center 2000 miles away from me. And so there’s a lot that goes into try to make that experience not feel like it’s 2000 miles away from you . And it’s really, really a hard problem.
Dave Mariani: It, it sounds like it’s magic. And then you have to do that obviously at massive scale too. Exactly.
Carly Taylor: Exactly.
Dave Mariani: Yeah. So that’s a, that’s a big challenge. So, you know, so how does, how does machine learning sort of work its way into the gaming industry and what you do
Carly Taylor: Yeah, so that’s a really good question. A lot of what we just talked about, right Like the clients sending game state to the server mm-hmm. right at the end of the match, you also have all of that game state, all of the inputs from the clients, we call it telemetry. Mm-hmm. gets sent to our data lake because we wanna be able to figure out, you know, what happened in the match. If everything is like playing the way it should, like how do you investigate if a map has good playability, right You need to be able to look at how people interacted with the world, right So all of that client telemetry and server telemetry gets sent downstream to our data lake, either during the match or post-match, depending on what kind of event that it is. and then once that’s in the data lake and it’s accessible, you can do all sorts of things.
Carly Taylor: I will say that, you know, you have a couple ingestion points for machine learning. You can do machine learning real time from those events that do get sent in real time. Mm-hmm. to the data lake, you can just mm-hmm. , you know, grab it before or as it’s being written depending on like how quick you need to do something. But you can also do like post-match analysis where you’re grabbing, you know, a match worth of data and looking at something. there’s also things that people do that aren’t in the game, right Like, we have a store, we have mm-hmm. the ability for you to customize your weapons mm-hmm. . So there’s lots of opportunities for machine learning there to, you know, figure out what kind of things are people wanting to play with. What, what’s your play style Are you like a sniper Should we be showing you like, Hey, here’s a cool new sniper like scope that you can use on your game.
Carly Taylor: Right and so I’d say those are more of like the recommendation systems. There’s also what happens before you play the game, right Mm-hmm. , like we need to actually build the game world. Mm-hmm. . So there’s a really, like, I’d say burgeoning community of deep learning engineers mm-hmm. who are trying to figure out how to better create content for games for like the game world, right Mm-hmm. , because you have tons of artists, why would they spend their time like making a new wood grain for a desk, you know, that no one’s ever really gonna look at when you can have some, like a deep learning algorithm make the textures or make something interesting for a room or figure out how to arrange furniture in a room based on where the door is, right Like that kind of stuff. That’s not, I wouldn’t say like the best use of an artist’s time that can just like, like you said, automate away this stuff that makes us feel lazy that we don’t wanna do all the time.
Dave Mariani: . Right Exactly. Exactly.
Carly Taylor: and the last one I’ll talk about was actually from a LinkedIn post I wrote this beak, which is something that’s really near and dear to my heart after we’ve launched DMZ, is how do you make in-game AI more human mm-hmm. . And that’s something I think about a lot because, you know, I think we’re all nerds here. We think about like, what does it mean to pass the Turing test mm-hmm. , what does it mean to like, look human, we see chat G P T, and we’re like, yeah, you know, what, where are these lines They’re getting blurred, right Yeah. Yeah. And I’d say that for like in-game AI and bots, there’s like a huge group of people working on either reinforcement learning or imitation learning mm-hmm. to try to make these things play more human-Like, so that, you know, if some, if a bot does something stupid, it doesn’t kind of take you out of the immersion of the game and you’re like, Ugh, I just remembered this is a computer because it did, you know, whatever that a human
Dave Mariani: Would never do. Right. Right. I mean, that’s like, so it sounds like it’s, there’s a lot of things, obviously there’s a lot of things you could do with machine learning in terms of enhancing gameplay and understanding what users are doing. Probably monetization is really important. Exactly. I would imagine ,
Carly Taylor: still have to make money, right
Dave Mariani: . Exactly. How much of it is like actually something that’s done post-haste or after the fact, offline, not offline, but in batch as opposed to something that’s happening in real time I mean, is it, you, you know what I’m saying, Carly Like, I do. Yeah. Yeah.
Carly Taylor: so for us, I would say that the realtime machine learning is mm-hmm. new, but we’re doing it , the post batch was more, was happening more often. And I say this after reading a bunch of blogs from different gaming companies, and I think that the consensus used to be that it was easier to batch xFi data from the servers, right Mm-hmm. , because they’re doing all this stuff, they’re already maintaining game state, so like the pipeline to get the data off of the server. It just used to be that they were architected to do that after the game completed. Right. And so they, you know, it would finish and they’d wrap everything up and then send that, payload somewhere. So the switch to get your server to start sending events in real time, it’s not, you know, a crazy engineering challenge. It’s something that Facebook and Google, like real time machine learning has been done for a long time. Mm-hmm. . But I just think that the scale and the actual, like, considerations about where those models sit is mm-hmm. now being reinvestigated, at least we’re investigating it heavily and from other people I’ve talked to, it’s, it’s on the horizon, but I wouldn’t say it’s like the norm.
Dave Mariani: That’s interesting. Yeah. That makes sense. I mean, especially doing that at scale must be insanely difficult. Yes. so I get that’s a, that’s a, that’s a amazing Kafka cue, , so,
Carly Taylor: Kafka, .
Dave Mariani: So, so, so, Carly, what advice would you give somebody who was interested maybe getting into the gaming industry
Carly Taylor: So I would say, you know, specifically for people in data, cuz those are the people I talk to the most mm-hmm. mm-hmm. , I always talk about the importance of having data projects. I think that, you know, if you are new to something or you lack some serious experience and depth, the best way to show your talent is to have projects that you can speak to something that you’re passionate about, right Mm-hmm. . So when you’re talking to a recruiter, you know, you’re like, I got this data and I did this and that excitement is something that comes across through interviews. Yeah. Yes. People love to see that you’re passionate about what you’re doing. Yeah. So I would definitely recommend like actually grabbing some real game data if that’s what you want to do. there’s open data sets from PUBG on Kaggle, same thing with Dota.
Carly Taylor: Mm-hmm. , I think there’s some Call of Duty, data sets on GitHub actually that you could, the call duty World League that you could grab and play with, and then just do some sort of project with it. If you’re a data analyst, right. Build a dashboard, do some data analytics, find some insights. If you’re a data scientist, build some sort of predictor, right Mm-hmm. , I think the Kaggle data sets were actually machine learning predictor. like the original challenge was asking for predictions to find who’s gonna win a game. Mm-hmm. . And so that’s a perfect opportunity, right You’ve got the question built in for you. It’s easy to tie back to the business use case. That’s something else I say like, as you’re doing a project, think about what would this do for the business. Mm-hmm. like, am I just building an algorithm for fun Mm-hmm. , because I think it’s cool. That’s awesome and it’s great when you’re in school, but you need to always be tying it back to a real world use case. That’s something that’s gonna help the business move forward. and a lot of those are, you know, if they’re already set up with a cool question that the business asked you, like they care about it for some reason. So that’s an easy win for you, .
Dave Mariani: Yeah, that’s a good, that’s great advice. and you know, I did, I actually, I actually did put, put a, a demo data set together for AtScale for using Dota Dota. Oh, cool. You did yeah, I did. That’s awesome. It was amazing. It’s like, it’s like the, the Twitter fire hose, but for gaming, I mean, it’s like, it’s a, it’s an amazing asset. So yeah, that’s a great advice. Go out there and that’s cool. Consume real data and do something
Carly Taylor: Interesting. I’d love to see that demo. You should link it
Dave Mariani: . Okay. Alright. Right. Yeah. Maybe I’ll, I’ll resurrect that and refresh it a bit. , it’s been a while. so, so just, okay. So we talked a lot about gaming. There’s a lot of stuff obviously going on and new technologies and a lot of hyped or overhyped technologies, as, as we know. so let’s just talk about the metaverse for a second. I wanted to get your opinion about whether you think that’s real or, or hype or, and where do you think it’s going, if anywhere
Carly Taylor: I think it’s a little bit of both, honestly. Mm-hmm. , like, I think that, the hype comes from people overestimating kind of like how soon we’re gonna get there mm-hmm. or like, how technically ready we are for some of these things that, you know, I think like people are promising investors, right Like, we hype up, like, oh yeah, yeah, the metaverse is gonna change everything. It’s gonna be this revolution. but it’s not coming next year. I don’t think, at least, you know, this is all me speaking very personally about what I see in, you know, the industry as a whole mm-hmm. , but I do think that there is an opportunity here and I, I worry that it has been co-opted by people that most of us maybe aren’t primed to trust. Mm-hmm. , you know, I, I feel bad, but I’m calling Mark Zuckerberg out personally.
Carly Taylor: On this one. I don’t think that he’s a beacon of, of what we would consider someone who’s like, like a trustworthy person who we wanna give all of our data to, right Yeah. Yep. That’s just kind of the way that it is. He’s gonna have to just accept that . Yep. I agree. Maybe isn’t even his fault, but, you know, like, get a PR team just like accept where this is going. Like no one wants to think that you are gonna own the future of the internet. Right. Because from what I see, what I love about the internet, and I think what a lot of people love is that it’s not owned by one company, right Mm-hmm. . Mm-hmm. . Like, you can go to different companies websites, but like aside from what people are alleging about Google, they’re probably the closest to owning the internet that we’ve seen mm-hmm. . And I think a lot of people resent that. So when we have these big players coming in and talking about like rebranding themselves as meta, right Mm-hmm. , like that to me is a lot of the hype mm-hmm. and it does a lot of damage unnecessarily.
Dave Mariani: yeah, I I’m with you on that. So let me ask you, as, do you have a VR headset I mean, do you
Carly Taylor: I actually don’t, which
Dave Mariani: Is crazy. You don’t, I mean, no, to me that’s like, that’s, that’s, that’s all needs to be said, right , it’s like, could you think that the first killer app, well, I know there’s another killer app that always comes first, but then Exactly. But gaming is second, is always the killer app for, for a new technology and, how come we’re not seeing more VR games Or am I not not close enough to it Carly,
Carly Taylor: You know, I think that there’s, my idea of the metaverse is that it’s all about concurrent experiences. Mm-hmm. instead of me going to, you know, email@example.com and looking at what you have going on, I could be somewhere in game or doing something in a world with other people and I could run into you or run into a net scale rep and see some things about it just in my day-to-day, right Mm-hmm. and, you know, not like in a spammy advertisement way, but just in a, I’m interested in this, let me go over to this part of the world. Mm-hmm. , let me see what’s going on over here. Mm-hmm. . and so I don’t know that VR necessitates that. Mm-hmm. , we already have a lot of like, like Call of Duty, right Like, we have these big world games with mm-hmm. 150 people mm-hmm. all experiencing something together. Mm-hmm. , I’d say that that’s like a fledgling metaverse, right Mm-hmm. like everyone is having some sort of concurrent experience. We’ve got the server authoritative state, right That game state was what I would call the metaverse and everyone’s connected to it. The question comes like, how do we scale that Mm-hmm. . So I think VR is a piece of it. I don’t know that, I think it’s a necessity. and my version to vr, just like very personally, I get super seasick.
Dave Mariani: . Yeah. .
Carly Taylor: That’s like, I need drama on boats. Like, I, I don’t know what it is. So for me, I’ve always been kind of like, I I, I haven’t wanted to do those things immediately just because of that kind of vertigo feeling. Yeah. But if they could fix that, like, I’d be super into it. But, you know, I also kind of don’t like the idea of having like my ears and my face covered. Like, I think it’s a very vulnerable
Dave Mariani: Position to be. Yeah. I am. Just think about bumping into furniture and stuff. It just, it just seems very unnatural, doesn’t it Yeah. Yeah. Yeah. Yeah. yeah. I like the, I like your take on that, that, you know, online gaming really is creating the sorts same kind of shared experience that you sort of would, would think about in the Metaverse. So, and, but we don’t think of it that way. Yeah. Cause it’s sort of always been there, but, yeah, I like your take on that. I don’t have a VR headset, my son does, but,
Carly Taylor: Does he like it
Dave Mariani: You know what, he doesn’t use it that much. Yeah. There’s maybe one game that he does when he does put, and he’s a big gamer. Yeah. and, but it’s, so there’s only a, a, a one game that he actually, you know, is engaged with with, and it’s another first person shooter kind of game. Oh. You know Oh, that’s interesting. but I’ve,
Carly Taylor: I’ve heard Beat Saber is fun. I think it’s like a techno like
Dave Mariani: Yeah. Sword. Yes. ,
Carly Taylor: I don’t really know what it is, but
Dave Mariani: Yeah. that’s Don, I hear that’s like, yeah, that’s, but, but I don’t know. Apple’s gonna come out with a headset supposedly this year or soon. Oh, stay tuned. Maybe they’ll, maybe they’ll fix everything. I
Carly Taylor: Don’t know. I guess that makes sense cuz haven’t the Samsung phones been like, you can strap it to your face Yeah. to be kind of like a VR
Dave Mariani: Headset, . It just sounds, it sounds very unnatural. That’s not gonna be mean. I’m telling you, you’re
Carly Taylor: Not gonna be walking around
Dave Mariani: . I just, I, yeah, no. so, okay, what about, what about, what about chat G B T and what do you think about that
Carly Taylor: So it’s okay. The first time I used chat, G B T I work in ML and I was like, holy expletive. Yeah. This is crazy. Yeah. I was taken aback by it. And as someone who kind of like understands the state of the industry, I wouldn’t say I’m an expert by any stretch of the means word, but like, you know, cursory overview, not blown away by things too often anymore. Yeah. completely floored. So in reflecting on that moment, I’m thinking, all right, I think the reality of the situation is that mm-hmm. , probably Google, Microsoft have bigger and better foundation models than chat G P T, you know Mm-hmm. , my guess would be that in-house they have some things that are even crazier, right Mm-hmm. mm-hmm. . But what I think happened here and I think was really smart from open ai was that they beat them to market and made it free use mm-hmm. . And so they set the precedent for what people think of when they think of something that blew their mind recently. Right. And I think we’re all kind of like, have some level of like technology fatigue mm-hmm. . So like, you know, for me, like an elder millennial, it’s not easy to blow my mind with things anymore. Like I lived through texting, becoming a thing and like the first iPhone, right Mm-hmm.
Dave Mariani: . Yeah.
Carly Taylor: So me too, for this to be the thing that blew our minds that might not even be the, the best thing out there. Mm-hmm. , I think still set them up in a way that like, they’ve done a really good job of like becoming part of the zeitgeist and just being like what we think of when we think of crazy ai. Right
Dave Mariani: Yeah. You know what it means. Like it’s, AI has always been something that you talk about in movies and the like, Uhhuh, , and of course us in our, in our industry, you know, we’re using machine learning to do things, but, to put it into, into a package, that a consumer can actually use and, and, and, and be impressed with. I mean, I was blown away too. Yeah. To me it was like, it’s almost like the internet moment for me. Yeah. where it was like, you know, just yesterday I said, you know, write me, an 800 word article about the modern data stack it did it. And it was like, of course it’s not very interesting. Oh, well, it’s, it’s, you know what It’s, it’s a good synopsis, but it’s, it’s, it’s, but it did it and it did it and, you know, you know, in, in about three minutes, you know Yeah. And that’s, that is insanely impressive. and I think that, I think that that maybe search, I think search and the way we think of search today, I think that’s gonna change. Right. I
Carly Taylor: Totally agree. Yep.
Dave Mariani: and so if I was Google, it’s like, you know, you always say, you know, like, you know, Google has like, what 85% of the market share for mm-hmm. for doing search and how would they gonna get disrupted Yeah. This is an example of eventually something comes along that is not gonna beat ’em at their own game. It’s just changing the game.
Carly Taylor: Exactly.
Dave Mariani: Exactly. And that’s what, that’s what this could be.
Carly Taylor: Yeah. I totally agree. I totally agree. I wish it did. As I was using it, I was thinking, I wish this had links to sources. Yeah. And then it would be basically a search engine, right Yeah. And
Dave Mariani: It’d be absolutely one. Absolutely. Yeah. Like, lifted up Microsoft, that’s probably, you know, the big investment. That’s probably where it’s gonna go, Carly next. Yeah. but, yeah, lately I I it’s hard to get on it actually. It’s always busy, so, yeah. Which means it’s being obviously successful out there. Yes. okay. Well the, this is amazing. I love your perspective on these new technologies. I love your perspective on gaming. so Carly, where do you think things are going from here Let’s just put on your, put on your future hat here. and what do you, what do you think is gonna be, is gonna blow us away or we’ll look back five years from now and look back to this moment and think, wow. is it chat g is it chat g p t or is it something else
Carly Taylor: I think that’s one of them for sure. Mm-hmm. , I think that, you know, all the things that we’ve talked about, the ways that that will change, just the way we interface with a search engine, right Mm-hmm. . Mm-hmm. . Now think about all the ways you interface with a computer. Like when you wanna, you know, I’m on Linux so I’m always trying to change things cuz all it wants to do is break itself. So let’s say that I could like have some sort of chat G P T interface that’s like, please fix my Nvidia drivers for the hundredth time today. Right Like, and it could just go off and find the right driver, download it like, like make sure everything’s working, run a diagnostic, right Because it understands what I’m looking for. It could go on a forum and say, oh, I can see here the latest release is broken. I need to roll back, you know Mm-hmm. one or two releases and do that by itself. Like that to me, I’m thinking even coding, right Like I do a lot of the same boiler plate code. What if it came in, you know, I think we already see this with a, what’s it called with Microsoft, co-pilot.
Dave Mariani: Oh
Carly Taylor: Yeah. So like, you know, up your AI pair programmer, right Mm-hmm. , it’s not gonna do everything for you or do a lot of the heavy lifting in terms of like figuring out the best way to approach a problem, but if it did the boring stuff, importing packages, you know, all that sort
Dave Mariani: Stupid stuff. We don’t do all the automation stuff. Yeah. The dump
Carly Taylor: Stuff. How much more like yeah, how much more could you get done How much more productive could you be And so I think we’re gonna see a level of productivity in everything that we basically do that interacts with the computer that we’re probably not prepared for yet. Yeah. . Yeah. And then it’s gonna make us all really lazy. , I’m be like, download a driver, what do I look like
Dave Mariani: But yeah, no, because then we’ll have more time to do other like higher valued stuff, Carly. Exactly. That’s, that’s always the, that’s always the idea, right Hopefully
Carly Taylor: I’ll have more time to scroll dog videos on Instagram
Dave Mariani: .
Carly Taylor: That’s just, that’s my ultimate goal. .
Dave Mariani: There you go. hey Carly, it’s this, you’ve been great and, this has been a great discussion. Thank you so much for, you know, for taking the time today to talk with, our listeners and, and you have a, a, a great background and a great perspective on where things are going. So, so everybody follow, follow, Carly, you, you’re a prolific, writer, , poster, about the industry and, and data and machine learning. So, check out Carly on LinkedIn. Thank you all.