HH82-20-05-22.mp3

Harpreet: [00:00:09] It's up, everyone. Welcome, welcome to the art of the Data Science. Happy Hour. Friday, May 20th, 2022. I'm super excited to be here again on Friday, every single Friday for the last 18 weeks. I think we'll be doing this each week at this, hanging out with each other, taking it virtually. But now things are coming back and people are going to travel. More live events are happening. Speaking of live events, there's a few coming up. But first, if you are in Denver. Or if you are from there and you're going to be there next week, May 23rd to the 26th, let me know I'll be in Denver during that time period. I'd love to meet up hosting ML Ops Community event on Monday, May 23rd at 4:30 p.m. at Denver Beer Company, and that is on Platte Street. So if you're able to make it out to that, please do then. Be excited to see you guys. So my Denver people holler at me. The Denver Data dude who used to be known as a Seattle data guy. I'll be hanging out with him, so I'm looking forward to that. Also supposed to be connected with and looking forward to that as well as be fun man. Excited to be there. Huge shout out to my younger cousin who graduated to the state university or something that teaches innovation right there, got his bachelor's in electrical engineering and told him to start pursuing some or learning some pipe on these people because it would be indispensable without that huge shout out to the sponsors for today's episode, the MLPs World Production Conference.

Harpreet: [00:01:46] You guys need to check this conference out. Look up to me with all this, this vague, mystical, dark art, I was just always a notebook data scientist, happy, working out of my notebooks and to a certain extent, a [00:02:00] lot of that, a lot of stuff kind of still is a dark art, but that's probably because I don't come from a background and I never felt like it was something that was necessary for me. But the more and more I began to learn about and more began to learn about this nascent field, the more exciting it is to me, the more energized that I feel by digging deeper into it. And I can't wait to learn more from June 7th through June 10th at the end of this World Machine Learning in Production Conference in Toronto, Live and in person in Toronto. Yours truly is going to be there. I'll be presenting a demo of Pachyderm, my first time ever presenting live in the studio audience and first time demoing the product. Not to mention that I've only been working a month and a half, so I'm sure there's nothing that can go wrong with me doing this, but I'm excited and happy to be doing it. So this conference can have over 300 machine learning and AI teams from all over the world. They're going to be sharing best practices around machine learning production. It's going to be a ton of virtual and hands on in-person workshops to help you build up your ML skill set.

Harpreet: [00:03:11] They're going to be talking about how to build an HTML platform from scratch, how to build real time machine learning features with a future platform model deployment. Ray How to build production. Ml Monitoring from scratch. A beginner friendly crash course in Kubernetes. Kubernetes. Kubernetes is awesome, but it is tough to learn trust because it can't be asked. But with this beginner friendly course, I'm sure you'll be picked up with no problem. Lot. A lot. A lot more going on at the conference. Be sure to check it out. We've got sessions even aimed at helping you understand this strategy while driving the technical aspects of machine learning as well. So check it out. Machine Learning Ops and Ops Conference and Ops Introduction in Toronto, June 17 [00:04:00] to ensure that our guys in in-person events are live back. There's a couple of people here that I've met in person can be one of them surge as well definitely strong person. One of these days we'll meet Russell and Maggie and then Eric and vision in person as well. But super excited to kick this off. If you're watching on LinkedIn. If you're watching on YouTube or on Twitch or wherever it is that you're watching and you've got questions. Feel free to drop your questions right there in the chat. Be happy to get to them. But we're talking about in person and just like being a person nowadays. What's lost with virtual communications? What is lost? What is something that gets lost in translation with virtual communications can go for?

Speaker2: [00:04:55] So I think right there is a perfect example where I can go for it and I have to unmute myself. I have to rethink about what I'm saying. There's this little bit of friction that doesn't come as naturally in conversation. You have to sort of wait for other people to finish talking rather than continuing thoughts. And I think for me, that's a really important thing when you're riffing off other people, when you're doing idea generation, not as big a problem when you're trying to report something. But I am someone who very much feeds off of the creativity of others. And the kind of snowball effect is very powerful for me, and I feel like that's very, very difficult to do on Zoom or on any of these other platforms. You know, I was with a lot of creators the last couple of weeks, and it felt like there was just like a spigot going on with the ideas just continually flowing. And I think it was just because there was that lack of latency between one idea flowing into another that that I really relish. And I'm trying to figure out ways to do more and more of this little hard being 1000 miles away from everyone else. But [00:06:00] it probably means I'm going to be traveling quite a bit more.

Harpreet: [00:06:05] Ken, thanks so much. Let's hear from from some surge on this and somebody else wants to chime in. I'd love to hear what you think is lost with virtual communications.

Speaker2: [00:06:19] Well, people don't understand your context as well, because if you're in the same room, you're it's the same lived experience at that moment in time. But when you're somewhere else, you don't understand what's going on. You know that perhaps my dog is distracting me or, you know, like something else is going on at the same time. And so it's it's really tough at the same time. I think it is it does open the door for distractions. Speaking of distractions, which is not as present, you're like, I think you're more present when you're with other people in person. I think it's really hard to be equally present when you're not.

Harpreet: [00:07:05] Yeah. It's very hard to like, just pretend like you're paying attention when really you're looking at a bunch of tabs on a screen. It's easy to do that through virtual communications, but yeah, that's 100% agree with you on that. Definitely are forced to be more in the moment and actually listen to the person. Let's hear from you, by the way, who shut out everybody else in the room with you again. Vijay, let me finish just the Russell ten and also Eric Sims, who is going next for. All right, maybe I'll just be contrarian. So I think what Ken is talking about is totally valid. And it's a pain that I have to wait until, you know.

Speaker3: [00:07:45] There's that time when two people kind of.

Harpreet: [00:07:47] Talk over each other for a second, and then we all kind of have to, like, stop and wait and then, like, nose goes for who's actually going to say the thing that they were going to say. And it's a pain. But I think part of that is that Zoom is like the dominant [00:08:00] tool.

Speaker4: [00:08:00] In the.

Harpreet: [00:08:01] Space. Like there are other tools like proximity chat that do. People can talk over one another just fine. You can ha ha ha ha ha. And people will hear you. And it's not like unmute. Ha ha ha. And then let go back to silence or whatever, you know. And so like I think that there are tools that could be used that aren't used like proximity chat is so cool. I don't know if ever use that, but it's cool. Like in that, like if I turn my head this way and you're behind me on the screen, you won't hear me as much versus and if I move my character closer to you, you will hear me. And if I move away, you won't anymore. So in that way, it's more like, you know, like real life. And I can't wait until, you know, I can meet in. I like VR, so I want to meet in VR and be able to.

Speaker3: [00:08:45] Maybe have a photorealistic avatar like I have right now.

Harpreet: [00:08:49] We'll see. I don't really care so much about that. But so that's that's one thing. And then the other thing about context, though, is I don't think people are paying as close of attention as we're like saying like, oh, they're they're forced to pay attention because they're in the same room. Like, no, they're not. People are tapping you, tapping away on their computer. They're distracted all the time. I know that for sure, because I'm distracted all the time in person, like meetings and stuff too. So like, I'm pretty sure I'm not the only one anyway, if I am sorry to everybody who's ever been in a meeting with me. But if not, then we're all human. But I do agree though, like where we are right now, we definitely still have room to improve, especially like in the. I mean, asynchronous chat.

Speaker5: [00:09:36] Asynchronous is a is.

Harpreet: [00:09:37] A big thing, like if you want to know. It's like, do you write out your thoughts clearly enough that somebody can grab your train of thought, answer your questions, maybe even run with it? If you suddenly had to log off from work for two weeks and were you able to like compose your thoughts off? It's just like trying to read somebody's comments in [00:10:00] a in a half written notebook or something like that. And I think I think that we have a lot of room to improve and grow in there because, I mean, we've been at it for a couple of years of trying to like being forced to like work together virtually. And I think we just have a lot of room to improve and I hope I'll be embraced that. Thank you so much. Yeah, it's interesting. I was talking to Alistair Kroll, coauthor of Lean Analytics, and this is the very, very beginning of the pandemic, I think were probably talking in June of 2020, and he had mentioned that Zoom should not be number one right now, but for whatever reason it is, which is interesting. I did not know that there were alternatives. The only alternative I know to zoom is that is Microsoft teams, and that's not that great an alternative. Speaking of conferences, they said we should create our own conference and line up sponsors and that can be hard, sell well hopefully can and I make this happen via the data community content creators would will not be doing it with Kate this year. Can I and talks to see if we can make this happen so that we will be bringing more details. Let's go to Russell and Dr. Russell and then put you on there while we wait for Russell. And then I would love to hear from you as well. So we'll go Russell, Vijay and Russell. Thank you.

Speaker5: [00:11:38] Harvey. Hello, everybody.

Speaker4: [00:11:40] So I've got a two pronged approach to this.

Harpreet: [00:11:43] Firstly.

Speaker4: [00:11:44] When you're in the company of somebody in real.

Harpreet: [00:11:46] Life, you you get to you get to.

Speaker4: [00:11:49] See the benefit of all of the, you know, the the micro actions of the person. You get to see the whole person's body. For a start, you have.

Harpreet: [00:12:00] Can [00:12:00] you hear me? Okay? Yeah.

Speaker4: [00:12:03] So you have ultimate confidence that they're wearing pants or trousers or a skeleton, which is always a good thing. But, you know, body language and, you know, the micro facial expressions and pauses in the speech, it's just it's better to be.

Harpreet: [00:12:18] You know, have that without.

Speaker4: [00:12:20] The latency of virtual communication, which is usually great, but not always. So then the other side of it is dependent upon the the tech that you're using. You know, you're reliant upon the ISP connection.

Harpreet: [00:12:36] Reliance upon your actual laptop.

Speaker4: [00:12:38] Or tablet or phone.

Harpreet: [00:12:40] Or whatever you're using on the app.

Speaker4: [00:12:42] Or the software you're using, and then the hosting system of the app or the software you're using. So there's.

Speaker5: [00:12:47] Far more.

Speaker4: [00:12:49] Opportunities for error in that. And um, a perfect example of this, I mean, if you, if you look back on any of the.

Harpreet: [00:12:56] Times I've been talking on these gems, more.

Speaker4: [00:12:58] Often than not, my camera pans me out. And I can assure you all in real life I don't just disappear in the middle of a conversation. You know, I'm present for the for the whole thing.

Harpreet: [00:13:11] Russell, thanks so much, James and Vivian. Let's go to Gideon and then we'll go. Yeah. Can you hear me? Yeah.

Speaker3: [00:13:26] Al-saadi, I was on mute. Yeah. I'm glad to join you. I mean, translate.

Speaker2: [00:13:30] I was driving from Pennsylvania to New York City, so it was a break.

Harpreet: [00:13:36] So I thought, I'll just drop. Just to hear.

Speaker3: [00:13:40] It was a good topic. Some of the points was already covered. To me, the important.

Speaker2: [00:13:46] Person that we missed in in the.

Speaker3: [00:13:49] Virtual meetings is the one you are in the room. You can see whether someone is getting your idea or not. Right. And you can change [00:14:00] your thought process if somebody is not getting it, maybe you want to say in a different way if you are on the virtual call it, since you don't see you don't know whether the way you are communicating is is getting across to other people or not. Right. So I think that that's the biggest piece and other comment that was made earlier. I have noticed and I am myself culprit of that is that when we are virtual, we think we are in the meeting, but we are doing multitasking. We have the time and I choose to do multitasking where if I am my voice important, I, I am not multitasking. But if I'm just listening, then, you know, I can say I'm still going to listen and I can pay attention basically to do that. One of the thing that in our office environment we are using it, Moodle, I don't know how many, how many of you are using Moodle, you are AL which is basically a tool. You can work in a very interactive way through that is a chatting and you can draw whatever you want other people are able to see. At the same time they can zoom in and zoom out. It's pretty powerful. We have been using in our office fairly regularly. It's pretty powerful tool to get that way. You can see who is interacting or not and literally whoever is in that room, what they're drawing and doing, you can see who is doing what. So then you can see that everybody is engaged in that discussion and idea generates and all those things because they are pulling their cursor. You can see everybody is there. So I think that is one of the tools I have seen effective.

Harpreet: [00:15:42] Thanks very much. I will have to look into that. A lot of great technology out there making this virtual, working easier, more streamlined. Thanks for sharing that one. Let's hear from Vivian. Good to.

Speaker3: [00:15:55] See you. Yeah. Hey, everybody. I missed [00:16:00] the beginning of the question, so I hope and like some of the earlier responses. So I hope that this hasn't just been said already, and everybody's going to think I'm done. But I actually think that there's lots of things that are better. Virtually like I. Did like a whole data science boot camp and stuff right after I started that, right after COVID began. And I liked it better because it was easier to like code along and stuff like the instructor would like put his screen up on Zoom and then you could have your screen right next to his screen and like be coding along and stuff. And that that's the sort of thing I like a lot. Instead of having to like hover behind someone's shoulder while they're like clicking on things, trying to see what they're doing or something, I think that's legitimately easier to do virtually and have someone screen up on your monitor. And I was also going to mention Mural, because I think that mural is a use case where I legitimately find it like better and easier to collaborate than doing like physical sticky notes on a whiteboard and stuff because that's, that's at work.

Speaker3: [00:17:07] What we have used mural for is like those kind of like brainstorming events of like, okay, everybody like come up with ideas, let's like generate ideas. Let's everybody put your ideas on a sticky note and then we can like group them together and find themes and discuss the ones we like best. That sort of stuff that you would normally in person do with physical sticky notes and then be moving around and stuff. But mural is so nice. I like it so much better. And then you have this record kept forever that everyone can refer back to. And like then. Then when I go about my work and I'm trying to think about what kind of analysis would be helpful for people, then I can go back to the mural and see what were the things that everybody brainstormed about what we'd like to know more about with our product and stuff like that. And then I have that record of that mural. It doesn't get lost with these like physical stickies that some [00:18:00] manager takes and puts in a drawer or whatever anyway. So yeah, those are my thoughts.

Harpreet: [00:18:08] Thank you very much, Ken.

Speaker2: [00:18:12] Yeah. I just have a thought. Not, not to like to go contra to what you're saying here, but isn't sort of the the benefit of these tools is that we can use them asynchronously, asynchronously, and we can also use them in person. Like, aren't they more powerful? Or I've never used Miro, so I could be completely wrong. But isn't there an added benefit of being around people and using this because you do get that continuity? Same thing with maybe even like coding in a classroom setting. I actually 100% agree with you. I prefer I went from exclusively in class education to online education and got a far better experience. But I thought that that was the fault of the organization and the professors. Right? There is a definite way that your professor could be coding on screen and be sharing their their screen with you on your computer in the classroom setting, as well as on the the non classroom setting or like the virtual setting as well. Right. So I would. I would. State that these tools can be just as useful in person as they can be asynchronously, and I quite enjoy using them in person. I'm also very extroverted. So I found that that's a very, you know, an amplifier on my part. One thing I did want to bring up is I forgot what it was. It's the 738 55% rule that I.

Speaker4: [00:19:36] Just recently wrote about.

Speaker2: [00:19:37] I'm reading Never Split the difference. And apparently 7% of information is communicated via our words, 38% is communicated with our tone of and 55% is communicated with our body language. So something that I've been doing, my my camera zoom is actually a little further out than everyone else's. And that's because [00:20:00] I find that there is extra value communicated by just having my entire body in frame rather than just my head. That's something that I think I don't know if it actually works. There's no concrete evidence that I've been able to communicate better because of that. But it's just something little that I picked up associated with the virtual communication part that is hopefully starting to blend a little bit more with what we would experience in a person to person setting. So take that or leave it. But I thought that was sort of an interesting observation or, or or look at something.

Harpreet: [00:20:37] Like that too, kind of moving through the way and then kind of expose more like a sweet post up. Let's hear from you.

Speaker5: [00:20:48] So when it comes to like this, I try to differentiate between like work and say conferences, right? So now with work, I'm liking the hybrid style, right? Using tools like we use Miro neuro board, which is probably similar to mural by the sounds of it. We use that at work occasionally for like retros and stuff where we want to brainstorm and being able to do stuff digitally is quite powerful because we have, we have experts in Sydney as well as people in Brisbane and people in Melbourne. Right. And we're able to work whether we're in person or whether we're in, in, in interstate, we're able to work well as long as we're all at the same time at the same board, we can access it digitally. And even when we're in person, we just have a big screen in front of us with that on and each of us has our laptop, so you can still interact with it, right? And we've found that that works pretty well either way, just like I'm saying. But then there's the other side of it, right? There's conferences. I'm in Australia. Do we get machine learning and data science conferences? Basically. No. Right. We've just never had had them. We've had one [00:22:00] or two. I think Europe was supposed to be here in 2020, but the pandemic had different, you know, different ideas about that. But like we rarely get conferences and the value of having conferences in person is massive.

Speaker5: [00:22:13] Right. And doesn't matter how many digital conferences are now available to us, it does still take away some of that interaction with the rest of the with the rest of the community at a direct level, like a lot of it, it comes down to just physical networking, right? I did a GTC, a couple of the GTC full day courses. Now, if I was doing that in person, just the conversations that happened in person would have been very different to and much less limited to the conversations I had in the side chat with one or two other people. And plus, it was at ridiculous hours, right? It was like I was starting my day at 11 p.m., finishing it at five, five in the morning. I would rather have flown out to the conference in person to do it. Right. So there's there's all these things that are that are slightly different. And I think we are still missing that. This is the other side is we do have other conferences and forums like like this one, right? Pre-pandemic, it would have been much less likely that I would have looked to things running internationally to say, hey, let's, let's take part in that. So there's always like plus or minus, but I don't think we can ever get to that point where we're completely replacing in-person conferences with a digital component.

Harpreet: [00:23:35] Anyway. You have got to make this happen in Australia. You have the power to do so, he shouted. Everybody else in the room to avoid his ability to get happy here. He goes, also here. What's going on? So just for you guys, we're talking about what's missing from in person communications. And it's interesting because Ken just released a video earlier today with Tina [00:24:00] talking about the title is the value of In-person Communications or something to that or beaten up. Definitely. Check out Ken's video. Finally, I'll drop the chat for you here. Let's go to then let's hear from Ben and then get a cool question coming up from another here that will touch on Levy's question has to do with senior folks like 10 to 15 years plus pivoting into data science leadership roles. But it still needs to be asked questions like Tableau and things like that. So, yeah, keep an eye out for that. We'll go to go to Levy to wrap up this conversation. Go for it.

Speaker4: [00:24:42] You know, I guess my perspective is kind of weird. I worked with really large companies before the pandemic and even before I started my business. I mean, back to like the Polycom. I don't know if everybody remembers them. Like those triangles that you had on the conference room table where you would make phone calls and there'd be I mean, I'm kind of going back there, but like that was the original. And if you worked in a multinational like that was especially when we were doing the offshoring and outsourcing, you know, in the early 2000s, it was that was how you communicated because you had a meeting at the end of the day to hand off your work to somebody else in a different time zone so that you could work 24 seven, basically developing 24 hours a day. So you'd have a handoff meeting at the very beginning where you took on work and you'd have a handoff meeting at the very end of the day where you handed off work to somebody else and you got really good at no video at all having to do it over the phone. And you couldn't like we would when phones finally got cameras on them and you could send pictures via email through your phone, that was like groundbreaking because I could do whiteboarding, click a picture and send it to somebody. Before that, we were doing like screen screenshot shots in paint. So I mean, I think we [00:26:00] got really good at having no tools. And then Google Hangouts showed up and Microsoft Teams in the video chat sort of showed up and help make remote possible. And I've been working remotely since I started my business in 2012 and being in regional offices, being in corporate offices with Multi-Time Zone regional offices, I think some of us just got good at it.

Speaker4: [00:26:29] And so when the pandemic hit, we just transitioned into something that we knew pretty well. And so when I look at what did we lose? You know, as long as I've had an existing relationship with people like I've met them in the past or I've worked with them for a very long time, didn't really lose anything because we knew each other, we understood each other. And that's kind of the nice it's the nice part about having long term clients is just really having those long relationships where everybody knows each other, understands each other. And even when I'm dealing with new people, the somebody can clue them in and go, No, no, no, he didn't mean that. Or So this is some backstory and some history. And I think that's the difference now is there's we've got a couple of generations that didn't grow up that way. We've got a whole lot of startup generations that, you know, all they've ever had is that office where it's like that office and nothing else, or maybe two offices total. There's this chasm because I know the people that I work with at large. Companies like the difference between work from home and being in the office was almost zero. We didn't really notice it that much. So I think that's I guess that's the perspective I wanted to bring in is really hearing people say it's different and there are things lost. It's kind of strange to me because I didn't. I think I just live in a world where that didn't happen.

Speaker5: [00:27:54] Can I ask you then just to kind of piggyback on that? How much of that is to do with like [00:28:00] how long you'd been at that company or with those relationships? Right. Like we're seeing the cycle rate of people moving companies a lot higher now than it was, say, 10 to 20 years ago. Like significantly so you're seeing the average lifetime of a developer at a company maybe two years, right? Two or three years even at really good work culture companies, just because that's the way things are going these days, right? Like I started a job, Mid-Pandemic and I ended up moving to Brisbane to get to know people in person more, you know, and that helps. And then now I'm comfortable to go back now that I know them in person a bit better, I'm comfortable going back to Sydney or whatever and work remotely and that's all fine. But I needed that initial extra, I guess, interaction before you can get to that point. So how much of that is relying on that existing? Hey, I've been working with these people for ten years and I know them really well.

Speaker4: [00:28:54] You know, it's a lot of yeah, it's definitely established relationships, but I also work with a different age group for the most part, which is you learn differently generationally because technology was different and we had to do, like I said, with the, with the Polycom, we had to figure it out over the phone. And, you know, you learned not to take things personally because you had no idea what the context was. And so you'd have side conversations. You know, we would have sometimes two or three conversations going on. And I think that's it, is that we learned how to make relationships with people that we probably would never meet. I mean, there were offshore teams that I managed for two years in India that I never met any of them, which was lame. But, you know, there just wasn't a travel budget. So I never got to meet any of the people on my team. Like I said, it was lame, but we made it work because we had to. We just figured it out. So I think, like I said, I think that's the difference is yeah, it's there's definitely the length of relationship that I have with clients, but it's also the age group that we're all in. [00:30:00] And the fact that we had to go through the sort of remote adolescence, we had to make some really janky stuff work, and we made all the huge mistakes that, you know, everyone's sort of learning from now. We made them all with way worse technology, and so we had a longer time period to figure it out. I really think it's just an old person, you know, different perspective, that being my age and I guess growing up with the lack of technology, I think that's, you know, as much as it's rarely a benefit. I think in this just one instance, being old was good.

Harpreet: [00:30:38] Back in the day. They used to call it telecommuting. Telecommuting. Yeah. If anybody else has anything to add to add him. More than happy to hear from us. Hugo or Tino or Tom, if you guys want to chime in here. No, I was talking about kind of the pros and cons, I guess now of the virtual communications, I guess. What are you looking forward to you? How do you contrast this in person versus alive kind of scenario after knowing so many people, I guess, from virtual. Same here. I did not know any of this if it was not for virtual, which is true. So it's really interesting question is what's it like meeting somebody from camera to to in person? I guess that's an interesting thing. Everybody hasn't been safe, but no, if not good to discuss, etc.. Interesting question. I think that's a good debate. Ericsson says you find out how talented people are. Yes, that is true.

Speaker3: [00:31:50] So I think I had a conversation with somebody earlier this week who reached out to me and has a lot of [00:32:00] experience in consulting across different industries. Some of that is like buy it related who's kind of struggling with getting, I would say, a little bit more data centric roles on the leadership side. And as somebody who has at least 15, 20 years of experience leading teams. So somebody who's very capable and the issue that he runs into and I'm sure a lot of other people like him run into, is that he doesn't have some of those buzzwords that. Are some of the data scientists on their team would have like some things in Tableau or Python or machine learning or whatever, whatever, whatever, right? And it's a challenge because even though he wants to kind of pivot there, he can because he doesn't have those technical skill sets. Now, the problem is in his job, he would probably not be building tableau dashboards or coding. Right? So it's just like, do you know this or not? And I guess when you write a lot about this stuff and you know what is kind of a good way for them to be able to do justice to their experience, but also be able to pivot in that direction. And you know, I'll do what I can in helping him kind of share his resume or whatever. I think I can. But I think it's a legit question for a lot of folks who are. You know, and the industry is struggling to find people at this point. Like right now, it's just nuts right now to get anybody in anybody through the door. So there is this big disconnect of who they think they want and who can actually do the job.

Harpreet: [00:33:58] So the question I [00:34:00] guess the question is I could kind of slow it down is how can somebody who's gotten 10 to 15 years of experience and maybe a kind of a role where they were doing some type of analytics or bi type of stuff. How does that person transition to leadership of more technical people that have a different set of capabilities?

Speaker3: [00:34:23] Or like, let's just say in this case, he is he considers himself more as a bi slash i.t person. So he's good with analytics, he's good with numbers. He also is a storyteller. He likes to consult so he could pivot himself in a more predictive analytics space, if I may say so, right. I think I've worked with leaders like that who have come with similar backgrounds. But I think if he's applying to those roles, they may say, oh, he does not have machine learning on his resume and he does not have X, Y and Z on his resume, which is totally irrelevant to the job that he'll be doing.

Harpreet: [00:35:06] Yeah. Let's go to his hand up and then after J would love to hear from anybody else who'd like to chime in, but maybe Tom or Tina or Kiko then anybody that wants to jump in after J just let me know. Just like these two handlers, I'm stuck and a.q. Yeah, sure. Well, that's really a good scenario. I can I can see that playing well.

Speaker3: [00:35:34] One thing I can tell you is that. That person who has 15 years experience. Let's understand what is the strength that person brings in.

Harpreet: [00:35:44] Strong domain knowledge, right?

Speaker3: [00:35:46] That is strength. You cannot gain that easily.

Harpreet: [00:35:49] So he has.

Speaker3: [00:35:51] She has that strength coming with that much experience. The second thing is, I heard that story teller. That person has basically is trained there [00:36:00] as well. Right. That's a key element of that. And he she has a background in the I.T. and VA. For me, I mean, there is some fundamental work that has to happen. There's no shortcut. Right. So I would say that if someone really.

Harpreet: [00:36:16] Wants to get into you have to get into the.

Speaker2: [00:36:20] Water.

Speaker3: [00:36:21] Right. This is you cannot sit outside and just imagine things will happen. Right. So if you want to get that getting into the water, that person is senior enough that they can carve.

Speaker2: [00:36:33] Out a problem.

Speaker3: [00:36:34] Right. Take one problem. You have a domain knowledge. Let's try to solve the problem using data science. A simple one. It doesn't have to be complicated. Right. So I think a real practical experience, I would say that you've got to get the water, get your toe and hands all feet wet and start small and getting into that.

Harpreet: [00:36:56] I would say that.

Speaker3: [00:36:57] Get a mentorship, right. So even your senior, I can tell you that I have gone to learn people from less experience than I in a certain technical skill area, have a mentor selected. You could be a data scientist, you know. It doesn't have to be that since you are senior, you cannot have a mentor. The person who's a data scientist. Right. Very open. Have a mentor. Get the direction from that person. Lay out some plan. Basically, take a simple problem, especially if you have a domain knowledge, somebody working finance, sales or any other area, they can pick up a problem in that space. So I would say get someone as a mentor. And then pick up a business problem, get into the water, solve this problem so the value, get the confidence. And I think.

Harpreet: [00:37:44] Slowly and it got to be a plan.

Speaker3: [00:37:47] Right? It's not only three months or one month, there has to be a plan. So that's my advice. Okay.

Harpreet: [00:37:56] Well, you can always send him a recording. It's going [00:38:00] to be on YouTube live and direct right after this. It's also on podcast. And speaking of podcast episodes, especially conversations like this. So be sure to check them out. Let's hear from Serge. And then after Serge, then Serge or Tino or Kiko would love to hear from you to. It.

Speaker2: [00:38:24] Yeah, well, I think there's there's two kinds of leaders, in my view. There's really like the administrative kind of leader and then there's a technical like leader in which is really on top of all the technical details. Either way, I think that the biggest value brought in by by leaders of any kind is, is not the technical skills, it's the soft skills, all the things that communication and coordination, problem solving and so on. Of course, it's important that they understand the subject matter well enough to weigh in. So like it does vary, I think even though it's, it's probably more towards soft skills than hard skills. I think there's still has to be a little bit of hard skills if it's an administrative leader and a lot more if it's a technical leader and that that will vary. So I find it kind of odd that it's kind of seen as such such a requirement on that level and. I as as you were talking about the transitioning and that I kind of can connect it with my own experience because I have a lot of experience prior to transitioning into data science. And it's a really bizarre thing for me to be the way people kind of evaluate experience moving in is different as someone that's purely entry level, because someone asked me the question, you know, how much experience you have in data science. [00:40:00]

Speaker2: [00:40:00] It always tricked me up because I don't know what to tell them, you know, how long have I been exploring data? You know, how many how how long have I been analyzing, you know, doing statistical analysis on data or making reports or using python or machine learning, you know. So the amount of years I can equate to each one could vary anywhere between 20 years and, you know, like five, right? So it's really hard to say. But sometimes what they really mean is how long have you been in roles that say data science on it? And I think I mentioned this to Tom the first time I met him, but I mean, if you're valuing me for that, it's only like three years. And even then I've recruiter's recently, you know, I have reached out to me and and they, you know, I say, well, if I were to transition from the role I have right now to something else, it would have to be, you know, a step up. It would have to be something more leadership. And they say, well, we wouldn't consider you for leadership because you only been a data scientist for three years. Right.

Speaker2: [00:41:03] And my management experience has been in another field. Like if, you know, managing web developers is much different than managing data scientists, you know, I did that for six years, so why can't I do it, you know, for data scientists? So I find it really, really iffy the way these things are. So I think the two something's up. I don't think that the concern should be, you know, what skills do I need to prove? You know, it's just can can there's probably one skill that I think is more important to prove than anyone else. It's not like, okay, has to be an expert in R or an expert in Python or expert in, you know, every single BI tool out there. No, it's more of a question of can they be an expert in communicating findings, in organizing projects. [00:42:00] So in that sense, you know, doing like the end and end to end pipeline is a good exercise just to say, okay, I can handle an entire data science project. I know what it's involved in it. It's not so much, okay, I can I can be the best at the at the programing part or the best at the data exploration part. It's just more like I can make sure that the whole thing is properly orchestrated.

Speaker3: [00:42:27] I think that's kind of the maturity that is needed for the interviewers who are who are there. I had a I had an interview, I don't know, somebody few months back and he was like, How long have you been doing data science? I was like. Kind of all my life, basically. You know, it wasn't called innocence then. And then it was like, okay, tell me what you've been doing. So I said a few things and then he goes, But how long have you been doing data science? And I was like, What do you mean what data science? So it was like, so there's this lack of understanding of what is actually needed to do the job. So I totally agree with with what you're saying. But the the problem is either you're working with a recruiter or you're working with somebody who's like, a little bit more on h.r. They don't know those nuances. And if they don't see those three things on their resume, you lose like a lot of good candidates because there isn't somebody kind of have has the nuance of saying that. But I I've had my technical managers and non technical managers and they bring their own strengths in each way. Right. So it's easier to go to a technical manager and say, okay, I'm having issues with this model. What do you think a non technical manager would be like? I have this model that's not working. Can you get me somebody? Whatever. So but but I think we are kind of losing [00:44:00] a lot of good folks because we have some individuals that are that just are not experienced enough to understand those nuances. Yeah.

Harpreet: [00:44:13] Yeah. I think we had a similar conversation a year ago around the topic of the right candidate for tech leadership, somebody who's not super technical as an engineer, moving to a tech lead type of role, but the perception that teams have of that person. And I remember them saying that conversation was that for people who are engineers, they want their leader to be an engineer or somebody who has gone through those ropes because there's that credibility or something to that, isn't there? But let's hear from Kiko. And then again, I'd love to hear from you on this because we're not necessarily talking about like engineers or analytical data, professional type of stuff, but still you then have to make you go for it.

Speaker3: [00:45:06] So I do kind of feel like because we're kind of going through like something similar where we're or not quite. So basically we're going through hiring, right? And we're trying to hire at like the staff up level, you know, like minimum is senior, but we're actually trying to hire staff up and like there is a difference between staff and senior in terms of like. Scope of leadership responsibilities but also to like it depends on the company like so some companies for example like they'll kind of put all the strategizing like from a technical tech stack perspective, like into the tech lead or the staff and up sort of ladder, I guess. And so you'll have some companies where like there's an expectation that the manager is like both a technical leader and also does like the human element operations team side of it. [00:46:00] But there's some companies where like there's this expectation that like the manager is literally just there as like an extension of HR to manage someone's career and like all the technical leadership goes into like the staff up level. So one thing I do think that needs to happen is like when you're hiring or when you're thinking about promoting, like there does have to be a very clear ladder and there also needs to be a very clear set of expectations that is very, very transparent. If it's in the hiring, then it needs to be transparent. Among all the people involved in that panel, if it's like in the promotion ladder, then like I feel the company should kind of make that available. Like it shouldn't be hidden, what the promotion process is and what the expectations are going up like the different levels.

Speaker3: [00:46:42] Because I feel like a lot of times like this thing of like when there's that kind of ambiguity, there's also a lot of room for bias and it's also like uncomfortable on both like the candidate and also the person who's seeking promotion because like if you don't give them clear feedback, like for example, we wanted to see more in this category and if the feedback is also inconsistent between candidates, to me that kind of says like the person doesn't know what they're looking for and like they're potentially like biasing kind of both the promotion and like the hiring, the interviewing. It's just like some of this is like, resonating because like. We've also kind of struggled like both at the staff and the manager level to kind of figure out like what is the right mapping of like technical and people skills to the candidate that we want, both in terms of hiring externally and also like promoting within. But sometimes, like I feel like when there's not a clear like criteria, so some things essentially become like sort of just like dog whistles for like not wanting to incorporate like diversity of experience or like taking a chance on someone who might fully be capable for that role. So I feel very strongly on this, like there should always be transparency and people should always be aligned in terms of what the expectation [00:48:00] is for both promotion and for like bringing people in. Because a lot of times I feel like nontraditional candidates, people of color, like women, like they fall through the cracks. A lot of times when they could be perfectly good sort of leaders or managers and all that in those roles.

Harpreet: [00:48:21] Thank you very much. Let's go to them and then we'll go to Tom. And then, by the way, if anybody's listening, if you're tuning in on YouTube or on LinkedIn or on Twitch, if you have a question, feel free to let me know. Go ahead. Drop it in the comment section, wherever you are. I'll be sure to add to the queue and then we'll go to Tom.

Speaker4: [00:48:45] Yeah I think as far as where your friends at advice for trying to get into leadership without technical skills it it really depends on the organization that you're going up against. Some organizations, especially in finance, they seem like they're a whole lot more comfortable in finance promoting people that don't have the same kind of technical skills into leadership positions. But the reason for that is because the the technical side of the House is usually really commoditized. In finance and in banking, except for fintech, obviously. But for most of the larger banking companies, they're like I said, the the technical side is commoditized. And any company you see where they treat data scientists, analysts, engineers of any kind as commodities where they're trying to minimize the cost as much as possible. They don't they usually will allow people that don't have as much of a technical background to go into leadership roles. And so there's there are opportunities, but it's just it's really company specific and the way data science is hard leaning right now and in a lot of ways I agree with this, is to put people who were data scientists into data science [00:50:00] leadership roles and to try to promote because and there's a couple of reasons for it. One of them is we have no leaders. I mean, the number of data scientists who are also trained mentored leaders with experience in leadership roles. It's so small. If you ask for ten years of leadership experience in a data scientist, you just narrowed the pool down to, I don't know, 50 people.

Speaker4: [00:50:24] It's not that small, but you know what I'm saying? It's just it's bad. And so we have directors who have two years of leadership experience and ten years of data science experience. We have people at the C-suite level who have five years total experience. And it's it's problematic for us because we need to train leaders. We need to promote leaders in order to get them in front of mentors. Because if you're at a manager level and you have aspirations to go to the C-suite, you need to have a mentor who's at the C-suite level in order to bring you into that role, because there's just no there's no training curriculum for that. And so that's why data science is just slammed so hard into bringing leaders up, especially in larger companies trying to bring leadership up, training data scientists to be great leaders because we have a technical advisory role that we have to play for senior leadership and for the C-suite. And that's at every level because at the management level you're talking to other managers and other organizations, director, VP, you're going level to level in other organizations and you're teaching what data science can do, what kind of problems can it solve, what opportunities are there for applying the technology? What are the pitfalls? How do we integrate just all of these pieces of institutional knowledge? Don't get there if all of your data science technical know how is at the bottom floor.

Speaker4: [00:51:56] And so that's why you're seeing so much of this, even [00:52:00] though in some ways it's nonsensical because yeah, you're right, the director is not writing code. I mean, I am still technical, but no, I'm not writing code. Well, okay, I am, but I don't. That's not my main line gig. And you wouldn't want me writing a ton of your code anymore. It's just not what I do. But if I didn't understand the research side, if I couldn't implement a structured research lifecycle, if I didn't understand how technical skills need to be layered in order for an organization to be built up who can meet particular types of project. I mean, you're hearing like you need to have a technical understanding to be at this phase that we're in right now, most businesses where we're building out the capability. And so it's kind of a long winded answer to say, yeah, it's not fair, but we're kind of this is what we have to do, even though there are definitely some people that are being pushed out. And to Miko's point, a lot of times that there's a slant I want to say to who's called not technical enough, which is something that we have to separate out from what's happening right now. But at the same time, this is a necessity. We have to we just don't have any data science leaders. And the only way we're going to get the know how up the food chain is by doing this.

Harpreet: [00:53:26] Ben, thanks so much. Let's go to Tom and Dr. Tom and go to Poster. And if you guys got questions, if you're doing this on LinkedIn or you got a question, let me know. Or I if there's no question, I'd like to explore the flip side of this question. People who are in, I guess, earlier stages in their career. How we talked about it a little bit, but let's talk about how you decide where to go technically for me, like, how do I know that that's where I want to go? Or is business [00:54:00] where I want to go? Or is data science from the field altogether? Let's go to Tom and then Coastal.

Speaker4: [00:54:10] So I want to tell explain something before I go into.

Harpreet: [00:54:14] This to our.

Speaker4: [00:54:15] Illustrious group.

Harpreet: [00:54:16] Here.

Speaker4: [00:54:17] 90% of the time, I'm communicating in one of two modes confession or encouragement. And when even when I'm doing the confession, it's with a motivation to be encouraging. So if I hadn't listened to Van Gogh before.

Speaker2: [00:54:36] Me.

Harpreet: [00:54:36] I might have.

Speaker2: [00:54:37] Been completely distracted by the.

Harpreet: [00:54:39] Chat.

Speaker4: [00:54:41] Because it has gone off the rails.

Harpreet: [00:54:43] I'm not complaining.

Speaker2: [00:54:45] I'm just saying our chat.

Speaker4: [00:54:46] Time has gone off the.

Harpreet: [00:54:47] Rails.

Speaker4: [00:54:48] And it's too much fun and it's sad that the YouTube crowd can't see it.

Speaker5: [00:54:53] But anyway.

Speaker4: [00:54:55] Then I agreed with all of your points, and I'm hoping what I'm going to say is going to add to it in Harpreet If at one point you need to mute me, I will take it as a kind suggestion and I will be glad to stop.

Harpreet: [00:55:09] And I'm not sure.

Speaker4: [00:55:10] This is going to come out in the best order. But I just want to say this to.

Harpreet: [00:55:13] All of us and anyone.

Speaker4: [00:55:15] That might be listening.

Speaker5: [00:55:16] Live.

Speaker2: [00:55:17] Or in the recording.

Harpreet: [00:55:19] If you get.

Speaker4: [00:55:20] Rejected because of an interview. Because I care for you.

Harpreet: [00:55:24] I'm begging you.

Speaker4: [00:55:26] Please do not take it personally. And if you want to go, look up some of the most encouraging stuff written about data science interviews, please go to our ultra tall friend. You can't tell, but he's six foot six. The Kiwi and Great Britain. Andrew Jones.

Harpreet: [00:55:46] I think some of his encouragements.

Speaker4: [00:55:48] About interviewing.

Speaker2: [00:55:50] For data.

Harpreet: [00:55:50] Science or spot on.

Speaker4: [00:55:52] And I might rapid fire.

Harpreet: [00:55:54] Some of my more.

Speaker4: [00:55:55] Embarrassing. Situations and interviews.

Harpreet: [00:56:00] I've [00:56:00] been promised the world.

Speaker4: [00:56:03] After an interview and then been completely ghosted. One of the more funny ones was the recruiter came back to me with his tail between his legs after telling me I was perfect for this role and said, Oh, this main guy at that company said, We see Tom as a new data scientist. And it went.

Harpreet: [00:56:27] Okay.

Speaker4: [00:56:28] And I said, Who are we looking at next? He said, That doesn't bother you? And I said, Dude, if we'd have been sitting in the office.

Harpreet: [00:56:35] Together, I.

Speaker5: [00:56:36] Would have just.

Speaker4: [00:56:37] Looked over at you and said, next.

Harpreet: [00:56:39] Because seriously. I'm 60 years old now.

Speaker4: [00:56:45] And in freshman physics lab 1980.

Harpreet: [00:56:48] I was already doing data science, at least regression by hand on.

Speaker4: [00:56:53] Dirty physics.

Harpreet: [00:56:54] Data. And I'm like. I was doing data science before.

Speaker2: [00:56:57] It was called data science. And you think I'm.

Speaker4: [00:56:59] New just because I only recently had data science role.

Harpreet: [00:57:03] Titles?

Speaker5: [00:57:04] Please, I don't want.

Speaker4: [00:57:05] To work for you. It's so then if you hadn't gone first, I couldn't have as.

Speaker5: [00:57:10] Eloquently said what.

Speaker4: [00:57:11] Ben said.

Speaker5: [00:57:12] Na'vi.

Speaker4: [00:57:13] The reason people are asking those kind of questions and Serge, the reason people are asking those kind of questions was best answered by then.

Harpreet: [00:57:23] I was on a first name.

Speaker5: [00:57:27] Email frequently basis with.

Speaker4: [00:57:29] The past CEO. I'm not going to mention the company because it's still a good company, but unfortunately.

Speaker5: [00:57:36] That CEO that I was having.

Speaker4: [00:57:39] Great communications with about data.

Harpreet: [00:57:41] Science to. I just said.

Speaker5: [00:57:44] Hey, I think your.

Speaker4: [00:57:45] New plan is pretty cool.

Harpreet: [00:57:46] But I don't think you want.

Speaker4: [00:57:49] Data science reporting to.

Harpreet: [00:57:50] It. I never heard from her again.

Speaker2: [00:57:56] In fact, I think I might have been targeted by her.

Speaker4: [00:57:59] After that anyway. [00:58:00]

Harpreet: [00:58:01] Guys.

Speaker4: [00:58:03] If you're not valued for your background, please. I'm begging you as your friend, just say next. Keep going around till you find the ideal culture, the ideal place where they value the way you think, the way you do things. And but of course, I'll always encourage this remain open to being humble.

Speaker5: [00:58:24] We're there.

Speaker4: [00:58:25] To be servants to the greater organization to show them how to get the best.

Harpreet: [00:58:28] Return on data.

Speaker2: [00:58:30] And if we just show, we really care to explain our hearts to them and.

Speaker4: [00:58:34] To help them appreciate what we're trying to do and really make.

Harpreet: [00:58:39] Our amazing arts count for their.

Speaker4: [00:58:41] Bottom line. They're going to see it. They're going to appreciate it.

Speaker2: [00:58:45] One thing I had to learn the.

Speaker4: [00:58:46] Hard way, though, is you have to be comfortable with releasing.

Harpreet: [00:58:50] Crap.

Speaker4: [00:58:52] The sooner you give them something, the sooner you can start getting feedback to make sure you're not just building your baby.

Speaker5: [00:58:59] And wandering off in.

Speaker4: [00:59:00] The weeds away from what.

Harpreet: [00:59:01] They really want. Me So talk to them a lot. Release crap as soon as.

Speaker4: [00:59:06] Possible so they can say, Oh, I love this. Or No.

Harpreet: [00:59:10] That's not where I was thinking of going.

Speaker4: [00:59:12] And then work it out. The frequent feedback.

Harpreet: [00:59:15] Is what you need when you're starting to.

Speaker4: [00:59:17] Help your group, your greater company get a.

Harpreet: [00:59:20] Return on data.

Speaker4: [00:59:22] Harpreet didn't mute me.

Speaker5: [00:59:24] And I'm just hoping what I shared.

Speaker4: [00:59:26] Was of some value to all of you. And now I'm going to go back.

Harpreet: [00:59:29] And watch catch up on.

Speaker4: [00:59:31] This. Gourmet donut type.

Harpreet: [00:59:35] Talk in the chat. Thanks, Harper. I think it's just a lot of conversation in the chat around fried fried doughy goods. Yes, I loved it. Tom, thanks so much. I really appreciate that. Absolutely. Encourage anyone who's coming from a nontraditional to data science background to check [01:00:00] out a book called Range. I think it's Daniel Epstein. It's called Why Generous Triumph and a Special World, Something to That Effect and just makes the case for why having a diverse background is really also making sense. Definitely check out that book 100%. We recommend it. Let's go to go to Costa and then also, Tom. To your point, you're talking about talking to people. One thing that I've learned from our recent podcast that Marc Freeman was on with the nutrition community, Marcus talking about this tribe framework that we've talked about in other platforms, too. But the first thing the tribe framework is talk. Talk to your stakeholders. Understand what it is that they want something so much that's going to cost to the after cost up. Let's go to a question that Gideon had that was touching back when we were talking about earlier. So it goes the.

Speaker5: [01:00:58] So I just want to kind of separate like a bit of a distinction between technical knowledge, technical background and previous roles as a data scientist. Right. There are two fundamentally different things, right? I think we're often we conflate, oh, they must have worked as a data scientists for X number of years because that's an easy proxy. When we're when we're creating a you know, when you've got a recruiter who's not a technical recruiter, who doesn't have an understanding of the technology, and they kind of create a recruitment package that's, oh, we've got to look for someone with technical experience for this data science leadership role. Have they been a data scientist and do they have leadership potential? Kind of becomes the question, which isn't necessarily the right question to ask. The question kind of becomes, do they need to be guiding the team from a people management standpoint? Do they need to be guiding the team from a strategic standpoint? Do they need to be guiding the team from like a product roadmap standpoint? Right. [01:02:00] Three very different kinds of leadership and three very necessary parts of leadership in a team. Right. So you need to and I'm talking more from the product world than, say, the experimental data science kind of world search. To answer your question. Yes, there are technical recruiters. It's really strange. And we have one at our at our company here. And he specializes a bit more in in specific fields of technology. And that really helps because he kind of understands the conversation a bit better. So being able to fine grain what we're actually looking for, it allows him to cast a wider net than we usually would be able to, but kind of back on point, right? So there's the three kinds of leadership that you need.

Speaker5: [01:02:44] You need the people management skills. It's one side of it. There's the product skills. And the product skills essentially needs an understanding of the customer. The problem being able to phrase the customer voice, right? Being able to manage that into a product roadmap, that's a totally different kind of leadership. And then the third side is the technical leadership. Right now, what we've found in the product side is and this doesn't come across as much when you have model decisions, continue on with the business kind of kind of thing. I'm talking about more long running. Okay? We have this product that's going to run for years and years and we're going to start accruing tech debt. We're going to start seeing that there are now not just user input to say what the next features need to be, but also the technical limitations that we need to go back and redesign. Right. And in order to do that, you need a technical understanding. Right. And you don't necessarily need to have a data scientist to do that. Right. There are plenty of people in the electronics and embedded embedded systems world. We call them systems engineers. Right. They'd be fantastic for that kind of role. Right. And it's just that ability to break down the problem and appreciate the technical blockers behind it. Right. And be able to communicate that [01:04:00] across the teams and across the product team and get get some hindsight. Otherwise, what you're not doing is you end up going through this essentially this open loop, right where the only closed loop is your user feedback at the end of the day, which is great, but you're going to end up with all of this spaghetti code and this product that's ill built for the systems but reasonably built for the user.

Speaker5: [01:04:21] Right? So that's kind of where the stress comes in. When you say, okay, we need to stress on we need some technical skills in leadership because you need that combination of people, management, a little bit of technical leadership, a little bit of product leadership. Right. And again, this this is different when you're talking about like senior C-suite, C-suite strategic leadership. And it's different when you're talking about, hey, there's this product tech lead kind of role. It's always going to differ, right? But if you if you don't have that balance, that's where you miss out. Right? But again, it's not specifically about have you been in data science? The moment we start asking that question of have you had a data science role for more than five years? You're forgetting how young those kinds of roles actually are. You're forgetting how young that industry entirely is, right? So it's kind of a poor proxy for do they have the right technical understanding and technical background to solve the problems that we have? And I guess you've got to understand what's your gap in leadership? And that's going to change from company to company. And if you don't have a gap in leadership that actually needs technical skills, then maybe you don't need someone with that strong technical background to fill in that leadership role. So.

Harpreet: [01:05:48] To give you an idea, you have to send us a lot of good advice there through the circle and back to the recording world on YouTube. Let's go to Vivian for [01:06:00] a question that she has to go for.

Speaker3: [01:06:04] Okay. So changing gears quite a bit here. Just I read an article about Apple announcing their new mixed reality device and wondering what people think about it, what the future of VR, AR is. And also disclaimer here I got a promotion, it works. And now I work on the Oculus team. So I will be taking notes about everything you guys say. Okay, just kidding. I won't.

Harpreet: [01:06:37] Stop. Congratulations. We've been at Facebook for how long? Like less than six months or something like that. Or a lot longer.

Speaker3: [01:06:44] It was like July or July.

Harpreet: [01:06:47] Almost a year. Already promoted that. That's what's up. Congratulations. That would be an interesting job to have as a mixed reality developer. It would be interesting, but yeah, I'd love to hear if anybody has any perspective here. Tina, you sit there next me. What are your thoughts on this?

Speaker3: [01:07:09] Yep. So I actually worked on Oculus as well. Yeah. So I was there for about a year or so. I was on the growth team and there was a merge with me. It was just like a merger in general. Another growth team. I don't actually know what happened, but yes, I was working on Oculus and I think I'm very biased as well. Like I think this is where the company has put its stake in like a very it's very, very obvious that this is where they really want to invest. And I think it makes a lot of sense as the next computing device. I do think it makes a lot of sense. However, it still is very difficult. I think like in terms of it's it's at a stage in which it's like you don't [01:08:00] know who's going to win this game. I think it's going to happen, but I don't know who it is that's going to actually be the person who like actually makes a device that has true product market fit, which I think it's not there yet for anybody. Another thing is like on the AR side is pretty interesting because I think AR is going to be much more widely adopted than VR is VR. I think it has very specific use cases that may not be something that is going to be as widespread a R technology. I think it's also a lot there's there's a lot more going on there that we may that we may think. And I think in the next few years as well, like, there's going to be a lot of products that are going to be released in that space. So I have my if I were to bet, I would be betting on the AR side to be widely adopted.

Harpreet: [01:08:44] And when you talk to say as the next computing device computing device, when I think of something like the Apple Watch or like a or a wearable device type of thing, not really consider these to be like computing devices. So I guess what does that what does that mean in this context as Oculus being kind of like the next computing device, does that mean be the next thing that we have with the phone is definitely a computing device, right? Is that kind of what you mean by that?

Speaker3: [01:09:16] That's what I think so, yeah. Like that's what I think. Wearable watches and stuff, they are not necessarily computing devices, they're helpful and they're wearables, but something that I think will very much replace the cell phone. And I think that's going to be on the AR side, like Oculus, where like on the VR side, I think it's going to be very specific to certain use cases. I don't think people are necessarily going to sit there every night and go into the metaverse and then just like not leave their house. I think that's I would actually be quite bad if that were to occur. But no, I don't think that. But on the I do think AR is going to replace cell phones at some point.

Harpreet: [01:09:56] Very much. If you guys are interested in hearing more about I [01:10:00] did a couple of interviews. One of them was with Paul. Paul McLachlan, you talked a lot about air. So we can check that out. So did a book with Cronin. We talk a lot about augmented reality and that episode too. So those are two options to check out if you, the listener, are interested. Let's hear from let's hear from from Ken on this. Ken. Ken, what are your thoughts on a VR?

Speaker2: [01:10:35] My finish, my cucumber. I think it's an interesting space. I remember when. The first Oculus commercials were coming out after the acquisition from Mehta. And I thought it was really interesting. There was like one commercial where two people are like clearly playing online with each other, but they hear, but they're in like apartments right next door and they're like complaining about the sounds from each other person. I thought it was really strange because they could literally just like go next door and meet the other person and get those benefits of interaction. And to me, I think it's incredible to connect with the world virtually. I mean, like I have a legitimate business that is through complete virtual connection, but I'm also on the other side of that where I believe that a lot of this, like, it's really bad for our mental health. It's isolationist. Like, we need people. We need to interact with people socially. We need to have space away from notifications and a lot of these other things to be able to to really live fulfilled lives and to find happiness in those types of things. So I view a lot of social media, augmented reality, virtual reality as something that detracts from peace of mind and is highly addictive and is dangerous in some sense.

Speaker2: [01:11:59] And [01:12:00] I'm a little fearful of the direction that goes if that becomes even increasingly more mainstream. I mean, I think all of these things are tools, and if you have the willpower to use it so that it can benefit you, they can be unbelievably valuable. But if we're talking about kids, we're talking about different groups that are immersed in this technology from the moment they wake up to the minute they go to sleep. I don't really know. I don't really know how I feel about that. I think that the solution probably is a technology related solution, and I've seen a lot of stuff. Even I have Oculus, right? I've seen a lot of stuff with mindfulness and meditation and those types of things within the app. So I think that's a good start. I just don't really know where it goes, and I'm a little more bearish than I am, bullish associated with the trajectory, just associated with the personal challenges, not about what they're like. The obvious benefits of the technologies are.

Harpreet: [01:13:06] Ken, thank you so much. There's a show on Amazon Prime called The Feed. Have you seen everything? To see that shot, I think takes it to the extreme with the metaverse and all that stuff. Super duper interesting, having fun and checking it out. There's different codes that are looked at at coastal perspective coming from that robotics angle.

Speaker5: [01:13:33] Yeah. Okay. So I want to add a little corollary to our old friend, Arthur C Clarke. Right. He talks about any, you know, any technology sufficiently advanced seems like magic. Right. There's a corollary to that. Anything that any any technology that we look at in science fiction and someone goes, hmm, damn, that's cool. I want that. We're [01:14:00] going to work on it as a race. Someone is going to work on it as a as a as humanity. Someone will be working on it. Someone has seen Iron Man and gone. I'm making it my life mission to bring those hollow, you know? Or is it that hollow phone of his surreal life? Someone's bringing those. I'd love it if someone brings those screens that float around in front of you to real life like a heads up display in your sunglasses. I'd love that. Right. Like someone's going to do it. Like, that's where watches, Apple watches and stuff came from. We've had that in movies for decades, right? James Bond talking into his phone and, you know, through his wristwatch and stuff like that. It's always about that romanticism of technology, right? We love it. We absolutely love it. So it's going to happen. That's I have no doubt about that. Right. The bit that I'm torn about is how do we use it and how do we make it actually significantly more useful. Right. Where it really, really helps is you're actually exactly right.

Speaker5: [01:15:02] It's in the robotic space, right? It's how do I use 20 robots out in the field remotely and be able to access the data in a way that's conducive to a human interacting with it? Right. How do we design complex systems? How do we debug complex systems from very far away? Right. Those are problems that this can really help with. How do we train people essentially think about think about training for for engineers as they're looking at a system. You can get various sensors, various histories of data coming straight into, hey, why is this part of my nuclear power plant, for example, breaking down? And I don't have to be the person there in front of the nuclear fuel rods risking my life because I've got robots doing that for me. And that data is fed through in a way that I can actually action on it. Right. How do we bring that immersive to the real world is that's [01:16:00] where it's really going to help. Here's this counterargument and the counter risk. And I've seen this with defense, right? Like we were talking about using augmented reality to help troops in the field, you know, just getting real time data assessment. There is this thing essentially where you've got data overload, right? There's only so much data that we can actually actively process. There's a subconscious processing of data that's different to your conscious processing data. You end up with analysis paralysis at a minute to minute level, right? That's one of the risks is how do you balance that? So it's all about execution, right? It's exactly like iPhone.

Speaker5: [01:16:40] This is no different. Right? You can have the greatest phone and you can have a rubbish app or you can have a great app, right? It comes down to how we design the apps and how we use it to the best of its ability. Right. The other side of that is the addiction side like Ken was talking about. Right? With a phone, you can put it away. It's not in your face, right? With something that literally is on your face that can get considerably more challenging. Right. But it's such a wide wild world out there. Right. Like some people find that they love being able to respond on their on their watch. Whereas for me, in order to put my phone away. Right, I got a watch that gives me notifications. Why? Because I can look at a notification and say, oh, that someone's messaged me, I can get back to that later. That's actually how I broke the phone. Must respond immediately kind of cycle. I grew up with Messenger and MSN and, you know, Facebook and stuff like that. That was that was my generation where we had instant messaging was the norm for communication with all of our friends. So instantly responding was huge. And using a technology like a wristwatch that only gives you a notification but no capability to respond was a good way to do that.

Speaker5: [01:17:56] Right. So there's going to be people who find good uses for AR [01:18:00] in their regular day to day life space. But then there are all these challenges, data overload, addiction to addiction to the interaction, but there are benefits to be had. There are genuine places where you can train a doctor. Right. What if you had a what if you had a surgeon? Right. Wearing wearing the right equipment, the right air glasses. Right. And feeding that data back to three or four students or practicing surgeons are trying to develop their skills or working with dummy models or, you know, haptic feedback VR systems. Right. But they're able to see the real surgery happening in real time and try to make those same decisions as they go. Right. What how does that increase their experience? Because we know with like surgeons, until you've done 75 to 100 of the same surgery, your expertize in that area doesn't really sink in. Right. So there's all of these use cases where augmented reality, it's really promising. But the fact that. The matter is, it's not going to happen because it's promising. It's not going to not happen because it's addictive. It's going to happen because someone saw it in Iron Man and went, Damn, that's cool. I want that right. We're obsessed with this romanticization of technology and that's what's going to happen.

Harpreet: [01:19:17] I'd be curious to know what the monetization strategies are for something like this because. These are going to be used as just another way to get the ads for the ads. That's a real concern of mine because. Like if you have third party developers to develop stuff for this device, then how are the monitors and what's their strategy? Are they trying to buy my attention? Because if people are trying to buy your attention when the things attached to your reality, it's kind of scary. So let's let's hear from green camp says got to go and get donuts can get the donuts but [01:20:00] can be still taking over next week I believe. Yes. Awesome. Taking over the reins for you next week, I'll be flying back from Denver, so you.

Speaker2: [01:20:09] Definitely got to give me a quick training. But more than that.

Harpreet: [01:20:13] Yeah, absolutely. It's easy. Just this is your training. You can't take it over next week. Then let's hear from you.

Speaker4: [01:20:24] What I understand about that apple, the actual apple headset, the the consensus was it's really cool. It's got functionality, but it's looking for a reason to be adopted. And I think that's if you talk about every platform that's out there right now, that's the problem is the the technology is still about 18 months away from where people will wear it. It's functional enough right now that you can use it. But there's a lot of people saying the cameras need to improve, the screens need to improve in quality. There needs to be there are little tweaks and technology pieces that just aren't good enough yet for a wider for the wider audience. And the other part is, it doesn't sound like there's a like Apple figured out how to make the iPhone something everyone wanted to have because there were things in there and like the App Store was a big piece of it. And the novelty of being able to do everything that you could do with that device and it doesn't sound like Apple's done that with their headset yet. And what's really interesting is in business use cases, this makes so much sense. There are mechanics that use headsets and they get training and real time guidance on how to do repairs from the manufacturer. So now a mechanic that is somewhat familiar with like a Ford vehicle can put their headset on and take like in front of them training where there's sort of this mixed reality that they're experiencing and they're able to do fixes with the instruction manual basically overlaid in front of what it is that they're building.

Speaker4: [01:21:59] So there's, you [01:22:00] know, when you say there's no you know, I don't mean like there's no killer app for this, that there's no use case for it because there are obviously a lot of them. But it feels like everyone's dropping their headset out there and saying, okay, what do you think you can do with this? And at some point you have to stop giving it to developers and saying, Hey, can you figure out what to do with this? And there has to be some sort of an application to it that makes it sticky. It makes it so that people want to spend money on it, because for most people, this is going to be like, you could go on a vacation or you could buy this headset. And to get over the people like me, I'm just an idiot. I'll buy it. I'm stupid. You know, I. I understand that I suffer from poor, poor impulse control, and that's the only reason why I buy these things. So the target audience like me, that just we're going to go buy it, we're going to play with it, we're going to figure out what we can do with it. You can't rely on us. We're not enough. There has to be something else on the other end of it. And if you look at what Disney has done with Disneyland's sort of augmented reality, where they have an app that's just dumb, simple started out as just here's a map to Disneyland. So when you were at the parks, you could walk around, you knew where things were.

Speaker4: [01:23:14] Then all of a sudden you had ride wait times, then you could order food and then you could use Genie Plus, and now you could book out, plan out your day with these fast passes and different types of experiences that you want. And then and it's just that incremental creeping where they're taking a phone literally as far as it can go. And I think companies like Disneyland have or Disney have this really compelling. You know, I'm wondering if there's something under the covers because they have a robotics lab, they have an extraordinarily advanced robotics capability. They published for a while. And then I guess somebody said, no, stop. And they haven't published anything publicly that I've seen for a while. But they actually have a fairly advanced group, and I'm wondering [01:24:00] if something like that is the business that's actually going to create either the thing that makes Apple's and Metas headsets sticky or if they've got something that they're working on from a technology standpoint that once they exhaust the utility of a phone, they transition into it. Because they've been playing around with wearables, they've been there's a lot of stuff that they've been working on under the covers. That's kind of where my head's at. Is is there is there another company that's sort of stalking? And they figured out what the application is, another building the hardware towards that use case because that would be an interesting show up and kind of take market share.

Harpreet: [01:24:42] Ben, thank you very much. Vivienne Hopefully some good, good notes there. And as always, this is recorded. So if you go back to YouTube immediately after this and get those notes. Transcription should be up in a couple of days to you guys. Well, go ahead. Wrap it up. I've got to get going. So I appreciate you guys hanging out. Thank you so much for all of you guys for showing up today. A lot of a lot of folks stopping by. So appreciate you guys coming and hanging out. Be sure to tune in to the episode that I released today with Professor Dr. David Spiegelhalter. He's the author of The Art of Statistics, as well as Covered by Numbers Statistics professor at Cambridge University. And yet he was I was introduced to him from Marcus de Soto, two guys who look up to because I used to watch a shit on BBC growing up. So that was super cool to chat with them. And the fact that I never thought my life of talking with professors from Oxford or Cambridge, so that was pretty cool. Definitely things that sponsored the World Conference Machine Learning and Production Conference happened June 7th through June 10th. Want to be sure to go and get your tickets. Easy discount code HARPREET for 50% [01:26:00] off. That's it for this, guys. If you're in Denver, let me know. Should be a message or LinkedIn email. I'll be in Denver at the 2030, the 26. I'm hanging out guys and harpreet.

Speaker4: [01:26:12] Real quick, I think we should thank Vivienne for stooping down to show up to this community again, even though she's a big shot at meta now, so honorific to have her show up.

Speaker3: [01:26:26] I'm going to try more. I just keep having busy Fridays and it's technically still a work day.

Harpreet: [01:26:33] So that's how I get promoted. Keep going. Congratulations again. That's so dope. So awesome. Excited for you and looking forward to to get more of these futuristic architectures and telling people like, hey, I know who helped build this thing. That's awesome. Thanks so much, you guys take care of a good rest evening. Have a great afternoon, weekend, wherever it is. And my friends, you got one life on this planet. And I tried to.