notebooks/data/hh101.txt

HH101_mixdown.mp3

Harpreet: [00:00:06] What's going on, everyone? How's it going? How's your body doing? We are here with the data science. Happy Hours. I'm live for the third time today. So you guys are not yet tired of me. You will be soon because I'm going live like crazy. I'm going live like crazy for the month of November. Today is my third live stream. I did a live coding session lecture kind of session earlier today talking about how resonant to change the world, really how the Skype connection changed the world, gave like a brief history of computer vision before deep learning, then talked about convolutional neural networks and the building blocks and then, you know, everything leading up to a to to resonate. The live coding session seen resonate in action, which is cool. Then I did a session with that. Jess Ramos I had a lot of fun talking with her, so hopefully got that a chance to join in with that. A couple announcements for what's going on next week. Next week I've got more live sessions happening next week, doing a live session on Thursday, 12 p.m. Central Time with Mark Ryan, who's the author of Deep Learning for Structure Data. I for 1 a.m. excited for this conversation because, you know, there's I feel like deep learning is a lot of hate on LinkedIn for some reason and I don't understand why and I well, maybe I have have you know hypotheses about why but but I want to debunk some of those myths and, you know, see how it could be applicable for those of us who are kind of working out of tabular database, things like that.

Harpreet: [00:01:31] So stay stay stay tuned for that. I'll be posting links on on LinkedIn for you guys to join in. And then again on Friday, I'll be talking about efficient the efficient family of models, and we'll see that implemented as well in my favorite low code training library called Super Gradients, which also happens to be the library put out by DC. I wonder. Coincidence also launched the the Community Deep Learning daily deep learning daily dot [00:02:00] community. Guys check it out it's I'm trying to build a one of a kind community for deep learning practitioners those people who are in industry who are solving difficult problems but also make it welcoming and inviting for people who are early on in the journey in the community. We've got like people like Co Sub and Richmond, al-Awlaki and Lou Rivera, who's there's people in there who are like doing like the ML ops side of deep learning and computer vision, which is amazing.

Harpreet: [00:02:29] And then we've got people who are joining that are kind of early on and then people like me who are kind of meddling in the middle. So it's the community for for deep learning enthusiasts, a place for you guys to come ask questions, join in on live events and all that stuff. And I'm proud of this, you know, building something from the ground up. It's it's hard work, but I'm happy to be doing it. So I hope you guys can join. That being said, those you guys that are watching on LinkedIn, all of your questions and comments are welcome. So please do let me know if you have questions or comments. And then also all of your engagement, all those likes, all those reposts, those are all welcome as well. Spread the word, man. Happy hour. What's going on? You know what else is going on? Layoffs, people. People getting fired. Left, right, center. I wonder who were who wore it worst. Let's go to Vin and talk about that. So first, just give us give us a recap of this week. Companies are getting laid off and who was doing it? You know, who was running at the worst? Go for Vin.

Speaker2: [00:03:26] I think there's something like 30, 35 companies this week, just like this week alone that ended up laying off in the tech startup space And bigger ones, you got Stripe that laid off, Lyft laid off. Twitter is the big one today. And when you say like, who were it worst? I think if you're looking at Stripe and Twitter, I think both of those are the worst looks because you've got companies that like, why are you all of a sudden dropping a ton of people? For Twitter, it's because [00:04:00] Elon showed up and it sounds like he violated a bunch of laws. In Ireland, you're supposed to give 30 days before doing a mass layoff. That didn't happen in the US in California. I think it's 60 days. It didn't happen. So. And not the system that he used was something like you got an email or you didn't get an email. And some people were getting locked out of their accounts last night before they got the email. So they were learning because they got kicked off a slack. That's not the great way to learn. You just got laid off, especially from a company like Twitter. When you listen to the way people that work there talk to each other. That's a community. A lot of people that work there long term love that place. And so laying them off that way, it's everyone now looks at Twitter differently. And it's the same thing with Stripe. And I think companies like Peloton have done terrible layoff rounds. That's another company that. And what are you doing? But if you look at most companies right now that aren't profitable. You know, their stocks are getting destroyed. And so even if there's no reason for them to lay off, they're going to be laying people off just because their investors are saying, look, you can't you can't keep losing like Uber.

Speaker2: [00:05:19] You can't keep losing like $1,000,000,000 every quarter. You just can't do that. There's no way to have any sort of valuation behind you. So that's what's driving it. Investors are basically saying, look, we're fed up. And I wrote a little bit about how different types of investors are cycling in last week. And one of the big themes now is when you go from long term investors to short term investors, short term investors are all about 12 to 18 months. And so if you've got initiatives that are running that are going to be profitable for 24, 36 months. They're not going to put up with it. They're going to be saying, just cut that entire group. And so that's what's happening at some of these companies. And the layoffs are being done so badly. You [00:06:00] know, there's no such thing as a good layoff. But if you can give people advance notice, if you can talk to people and say, look, this is coming, we're in trouble, like Google's great job, kind of laying the groundwork. Back in June, July, Metta did the same thing. We're tightening our belt. We're going to get smarter about things. Microsoft in different groups that were going to be impacted. The same thing. They said, you know, it's coming. We've got to reevaluate. We've got a freeze. We're going to be smarter and more strategic about how we hire. So there's companies that are doing it better where they're basically saying, look, this is coming and they've been clear about that for months. So no one's no one's being caught off guard. Everybody has enough time to prepare. But when you tell somebody a week in advance piece, that's just. That's the worst.

Harpreet: [00:06:51] Russell, I'd love to hear from you. You had some commentary before we were on on on air. We'd love to hear from you. Shout out, Joe. Risto. Good to have you here again. We're talking about layoffs and things like that. So if you got any. Yeah. If you got anything on your mind in regards to that, I'll call you up after Russell and then shout out to everybody on LinkedIn who's watching. And Jennifer and A.V. guys got commentary that you wish to contribute. By all means, let me know. Russell is here from you.

Speaker3: [00:07:18] Thank you, Abby. Yeah. So my comment was, as an outsider of some denomination being from the UK rather than the US, I'm interested to know what the perspective opinion is on these types of layoffs from the big companies, especially Twitter, which I look at being an extraordinary event after Musk has purchased the company and then affected these layoffs, plus the others like the Amazons, the Metters, the Stripe's, etc., that are huge, big companies that have hired hard for a long time and had a lot of workforce and are looking to make more efficiencies now. So it could be considered something that could have been strategically [00:08:00] forecast a few years ago possibly. Is that different to any other technology companies that are operating across the wide spectrum from, you know, web3 into data, into anything else that's technologically driven? And if it isn't, is there hope for those that are subject to these layoffs to find employment elsewhere in other companies? And where might they best look at this time?

Harpreet: [00:08:30] Vivian or Joe Diaz won't take a stab at this.

Speaker4: [00:08:34] Can you paraphrase your question? That wasn't.

Speaker3: [00:08:38] Sure. Okay. So so I was saying, so the people that have been laid off now, the big layoffs from the tech companies, Twitter being kind of in the room on its own is Elon Musk just took it over and the other big tech companies have hired heavily for a long time and had a lot of people. So maybe it could be considered just an efficiency drive after so many years of a lot of hiring. Is that different to other technology companies in similar environments that aren't so big or haven't gone through such extreme situations such as Twitter, etc., that may not need to lay off so much? And therefore. Is there additional advice we can give to those that have been subject to the layoffs of what type of companies to look at if they are exiting the Twitter's, the methods, etc.?

Speaker4: [00:09:27] Yeah, for sure. I mean, I know startups that are hiring right now and, you know, companies that are still hiring. And so I would say that for those companies this is a blessing. This is this is the one that they've been waiting for. If anything, it's like if you know how hard it is to find engineers, especially in the Bay Area, it's insane. So, yeah, I would say that for people looking for work, you know, I mean, keep your chin up. I know it sucks and I've had that happen before. You know, getting laid off and it's, you know, it hits your ego. It hits your pocketbook, obviously. But the thing you got to realize is, you know, there's you [00:10:00] know, hopefully you have a good network of people and hopefully people are looking out for each other. But, you know, you'll find something, hopefully. But the big tech companies, it seems I mean, there's been a lot of studies on this, or at least a good analysis, I would say, of like the the head count versus the revenue and versus the the output of these companies. It was just pretty clear. A lot of these you know, in hindsight, everything's really clear. Now we can, you know, trash talk the decisions that everyone made for the past few years because this is the easy target. But, you know, there's definitely there was definitely a lot of hiring and kind of overshooting things. And but the strategy that worked in a zero interest rate environment doesn't work now. And rates are high. That's just it's Warren Buffett always says, you know interest rates or the gravity on asset prices and when the party stops, you know, kind of find a new party, whatever that looks like. So, yeah.

Harpreet: [00:10:52] It's an interesting point you made there about, you know, how hard it is to find engineers. I'm wondering, is there a is there is there a I don't know if class is the right word. Is there like a department or function that is usually up on the chopping block before everyone else? I mean, my intuition tells me that if for in a recession we're going to downsize, we should probably start getting rid of HR people first. But I don't know if that's correct or not, but I'd love to get your insights on that event. By the way, those of you guys watching on LinkedIn, smash that like, yo, smash that like and let me know if you got questions. I don't know what's been going on with LinkedIn. I can't see comment during my livestream. So if you do have a question, please do send it to me as a PM and I'll make sure we get it up here. Dm Whatever the hell is called. Go for it.

Speaker2: [00:11:38] Vin Yeah, don't be a recruiter right now. Let's just Oh my goodness. I feel horrible for recruiting departments that are losing 75. I mean, I've heard ten person teams going down to one. You know, it's just brutalizing right now, especially in tech companies who are looking at the recruiting [00:12:00] team and saying.

Speaker5: [00:12:00] That.

Speaker2: [00:12:01] You're all gone. Marketing is another one that's just been getting hammered and it's not marketers. The the top end marketers themselves, the people that are really good at being creative and building the campaigns. It's the administrative and support staff that have just been whole, you know, and at the sort of a lower entry level in marketing, it's been brutal. If you're in any of those types of roles. And so if you if you're thinking about getting into data science from a different role, it's, you know, race into maybe an analyst role, take something that's a little bit lower down right now, chase into maybe a software engineering role or data analysts role, try to get into something out of recruiting, something that's outside of marketing. Like I said, there's entry level roles. Or if you can get into one of those creative roles, great spot to be in the other teams that get absolutely devastated, our innovation and advanced R&D teams. Any company right now that's looking at an initiative, you know, 24 to 36 months out, like I said before, it's going to produce any sort of revenue. Those are going to be on the chopping block. And any teams that are the primary engineers, developers, data scientists, they're going to be going with them.

Speaker2: [00:13:19] In some companies you're going have the opportunity to transfer, which is cool. There's there's a kind of stack ranking, though, so if you're not one of the hyper. Formers on that team. You may not have a seat when the hokey pokey is over. So it's really looking at yourself with Does your team support core business? Is it essential? Is the role that you have essential, or would a slowdown in business mean that you're not as necessary as you used to and you don't need as many people to service the accounts that you have? That's another one that's going to be in trouble. And if you're really if you're a low performer right now to company. Great time to cycle [00:14:00] out because in the next 12 months, if you're in that bottom 50%, you're pretty much I mean, I don't want to say guaranteed because nothing's a guarantee. But if you're in that bottom 50%, you're in a lot of a lot of danger of being laid off. And sometimes it's just better to start clean. You can end up going from a low performer at the bad job to a high performer or a better place for you.

Harpreet: [00:14:23] As a developer relations professional who's typically housed in marketing, your words frighten me. Should I be? Should I be worried? Should Devereaux? Is Devereaux and developer advocacy, especially those within marketing functions at risk? And if so, what can I do to to help secure my job and identify? So you had your hand up. So if you had a question, we will get to it to you right after.

Speaker2: [00:14:45] Go for this. If you're in Devereaux, you're. You're going to be safe at some companies. If you're at a machine learning company, deep learning company, anything, that's data science first data first Data engineering first ML Engineering, ML Ops. Yeah, you're pretty safe because without a developer relations role, I mean, it's really hard to get anybody to buy until they understand why they're buying, what they're buying, what they'll get out of it, how to pitch it to the sea level so that they can get some budget and approval for it. So in those companies they understand the ROI, or as long as they do understand the ROI, then you're in pretty good shape. But in other companies who are just toying with the devil, especially companies where software isn't their core business, that's where there's a lot of risk because some companies are looking at it as a growth area and so far it hasn't come, it hasn't gotten to positive revenue generation and it doesn't really have a good path to profitability. Those are the areas where marketing spends on anything really is in jeopardy.

Harpreet: [00:15:53] It makes me feel better. Thank you, Vin. Jennifer Yager, Hand-raised Gopher it good to see coach lab in the house coast have got the mic [00:16:00] and the headphones. They goddamn look at you man. Go for it.

Speaker6: [00:16:03] Jennifer Really, just to add on to what Vin said, anything that isn't part of a business's core value chain is going to be at risk like it until we also put PMS and operations in with the other cost centers that are evaluated in times like this, or teams where management believes that a function could be done centrally and support a broad audience, that's another one that's sometimes at risk. Now, let me tell you the flip side, and this goes back to Russell's question. A friend of mine was was given notice at Intel several weeks ago. I think it's just two and one half, maybe three weeks ago now. He is already deep into interviews expecting two or more offers next week. So, yes, it it stinks. And I am troubled that Intel is letting him go. But at the same time, I'm thrilled that there is a marketplace for analytics for engineers. It's still out there. It's a shift.

Harpreet: [00:17:09] You had 20, 20,000 layoffs or something like that. That's just huge.

Speaker6: [00:17:15] If Bloomberg was not accurate, it was an inaccurate rumor. But yeah, it still is.

Harpreet: [00:17:20] So yeah, yeah. I remember being at the Intel Innovation Conference just a few weeks ago, months ago. I can't remember. It was recently ish. They're coming out with some new like GPUs and stuff that I'm really excited and excited for. That's going to be definitely, definitely a game changer. It Koza. What's going on, man? How are you doing? Nice setup.

Speaker5: [00:17:43] Yeah. So it's with a bit of an update. I don't know how this sounds. I'll find out when I hear the record.

Harpreet: [00:17:49] Man, it's good. Do you want to be hosting a takeover? Looks good.

Speaker5: [00:17:53] Depends. Do you mind the unnecessary levels of reverb? But. So I just [00:18:00] got this thing plugged into my mixer that I use for practice and stuff, so I just thought I'd update it. The job market in Sydney is, is is weird. It's it's in a strange spot right now. Like, I mean, I'm just listening to what you guys are saying so far and it's weird because we've got the same kind of thing, right? Like we're seeing some companies laying off 30% of their force, 40% of their force in Australia. Right. And not just in Sydney like Brisbane and Melbourne, etc.. And we've got other companies that are doing pretty well and, you know, just snapping up talent. So it's weirdly a bullish market for people looking for jobs. But also at the same time you're seeing people on the other end of the spectrum as well. So it's it's strange. Again, it comes down to something we've discussed a few times in the past, right? Is that where is that band of 4 to 10 year kind of experience, people that you know, have enough experience to make a significant impact but also aren't right at that top end of the spectrum in terms of pay for companies to actually be able to afford people.

Speaker5: [00:19:03] So I think that's what we're seeing play out at at this stage. And I mean, we've we've got like a kind of a controlled inflation situation going on in Australia right now where, you know, every every month or so, the Reserve Bank is bumping up interest rates by like 0.2 5.5% kind. Just to kind of avoid jumping neck deep into into a recession. I don't know the first thing about that. But what I'm seeing is companies just getting really wary on where they're investing their their talent and just kind of consolidating it into experienced talent more than anything. Now, I just don't know what that looks like. I have no experience at any of this. These the first time that I'm seeing any kind of significant economic change in my career. So I'm just like keeping an eye on the horizon and going, okay, interesting stuff. But at the same time, there are so many companies hiring that. I don't know. [00:20:00] I don't know what to make of it. Right. It's. Still seems pretty positive a lot.

Harpreet: [00:20:05] That Vivian got said. Oh, that's not Vivian Venus been, but a shout out to Vivian. Good to have you here, Vivian. I've been saying, can your Fed call our Fed and work on that not jump jumping face first in the recession thing. Talk to us a little bit about that. And I want to go to Joe after that, because, Joe, you were recently in Australia. I'm wondering what you thought she had around the Australian like data science, data engineering type of market. But first, then then Joe, and if anybody has questions whether you are on LinkedIn or YouTube or wherever it is that you're watching, please do let me know. I'm happy to take all of your questions and comments.

Speaker2: [00:20:40] Yeah, I'm just our currency in the US right now is like a sledgehammer or a wrecking ball. More like when you look globally trying to do business right now in the EU or anywhere is so much harder than it was six months ago. That's kind of it. We're we're raising rates really, really fast comparative to most other countries. And so the US dollar is just going it's so much higher than all other currencies that we're taking hits with all the companies that are doing business internationally. And like my courses internationally have basically they're 20% more expensive now than they were six months ago. And trying to normalize pricing internationally right now has been a nightmare. So there's yeah, we need to we need to call up maybe some people on the Fed and say, you know, yes, raise interest rates, maybe not as.

Harpreet: [00:21:37] Fast.

Speaker2: [00:21:38] Maybe maybe slow it down, you know, stretch it out a little bit more. We've got time. We have months. We can. Slow down a little bit. Maybe not 75 every single time.

Harpreet: [00:21:52] As somebody who is going to be building a house soon, I'm hoping that these rates cool down.

Speaker4: [00:21:59] Ooh, I'm.

Harpreet: [00:21:59] Sorry. [00:22:00] Quickly. Well, you know, we're still the land is you know, land is secure. We just need to.

Speaker4: [00:22:05] I just finished Paul Volcker's autobiography. That was fascinating. It's called Keeping at It. And Paul Volcker was the person who crushed the back of inflation back in the the early eighties. But it was interesting getting his take because he there was a bit of a head fake when. So inflation for context had been running rampant in the seventies and through the early eighties and the order of like double digit inflation. Like I think the Fed funds rate was like 11% and mortgage rates were 18% of their high, which when you think about that, cost of borrowing is insane. I mean, inflation was obviously running really high as we had to raise rates. And so his whole thing was you just have to be very merciless with this thing. It's a monster that if you don't kill it, it will come back. Which it did. The head fake part was, you know, inflation looked like it was under control. They took the brakes off and came roaring back. And so in Volcker's view, and I think this is what Powell is doing, is it just you have to kill this at all costs. You have to crash the economy and there's collateral damage and it is what it is. It sucks. But you get that or you actually get a much worse situation with inflation. If you've seen hyperinflationary economies, that's not very fun either. So it's but that's what happens when you print trillions of dollars and you carpet bomb the economy with it. And so that's super fun. Australia, though, let's talk of Australia. What a great country I hung out with close to. But it's great to great to meet you in Sydney. Yeah, the day to day community there is awesome. It's it's alive and well. I feel like there's, you know, it's yeah, it's going strong I suppose, in Melbourne and Sydney and I felt like both communities were awesome and everyone seemed pretty positively as far as I met the economy in Australia felt like it was different.

Speaker4: [00:23:51] I talked to an American who actually lives there and he felt like it was America. It's like America, but things just move a lot slower there. It's only, [00:24:00] you know, and America is a land of like extreme ups and downs. There's not really like a middle here. It's not like you sort of, you know, when the highs are highs and the lows are lows and we're going to the low lows. But when it's high, it's real high. It's pretty schizophrenic around here. So it it's really felt a lot more even keeled. But even so, you could definitely tell that, you know, people were kind of skittish when I would talk to people about the prospects of their employment and whatnot. And it's the same here. So it just seems to be everywhere. And same with the UK. I was just in London, actually saw Russell there, you know, and that was it was cool to hang out, but same sort of thing, you know, I don't think anyone was really, really stoked in the situation. I mean, I wouldn't mind. Is it Russell? I mean, your Prime Minister was on the way out, so I was like, if people ask about America, be like, well, at least we're not. We don't have that problem this week. So but yeah, it's strong American dollars. It's a good thing if you're traveling, but it's a crappy thing if you have to export anything. So it's not like going to Australia was awesome. In England, I felt like went to Mexico, so that was great.

Harpreet: [00:25:04] They get paid in US dollars for Live in Canada and it's beautiful. I love it. Hello.

Speaker4: [00:25:08] Lucky you. Yeah.

Harpreet: [00:25:10] I make more than most doctors in Canada, which is funny. Shout out to people For some reason it's hard for me to see live comments on LinkedIn, but I do see a few here. Shout out to Rodney Beard. Man, I haven't seen Rodney or heard from Rodney in a very long time. Good to have you back here, Rodney. Warren Simmons is saying I have seen several recessions already since 1990. In the markets where I've worked, one things holds true Keep true to your ethical standards in whatever you do and maintain your standards. Be compassionate with your colleagues who get the chop. What goes around comes around. What are your thoughts on like Elon Musk? Like just blame everything on Elon Musk. But those Twitter layoffs, those fake fake Twitter layoffs, like he hired actors to to pretend to be like, fired. And then all I saw all week long on my Twitter feed was people poking [00:26:00] fun at that like I was the PM in charge of, like, the bookmark thing or something like that. Yeah. What are your thoughts on that, hiring actors to pretend like they got to that?

Speaker3: [00:26:11] Did you see the names of the people out there? Were strategically chosen names altered in a certain order?

Harpreet: [00:26:18] Yeah, yeah, yeah, yeah. That was. That was interesting. Yeah. So. So how do you tell if a recession is going to come for your particular role? What are some warning signs you should look out for? That the support that then will go to use the resident expert on layoffs. What signs should I look for at my company to kind of gauge whether or not I should temper my fear of getting laid off?

Speaker2: [00:26:58] The big one is when people stop caring about what your deliverables are like. That's a huge that's big. And that usually happens maybe a month or two before you end up getting laid off, because these decisions are usually made beginning a quarter or a quarter in advance because there's a lot of reporting requirements and that sort of thing. So you'll know about 30 to 90, 30 to 90 days in advance if your workload suddenly isn't on anyone's radar. You're not getting invites to meetings that you used to. No one cares. Pms don't talk to you about your deliverables anymore. If you're asking what are we going to be working on after this project? And no one really knows. If you're supporting a business unit that they keep talking about winding down and they don't tell you where you're going next, that's there's most leaders will give you the hint, hint, nudge, nudge, warning that something's going to happen. Nobody [00:28:00] can say you're about to get laid off. But if you're looking for, you know, kind of your workload falling off a cliff or your involvement in other projects falling off a cliff, that's your leader is giving you the the. Yeah. Know your jobs. It's totally safe. And the biggest rule is if you have a first round of layoffs and somebody specifically comes to your team from a very high executive level position and tells you all your job is safe, it's not, you're done. You're in a lot of trouble. That's it's like the biggest lie. And it only happens when they intentionally come to your group and tell you, no, everything's cool. It's not.

Harpreet: [00:28:43] Anybody have thoughts or comments on that? Vivian, go for it. Good to see you again, my friend. Been so long.

Speaker6: [00:28:49] I know. I'm sorry. Actually, my partner, Eric, had surgery. He had an acoustic neuroma taken out of his head, which is a fun thing. I encourage everybody to look up. Anyway, so that's. I've been, like, dealing with him having that surgery lately. Anyway, I wanted to mention also that sometimes you just really don't know when you're going to be laid off. And it really comes as a surprise and there are no warning signs and. I don't know. I guess I just feel like as someone who's been laid off a couple of times. When the market starts looking scary or we're in times like this. I just always like. I have a card in my back pocket of what I'm going to do, you know. And I definitely don't think that that's a stupid idea to like. Create your backup plan. But as someone who has been part of layoffs before, the good thing is that [00:30:00] I've never I've never been part of a layoff in which there was truly no like. Compensation or something like at least like a few weeks or something of pay first. And so then then that's kind of how I like usually think about it is like, okay, if I found out tomorrow that I only have like four more weeks or like two months or whatever of, of work of pay left, like what, What would I do? And like, how could I prepare myself to like, if I get that news, like be willing to jump so that I can, like, act quickly. So I don't know. I think that it's great to try to look out for like what I was talking about, but sometimes it truly is like a surprise, even to your boss or even to your boss's boss and like. I mean, unfortunately, that's the world we live in. And sometimes you just really don't know. And it really comes as a huge shock and surprise. And the best thing you can do is like, be ready to spring. Should the bad news come?

Harpreet: [00:31:05] But sincere that metta and that ever happens to you. You could always teach people how to get into data science because you have that clout.

Speaker6: [00:31:13] I mean, I guess so. I don't want to lose my job. I like my job, but I do. There are a lot of people that have been like, Oh, well, you're fine because you now have men on your resume. And I guess in perspective, I, I am grateful for that. But I also still really don't want to lose my job.

Harpreet: [00:31:36] Yeah, Yeah, I've been thinking about that. Right? Like, okay, like if I was to ever lose my job, at least I think I know what I need to do in order to, like, pick shit up and do stuff like on my own immediately. Like I've taken enough of those of those marketing courses to know what I could do to kind of build something from scratch. But yeah, [00:32:00] I'd never, never want to actually lose my job, especially because I love it, dude. Like I, you know, I see all these posts on, on LinkedIn from these, you know, and Twitter from these people talking about the rat race and, you know, working 9 to 5, blah, blah, blah. It's like, dude, like, I kind of like having an actual job. Like, I love it. Like it's cool. Like, I love the work I do. I love the company I'm at and it's a lot of fun. Kosta Go for it.

Speaker5: [00:32:26] Yeah, this is something that's been playing on my mind a fair bit in the last, like maybe the last week or so, right? I mean. There is. I can't remember what there's a particular term for this, but there's a law. Someone please shout it out if you know which it is. But. When negative news or lies garner more attraction than truth or positive news. Um, and while I well, I'm in no way telling people not to speak out when they're facing negative work circumstances. I think we can very quickly and easily rabbit hole into just, you know, the algorithm feeding us more examples of people hating their job. Right. It's so easy to fall down that spiral and rabbit hole and then you start turning around and looking at every job and every company and every person is, Oh, they're out to get me and they're out to, you know, lowball me on my salary and do all sorts of terrible corporate corporate crap to me. Right? It's it all it almost turns into at some point, I want to turn that noise off in the week. Like earlier this week, I was like, actually, I'd like to hear a few more positive stories about people enjoying their job or people excited about the work that they're doing right, because everything we're doing all the all the technology that we're working on, we all got into it because [00:34:00] it's crazy exciting, right? And we don't hear those stories shared nearly as much. Either that or the algorithm's been feeding me some seriously negative stuff for the last two years, right? Know, one way or the other, I'm not entirely sure which, but I do think we do tend not to share as much the positive aspects of day to day work.

Harpreet: [00:34:26] Speaking of positive things to shout out to, Eric Sims just got his house, so that's positive news, man. Eric Simms If you're listening for watching Shadow Team and super happy for you. Super excited for you. But yeah, like I love being a developer advocate. I like it better than being a data scientist. To be completely honest, I absolutely love Denver more than just being a data scientist, and I think it's just personality thing. It's like I get to teach, which I like to do get to create content, which is fun, and I don't have to worry about fighting uphill battles for data strategy and getting that shit, you know, done. There's other strategy battles that I got to fight, but for the most part, I'm excited to fight those type of battles. Then also some commentary on that on that as as you crack it up.

Speaker2: [00:35:23] Yeah there's it's Yeah. Strategies. It's a good business to be in right now but yeah, there's a whole lot of knife fights happening at the sea level right now and it's because we're swapping from FOMO, running the business and technology running the business because of FOMO to strategies back. And many of the strategists now have axes to grind because they've been kind of slapped around for the last five years. So there's yeah, there's some knife fights going on right now. There's some interesting politics coming back into play. And it's [00:36:00] weird to watch the pendulum swing so hard, so fast to strategy being everything now. And any time you hear anybody talking on an investor call, it's always strategy. You know, we're going back to core strategy, returning to a more strategic way of this, that or the next thing. So yeah, I'm having fun, but I can imagine literally no one else would.

Harpreet: [00:36:23] Is there a difference between like strategy and process? One thing I know about myself is that I don't like too much rigorous process for some reason. Too much process feels like it creates too much chaos and entropy for me. Like I prefer to to have as minimal process as possible, but I feel like that is different than strategy.

Speaker2: [00:36:46] Yeah, process is more your workflow, and so you can have a process for strategy planning, You can have a process for implementing strategy. So there's connections there, but that's really the steps workflow work, products, judging quality, assessing when you're done building timelines and that sort of thing. So strategy can dictate process and strategy contains processes and frameworks and systems. But when you start talking about real digging down to the process, like what you're saying, that's more workflow and trying to maximize optimize. So people really like structure. And I think the more creative you are, the more you rebel against structure, but the more you end up benefiting from structure because you you'll realize your creativity can actually be applied to making your job better. And it's one of the things Bill Gates used to say, I hire lazy people because they always figure out the fastest way to do something. Lowest effort, easiest way. And he was really talking about optimizing processes. It was just a fun way of saying the same thing where Irish people that optimize because they don't want to work hard on a particular item. And so there's some strategy [00:38:00] to that. And even creative people, you can build more, you can create more if you have a structured, rigorous process for it. It just feels like you're constrained because you don't really get the opportunity to continuously improve. But if you look at concepts like lean or real manufacturing types of continuous improvement and Six Sigma, it's a really creative process because anyone at any point can step up and say, Look, I think that's busted and here's a better way to do it. And if you can prove it, your idea now is the process.

Harpreet: [00:38:35] I like that a lot. I think even a higher, lazy people. I love that. This brilliant. Yeah. Like the type of process I hate is just like the nitty gritty ness. Like, you know what I mean by that is like, okay, if I'm, if I'm doing a certain thing, then there's like 12 different other things that's associated with that. One thing that I need to get done and I feel like it gets in the way of getting the most important thing done, that I'm being kind of abstract there, just being a process. Kosta was kind of moved into role, where I think process is a lot more important process strategy you're doing like the data pipelines and stuff like that with, you know, with respect to computer vision and deep learning, which is, which is interesting and fascinating. And you know, you're mentioning to me that you're reading Joe's book a lot for for for guidance there. Talk to us about that. How's that shift moving from the engineer model builder to the.

Speaker5: [00:39:35] So so I'm moving back into the pipeline building space like I probably over the last year, what happened was I went from full time model builder to a pipeline builder because we don't have our data labeled because it was at a new role. And then the last probably eight or nine months have been working predominantly on the, you know, operationalizing those pipelines. And a lot of it comes down to what processes do you have in place to use [00:40:00] a lot of the tools that you've got because you're never going to have like it's a it's a beautiful state that we all dream about as engineers where everything is automated and everything is got code for it, right? Not everything is code and. In order to plug in all the gaps. That takes a lot of development time, right? So where you can't automated or where you can't design it out, you have to have some element of process thinking that that comes into it, right? And even when designing your systems, having that understanding of what the actual process is often helps you find really efficient solutions to it from a code standpoint. Right? So. I found. And I always find that it's different kind of thinking. Now, I'm sure you guys would have all heard of disc analysis at this point, but for people who haven't out there, Vince got that smile on his face. I'd love to know what your thinking is on disc profiles.

Speaker5: [00:40:59] But from what I've seen, this sort of for people who haven't heard about it, disc is basically dominance. I think its influence, stability and compliance right now, typically high compliance personalities, not me, are typically very process oriented and are very happy to follow protocols to the team because they see the value in that right and they're able to follow that. It's just a natural personality thing. Now everyone has like an adaptive and a natural personality. My adaptive personality is a lot more process oriented, so I can switch into that. Earlier. Earlier on my career, probably five years ago, I wasn't able to switch into that at all. Like, that's not my natural mode of thinking, but I've started to see that value in having that side of it so I can switch into that as needed. The interesting thing is. When you've got a team that. All of us, you know, don't have that aspect in our natural personalities. It [00:42:00] makes it quite difficult to come up with solid processes and actually follow them because there's a much bigger likelihood that we'll want to design our way out of trouble. Right. Or use our expertise and our our own skills because it's just quicker for me to throw together a quick script to do that, as opposed to think about a process that's kind of bulletproof that anyone could follow. Right. But the downside of that is that you've got a team where people are working for long periods with their adaptive personality, and that's pretty tiring, right? Like, try that for six months, ten months.

Speaker5: [00:42:37] It gets very, very tiring for a lot of people. And that's that's quite honestly the truth. So it kind of comes down to what side like what what stage is your is your organization out or your product out or your your process and tooling at and what's the right balance of personalities that you need on your team Because skills can be taught, right? Natural personalities, having people who can think a bit differently, people with a different background and experience, right? The the only reason I have even remotely some idea of what, you know, a high compliance person, how you would do or process thinking would do is because I was really bad at it five years ago, spent some time in manufacturing and they showed me that weakness in a big way because the people out there were fantastic at it. Right. And I just didn't have a clue. So over the years, I kind of built up on that weakness and just said, okay, I need at least a bare minimum and to the point that it become something that I can switch on when I really need to. Right. But from a process thinking standpoint, that's the main thing is you've got to be able to identify the value in processes. We, especially in the data science and the machine learning kind of community, I think so many of us prefer to be model builders, right? We prefer to be.

Speaker5: [00:43:59] And [00:44:00] the weird thing about that is, I mean, look at the code, look at data science code. It's low compliance. How much testing do you see in data science code? How much are engineering rigor do you see in data science code, Right. Like you're laughing because you know exactly what I'm talking about. It's like we're only starting to turn that curve now. And people like Mike Kerr, people like Joe, they're ahead of the game in terms of looking at things like envelopes and data and data engineering and things like that. Because there's that realization that, hey, actually guys, we can't do this all as experimental notebooks. We need to start thinking about how do we apply engineering rigor, and that's both from software engineering rigor, but also process rigor that we've had in every form of engineering for the last few hundred years. Right? So that's where like I'm leaning back on some of those books, like I've started reading more of Joe's book to just get a better understanding of more data pipelines because I'm heading back into the data pipeline aspect of things with the job that I start on Tuesday. Super excited about that. But basically, yeah, it's we just need to understand that there is a bit of a heavy feel towards your more experimental personalities in the industry and we're slowly seeing that turn the curve right.

Speaker5: [00:45:17] We're starting to see a larger variety of personalities enter that barrier for expertise get lowered so more people from different backgrounds are coming in. I think we'll see that level out as we need robustness. We'll bring in the practices that robustness needs. Put it this way, right? Nothing, nothing that was designed as a safety need mechanism was ever designed before it was needed. Right. Halos on Formula One cause didn't exist right for a very long time. Like the front stalk of the halo didn't exist. And I was watching a I forgot who it was. But there was a video of a Formula One driver from a few years ago before the halo was there, getting hit above the left eye by a spring. Right. [00:46:00] And it's just mind blowing that that's a danger situation that they hadn't considered. Now the halo was brought in to deal with that. Go back to what was it, the sixties, the Tylenol poisonings. Right. And I think it was Chicago, right? We didn't have like tamper proof bottle tops on that. That seems so obvious now. But we wait until we need robustness. We wait until we need systems before we actually build them. And we're starting to see that acceptance of, hey, we need good data engineering come into play. Right? And probably that reflects at a strategy level like what you were saying before, been where previously it was just we just need to go fast and now they're like, hang on, we need to do this.

Speaker5: [00:46:41] Right? So that's the main shift that I'm seeing from a high level. And I'm I'm kind of excited to head more and more into the hey, build more data pipelines, build more, you know, ML engineering stuff as opposed to, you know, the models themselves. And I am starting to see that that value be recognized more by employers and companies, right? Whereas previously every company was just looking for data scientists. Right now companies are actually like, no, we have we have data scientists. We also need ML engineers. We recognize they're different and the value that they add. So it's a little bit of an ego swallow, right? Like I spent like a couple of years learning, doing a master's degree on how do I build models and all that stuff. And now I'm not building any models, am I still adding value? That's the ego pill for me to swallow and understand that yes, I am in these ways that other people probably can't add value in, right? So those are the things that are going through my head. When you talk about, hey, process will hold versus going from building models to building processes and building pipelines, That's the mindset shift that I've been kind of traversing the last year or so. I'm quite enjoying it actually.

Harpreet: [00:47:55] That's awesome, man. Yeah, I've [00:48:00] never been like part of a proper engineering team and I have never had anybody like to show me like the ropes. I've always my roles have always been like, you know, statistician, data scientist and not, you know, much engineering like people around me. I wonder, like, how can somebody like, you know, that that being a solo data scientist, maybe they don't have all that support or a robust team to learn all these practices, like what do they start doing to, to learn these type of, you know, mindset shifts that you're talking about, essentially.

Speaker5: [00:48:36] Go join those teams. Yeah. Like all due respect to all the books and courses out there, they will help you. You'll get that 10%, 20% knowledge that you otherwise wouldn't have access to. That's just helping you get access to information. That's second hand, third hand learning. Right. Go join those teams. Otherwise you're never going to get to see it in practice. Right? Like, honestly, I was at a robotics company from 2020 through to 2021, mid 2021, and it was great. I was building models all the time and I was building stuff that I look back at today going, You're insane. You're wasting a lot of energy building that stuff, right? If I just built it right and I didn't have the team around me to tell me what the right way is to build it. So I ended up building arguably some really janky shit, which still runs well today. Right? Like, that's just I'm thankful that I wasn't that much of a dope, that I built something that's going to explode the moment that I leave. Right. It's still running a year or so later, but what, I do it differently now. Absolutely, because I want to join that team to have a look at how to how do teams outside of robotics do this? Because I know teams outside of robotics do this right, like secede for models and model development pipelines. That's not rocket science like, I mean. [00:50:00] It's it's really not like many, many, many teams are doing this to the point that we have like literally dozens of off the shelf tools begging and vying for our attention.

Speaker5: [00:50:11] Right. So how do we actually do this? That was my kind of initial quest, right. And now kind of extending that further. Right. So I wouldn't have gotten that opportunity to see that firsthand. And I don't think the lessons learned would have been there first hand without the right team. And there you can learn to a point on your own and then you start to plateau. And then you need seniors. You need principals who have four or five, ten years on you. Right. Like it's it's yeah, you need that. You really do. And you need the the time on the job to be able to focus on that as well. Right. When you're highly operational where the the you know, the pipelines are already up and built, you're not going to get many opportunities to go through and examine how the pipeline is built because you're focusing on operationalizing those pipelines and actually using them to deliver the outcomes. That is a very different role. You're not going to get the chance to spend the time in building the pipelines and re architecting and getting your designs reviewed. Go from literal design review stage, right? It's just not going to happen nearly as much. So, yeah, no, go join those teams. Find the people with gray hairs more and more gray hairs than you and ask them every question that comes to mind and just drain their knowledge. That's kind of my goal right now.

Harpreet: [00:51:34] Don't get fooled by people with gray hair because I have a lot and I know nothing. So Rodney St excellent points coast up. I finally see comments on the LinkedIn post but yeah, excellent points and yeah, any other questions, comments going on? Anything else on anybody's mind. Bingo. For a man.

Speaker2: [00:51:53] Yeah. Actually I have a question for Joe. I was just you kind of listening through all of this. Are you seeing a change from [00:52:00] partnering data science or data engineers with data scientists and more preference for data engineers and data analysts kind of backing away from the data scientists being involved at day zero and seeing more hiring towards that direction, where instead of the data scientists being dragged in, it's really a data analyst being partnered with data engineer and getting more value out of that.

Speaker4: [00:52:28] I haven't seen anything change just yet, but I sense what you're, you know, I'm picking up what you're putting down. So I think that that might be the case soon. It's Data Scientist is a very nebulous title to begin with. But, you know, so what I've seen is, you know, it's the same stuff we've been talking about for a long time, right? It's the the artists of nebulous data scientists. Basically. It's like, what have you been hired to do this whole time? And now you're about to find out. And so any any anything that hasn't been yielding results, I think will be. But I can say the same thing about data engineering, to be frank and analysts were if you're not if you haven't been producing anything of tangible value for I guess during the good times, then, well, you should either start doing that it's very, very immediately or you won't be doing much of that at that company you're at. So that's that's a tone I'm starting to get, is I think there's definitely a refocus in general just on every conceivable way of either trying to make more money while you can kind of, you know, refocusing things or just cutting words. Definitely not going to yield anything for a bit. So that's the general sense that I get overall. But I haven't seen any specific, I would say rules being targeted just yet, yet being the key word. What about you?

Speaker2: [00:53:53] Yeah, I've been hearing about the we're focused more on hiring analysts, but [00:54:00] I'm wondering if that's just, you know. People talking about it a higher level or if that's actually something that's ground floor happening yet because I'm here. That's kind of the thing about when I hear someone say at the sea level, hey, we're going to focus more on this. We're going to focus more on that role and we're going to do this kind of partnership. It's like, okay, yeah, what are you really? And so I never know when it's real or if it's smoke.

Speaker4: [00:54:29] Well, the thing with analysts, too, is you need something to analyze, right? Like, there has to be some sort of, you know, you're moving the needle on something, but. So I guess it depends on how much where that function has has become before things kind of dropped off a cliff. So if if those things are adding value, great. You know, hopefully it continues. But, you know, and I think I was like the thing that most companies I give them the benefit of the doubt. And I hope that they're, you know, investing the money wisely. But, you know, endless amounts of free money mass, a lot of stupid mistakes, too. And you're about to find out, I guess, as Buffett says, who's been swimming naked when the tide goes out. So it might be a huge nudist colony for all I know, but. Yeah. Yeah.

Speaker2: [00:55:11] It's do I mean less data science magic being panacea? Yeah, it it's kind of interesting that I think from what I'm hearing, leadership's asking the question, you know, what's the difference between the data analysts and a data scientist? Which ones do we need? When do we put one on a project, not the other? Because, you know, the salary is like half. You can hire a data analyst for anywhere between 185,000 120,000. Trying to get a data scientist for us from 200 K now is it's hard. So that's what they're looking at is there there's actually like a conversation that I'll hear from time to time, which who do we need on this? What department do we need on this? Which, you know, should we be hiring more of this versus that? [00:56:00] That's the conversation I'm hearing. But like I said, I don't I don't know if that's going to turn into something. Or is that just this month's conversation?

Speaker4: [00:56:09] Well, these things have a weird way of working themselves out to because it's like, you know, execs talk to each other, right, at different companies that are trying to compare notes. And I feel like it's it's like a lot of things where people follow fast, even if it makes no sense. And so it's like, oh, I'm going to cut my data teams. I get f them, let's just cut those guys too. So it's, you know, because a lot of this I mean, recessions are a weird thing where it's, you know, you're trying to forecast the future and recessions occur precisely because everybody starts cutting and cutting back. Right. It's not like these things just sort of happen in independent. It's demand drops off, but demand there's a lot of reflexivity, as a source calls it. So these things have feedback loops. And when I talk to each other about their hiring or more to the firing plans right now than I think that there's a lot of comparing notes. So, you know, it's this is this isn't like a fine art, the art of like cutting your teams, right? It's like, you know, how do you I mean, it's a question that I think a lot of people here have been mentioning over a while, which is how are you assessing ROI in the first place? And I've been talking about this in the you know, in the boom times. And if you're tracking how ROI has been calculated, I suppose it should be a pretty easy exercise for you to figure out. You either keep or you cut. But if it wasn't providing ROI back then, I don't know why you're keeping everyone around in the first place. It seems a bit silly, but we might think we've managed to do a really good job at running things that have the appearance of businesses but aren't really run as businesses. So that's changing.

Harpreet: [00:57:47] I'm curious about that. That that last team about running things that have the appearance of businesses but aren't really businesses like what would be kind of an example of that.

Speaker4: [00:57:57] Well I mean I think the entire startup ecosystem, for [00:58:00] example, is an example of this, where to call it as it is. I mean, you have a lot of companies that were flush with VC cash and or were incentivized to grow with. I mean, when I would talk to startups, I asked them already incentivized by revenue, profit or logos and logos was the driving factor for a lot of these companies in terms of, you know, I need to get more logos and that's how I'm going to get my next round. Revenue was, I mean, so you'd see these kind of dinky deals coming through, but it's all long, your logo collecting, that's what mattered. But the last time I checked, I don't think logos paid the bills for real businesses by banner. I try to run our business on logos alone. I mean, we got our business in like 2 seconds and that's the reality of it. But and most businesses are they do these weird things. They make revenues and profits. It's a very strange idea. And, you know, you can look at Wikipedia and find these how it's calculated, but you know, this this this fad of profits and revenue, it's slowly catching on. And so that's what I mean. These by the by the official term of what a business does. Right. It generates profits and cash flows and returns. Those are shareholders. That's literally the mechanism of what a business is. Right. Until then, it's like I said, it's it's it's it's something that's on its way to becoming a business. But I wouldn't say it's strictly defined as a business. So that's what I mean. It's play business.

Harpreet: [00:59:27] So. I like that. Thank you, Joe.

Speaker2: [00:59:33] To define what a laying of business is. My daughter makes more money every month than Uber does.

Speaker4: [00:59:40] Ouch. But that's the whole point, right? I mean, think of how much money went to Uber. You know, I mean, colossal amounts. And I don't know that they'll ever turn a dime of profit that's going to recoup that right Or so. And you can you can play all the shenanigans you want about EBITDA and all this other stuff. But at the end of [01:00:00] the day, it's either making money or you are and you know, smuggler calls it like bullshit earnings. So, you know, you've got everything except the things that matter in a business. But anyway, I got off my soapbox and.

Harpreet: [01:00:15] But what would happen like what would happen if, like Twitter, Twitter, Uber disappeared?

Speaker4: [01:00:21] Like a cabs.

Harpreet: [01:00:23] Cabs.

Speaker2: [01:00:24] Literally nothing. I mean, I would feel bad for the engineering team because they're amazing. I feel bad for all the people that work there because they're all literally trying their best. But how can I just I, I ask this question a lot over the last two years. How can a company that loses $1,000,000,000 a quarter be worth anything? How do you tell me that their stock is worth $10 more today than it was yesterday when they still aren't profitable? Like, what do you you know, we're supposed to be making guesses based on forward looking earnings. But if their forward looking projections are losses, how is their stock value not negative? I'm just you know, I'm just doing math, right? I'm not I'm not talking crazy.

Harpreet: [01:01:09] Where where are.

Speaker2: [01:01:10] They saying it's not he's not nuts. It's all this is the entire startup industry. Talk about how many companies have never made a profit. I mean, Uber's been around forever. It in a year. There have been profitable quarters, but I don't think they've had a single profitable year. Even when they sold part of their business to China to I think it was Didi. I even think that year they weren't profitable and they sold it for like 1.2 billion. So how is that a company? I love them. I have been with Uber and an Uber customer since forever. Like since the very beginning, I've used Uber. I love them. But how do you make that much money and not make any money?

Speaker4: [01:01:50] Well, I think it's more of a security, actually, you know what I'm saying? Like is traded as a as a financial instrument that people will pump up and sell the next person and so forth. And again, there's nothing wrong with it. I mean, I've worked [01:02:00] at startups and Lord knows I got startups asking me if I want to join them right now. And so I know how the game works. And it's like when you're in that game, you play that game and you got to know how that game works, right? And that game's a lot harder than it is than it used to be because it's harder. I would say it's going to be easier in some ways to be able to hire more easily to know. And that's been the bottom, the choking point for a lot of these companies. You just can't find anyone to work there because, I mean, they all the talent's been locked up in Facebook and Twitter and all these companies and Google and everyone else. But I wager, you know, a regular source letter to Mark Zuckerberg last week from the alternative capital letter is like you got to be cutting a lot of people in that it's you know, no offense to Viv, but that's I mean you throw it some pretty large numbers.

Speaker4: [01:02:44] I mean so it's you know, things are going to happen I suppose. But if that's going to unlock a lot of talent that now can go to these startups, I think that's probably a good thing. At the end of the day, like startups, you know, I'm not I wasn't here to bash startups. I think the business model is what it is. It's like you until you generate revenue and things that would, I guess, qualify as a, you know, a business. You mean you literally are like on life support because you rely on VC funding. But again, I work these places, they know how it goes. It's like I'm stupid and just talking too much crap. Like, so you just play that game and hope it works out. So but you know, in this case, I think it's it's going to get easier because you're going to have less competition. So.

Speaker2: [01:03:23] You know, I just wonder how long you can I mean, there's got to be a time where you say, look, how are you not profitable yet? You know, I get every startup needs a runway. Every startup needs someone to take a risk and a chance in it in order for innovative technology to ever make it to market because it's expensive. You know, ask Mehta how much the metaverse is costing. It's not easy to build that big of a platform. You know, it's going to cost a lot of cash. You have to be able to look at your investors and say, I don't care how low my price is, you can put my share price to whatever you want to. I'm in charge. We're doing this. You have to have Mark Zuckerberg [01:04:00] nerves and play that game of chicken. And Amazon did the same thing. You know, Wall Street Journal wrote him up, wrote up Bezos and said, why don't you stick to groceries? Well, because US is now massive. That's why it's now saving Amazon's grocery business, which is really doing terribly so.

Speaker5: [01:04:20] Right. But the difference there is that at some point, the pay, the risk paid off. Right? Right. There was a legitimate. Plan for. Under what market conditions will that move of focusing on distributed like web services actually pay off?

Speaker2: [01:04:37] Well, and I think everyone says they have a plan, like a path to profitability, especially now everyone has a path to profitability.

Speaker4: [01:04:44] You have to say that. Would you would you be otherwise like not? We really don't know how we're going to do this.

Speaker2: [01:04:48] Well, I think that's what a lot of companies have been doing, right? Like until the last six months, they've been going, yeah, some day. I mean.

Speaker4: [01:04:56] It's like, oh, yeah, we have a we have a plan, right? I mean, I've seen all these plans. They're basically facsimiles of other startups at the same stage because like, here's what we need to either do, here's a trajectory you need to be on for all these different metrics in order to get our next round or to IPO or something with nobody's IPO right now. So it's like, you know, here's our plan to be at least like burn neutral, whatever the hell that means, or just extend our runway for a couple of years. But it's like, I don't know. I mean, there's a certain input that you need, which is revenue. And that's that's going to be locking up pretty hard. And so, you know, there's only so much money to go around. And that's I don't know.

Speaker5: [01:05:30] I mean, how much does that come down to the right horizons? The right like because you can make a plausible plan for the next two moves in chess in any situation. Right. I've done that. I'll play that up to like ten. Right. And what horizon does that break down is kind of my question.

Speaker4: [01:05:46] I mean, whatever horizon is going to satisfy the people who are asking the question, Right. That's that's that's what it comes down to. It's all bullshit. At the end of the day. It's like there's there's a reality of the market and there's reality of what you're and there's the story you're trying to spin in order to get what you need. And also [01:06:00] and there's a story of what people who are listening to you want to hear so they can tell people their their stakeholders what the story is. This is how this works. It's all just, you know, Fugazi. Fugazi is the old saying goes in the famous movie. So this is so I'm not out. I have to take off, actually.

Harpreet: [01:06:26] So good to see everybody. Thank you, Joe. We got you like a ghost.

Speaker2: [01:06:32] I think you're asking the right question. What's the runway and how far ahead can you look? But like I said, at some point you have to. Your story can't be a story for ten years. I just I don't know how you keep plugging cash into something. As a VC that hasn't been profitable for five years and they're talking about maybe never being profitable. Like I get. Uber has value. If you broke it down as just a you know, the technology has value, the infrastructure has value, the customer ecosystem has value, the network and marketplace they've built, Those all have value, but they're not profitable. So yes, they've created an asset. However, the value of the asset to investors seems to be higher than the value of the asset to customers, which is where I, I struggle. How look, at what point do you give up on it and just say it's going out of business?

Harpreet: [01:07:37] I'm still perplexed by Uber. Like where are they losing money at? Because, like, they don't own any cars. Like, you know. Is it because they're paying the drivers pain insurance? They have too many engineers. Like, where are they? They lose money or are they not charging enough? Because, I mean, it's not like it's it's not charging enough. Yeah.

Speaker2: [01:07:59] Yeah. Yeah. [01:08:00] Their problem is and I don't I'm not bashing Uber. I love them. Like I said, I absolutely love the company. And the concept and what they do like is awesome. I'm trashing the business model a little bit with respect to the marketplace. And I think the problem is that they expected self-driving cars to come sooner because that's they're only out when you take the labor cost away, all of a sudden you go, Oh, wait a minute, this business model works really well if you have autonomous vehicles, but as long as there is a person in the driver's seat. This doesn't work. Amazon's got the same problem. Amazon Prime has the exact same problem. While there's somebody in the driver's seat, their basic economics is going to continue to get worse. So and I think a lot of companies are getting to this where it's a race to the bottom. Labor is a commodity, but we are in a labor shortage. So all of a sudden, the things that they've been relying upon aren't there anymore. And that runway to self-driving cars where labor is eliminated isn't where they thought it was. And like I said, for a marketplace like Uber, there's value, but they're not charging enough. And the problem is they've trained customers to be very price sensitive. They've trained their customers at a particular price point. And now they're in trouble. Mm hmm.

Speaker5: [01:09:29] So it was just too early in the game then, because, I mean, realistically, for them, they're reliant on many other companies reaching the self driving stage because, honestly, how many how many self driving startups have you seen actually stand up to the existing automotive industry market? Right. Like the dice and pulled out of the electric car market? Forget the self driving car market, right? I was there when it happened, actually. I was at Dyson when it happened. It was quite an interesting week. That's all I'm going to say. But [01:10:00] I mean, rivian they're standing up, I don't know, on their profitability. I haven't looked at all, but they're the only name that's other than Tesla that's a nontraditional, like automotive company. Right? And the other automotive companies are catching up quick. Hyundai and Kia just released their flagship electric cars. Right. So they're going to catch up to, you know, the likes of Tesla and Rivian pretty fast in terms of their actual technology offering. Right. But in terms of self driving, how much does that actually like? The point for Uber is that they're completely reliant on the industry of the entire automotive industry to catch up to what they need to make that profitable, Right? Amazon Prime, on the other hand, they're really just waiting for drone regulation and stuff like that becoming a little bit easier and that may be more accessible. But the funny thing to me is like I understand Uber not being too willing to build their own self-driving cars and stuff like that. They've they sold their self flying aircraft business at a loss or something, right? Like if I'm wrong, but someone like Amazon, I'm surprised they're not hiring more robotics engineers to work on this self flying drone problem. I mean, package delivery drones is not a very difficult concept to adapt existing drone technologies from like they should be. I don't know, like, I mean, if that's their plea, why don't they go out and hire bunches of robotics engineers and try to solve the problem themselves because there's no established industry for that, Right.

Speaker2: [01:11:38] Well, you got to think about it. If they live up to high reliability standards, five nines, that means one in. If I'm remembering. Right, one in 100,000 failures. If you do 100,000 deliveries a day, that means one thing kills somebody every day.

Harpreet: [01:11:55] Yeah.

Speaker2: [01:11:56] And so five nines all of a sudden doesn't sound so amazing. You [01:12:00] know, if one drone falls out of the sky every day or every 100,000 deliveries you imagine, like every 100,000 plane flights, something hits somebody in the face.

Speaker5: [01:12:14] It's it's not a problem. That's the reality of of a robotic system. Right. It's that much more dangerous because you're dealing with it in the physical world. It's passed litigation now where you can just fire or sue a driver. Right. It's about is the technology willing to stand up to it? It's possible, but that market's not yet big enough to, you know, to make it affordable for a company to do that. So, I mean.

Speaker2: [01:12:41] It's like Elon Musk said, I mean, this is a hard problem to solve. And as much flak as Elon gets, he's not an idiot. I mean, he he used to be a very talented engineer. I don't know how he is now, but I'm just saying he used to be at PayPal, no joke. He was a talented engineer. He understands engineering challenges, but he's had the same realization. Everybody who's tried to tackle self-driving cars has had, which me included. I thought we'd have them by now. I bet somebody a very nice dinner in 2017 that we would have significant disruption by now in the the automotive industry and trucking and everything else. And I was wrong as anybody else was. So not pointing fingers. I am the idiot I speak of. But it's a harder problem to solve is we don't know how to get self-driving cars through people's perception. And it's the same thing with drones. I mean, if I drove these same frequency as a drone did, or the same frequency as a self-driving car did, I'd get into more accidents than they do if you if you did the metrics on a self driving Amazon versus a human behind the wheel, Amazon, all of a sudden that self-driving car looks really, really good. Nobody says that about Tesla's autopilot. [01:14:00] You're not talking about a comparative. And that's one of our problems in data science is we don't know how to talk about reliability. We don't understand how to publish reliability requirements, human reliability, how many crashes per mile do we have with a human driven semi versus a rivian? I think rivian's do semis, right? That's rivian versus a rivian. So how many crashes? And if you look at self-driving cars, Tesla in driving that autonomous mode compared to a person, how many crashes per mile do we have? When you begin to advertise it from a reliability standpoint, you can say this must be at least as reliable as a person. And if it meets that reliability standard, you're good. Because, I mean, what are you supposed to do? Make it better than us? Really? Come on.

Speaker5: [01:14:54] Well. Well, that's. That's the eternal perception of all things robotic, right? Like on a manufacturing line. A human visual inspector is probably going to have 80% accuracy over over the course of a day, a week, a month, a year, a career, a model that's supposed to detect defects in complex plastics. Yeah, sure. That's expected to be 99.9%. Right. Like you see it all the time, Right. We just expect perfection because there is no how do I put this? There's no throat to clearly choke.

Speaker2: [01:15:30] So what do you somebody who's a very well known expert in robotics once said to me privately that the the difference between a human and a robot is that humans are way better covering up their messes.

Speaker5: [01:15:48] Oh, man. That is so true. You can't hide the magic smoke, guys. Trust me. We'll try it. You can't hide the magic smirk. Robots just hate giving that stuff out.

Harpreet: [01:16:02] With [01:16:00] that, let's go ahead and wrap it up, y'all. Great discussion. Thank you all for being here. Thank you all for watching on LinkedIn. I thought for the longest time LinkedIn algorithm was punishing me, but we've had steady, you know, mid-teens viewership in this video. And I got like the biggest engagement on the post that I've ever had recently. So that's pretty cool to see. So I think LinkedIn is now rewarding me again. So thank you all for being here. I don't know why I'm going off on that tangent, but thank you all for being here. Thanks for joining. Have a good rest of the afternoon. Remember, my friends, you got one life on this planet when I try to do something big.