Facebook F8 Event Highlights, Amazon Shareholder Letters and Sleep with William Bourke

And we're live! What's going on, guys? We're back for another show. Or actually, it's really the first show. First of many. First of many. Yeah, exactly.

See how this one goes.

I think you may have just heard the intro. Psst! We haven't actually done an intro yet.

Pre-recorded intro.

Yeah, exactly. So, we're going to do the intro based off what we talk about today. But otherwise, the premise of this show is we want to We want to talk about— we go for these walks after work or at the end of the day, and we talk a lot of technology, business, life in general. We thought, why not record them and post them on the internet so other people can enjoy them too? And really just so we can look back on them in 20 years and go, what were we thinking?

Well, maybe here's a brief intro. I'm a business person learning AI, and you're an AI person learning business, so kind of mash in that situation. It's just good to throw around ideas. I have ideas about business and how business can integrate AI and you're the tech side of things, so you share the spotlight on that side.

Still a newbie, still a newbie on that part, but I love learning. So, well, you've got the AI shirt on, man.

I need to—

I should have got me one of them. Look at that. I don't know if you can see it, but what's, what's that from?

MIT. It's artificial general intelligence.

Which is an epic series on YouTube. And by the way, everything we talk about, we'll wrap it up. It'll be in the show notes somewhere in the internet either on my blog.

I was about to say, if we reference something, we're going to chuck it in the description so that people can— Yeah.

Yeah. Well, that's enough of an intro of a second intro or whatever. But the show is going to be— we're not entirely sure what it's called. Maybe Burke Brothers Podcast, something. If you've got any ideas, send them our way. Otherwise, 3-ish things, because that's what we want to try and keep it to, just 3-ish.

3 points every week.

3 points every week or something. All right. So let's, let's kick it off. Amazon shareholder letter. Yeah. That's, we're completing each other's sentences already. All right. So what did you, did you read the whole thing?

Yeah, I read most of it. Well, all of it. Sorry. There's a few things that really stuck out for me and I'd like to start off by—

Wait, wait, wait. What is Amazon shareholder letter?

So 2 weeks ago, Jeff Bezos posted the Amazon shareholder letter for 2017. Yeah. Reflecting on their results of how they went and updates on how the company's going as a whole. Pretty much issued to the whole company, all their employees and the wider community. So shareholders to pretty much provide a general update on how the company's operating.

Yeah, here's what we've been doing. Here's what we want to do for the future. Yeah.

So the one thing that stuck, stuck straight out at me was Amazon Prime. Prime? Wow. Yeah, like 100 million people per year and now using Prime.

100 million?

Yeah, across— that's, well, that's globally. But yeah, I initially, when I read it, I'm like, no way, if that's the US.

It's not here in Australia yet, I don't think.

No, it's not. Yeah, because while obviously Amazon's recently new, they just came to Australia, so they obviously got a lot of distributions and logistics to sort out here. And obviously, like, obviously different legislation and pay wages. So they've got to sort out a whole lot of logistics to get their prices down to what they're known for over in the States and obviously other places like Europe, I would imagine, to being like the low-cost delivery provider. Yeah, the owner of the last mile delivery, that's what they refer to.

Last mile delivery, what does that mean?

So the last mile delivery is like basically the last— so being able to get you a package very quickly.

Yeah, like owning that last mile from point A to point B in the delivery, like having that efficiency, being on that last step of the pipeline, which is so important because I think I read in the shareholder letter as well that in 2017, third parties sold more on Amazon's site than Amazon themselves. So like Amazon Marketplace, which I'm not entirely across, but that's kind of insane to me that they just open it up. It's just this one massive marketplace for buyers, sellers of all people, not even— and Amazon are just like the hub. They're the hub. They even do the fulfillment and everything.

Allow people to use that, obviously their distribution centers and their logistics to sell their products, and they just take a share of the profit, obviously. Yeah, a percentage to cover what they spend and then also what the profit— obviously to cover their expenses and then also a little bit extra for them on the side. Yeah, pretty much just to own— just a way to own more of the market.

The Amazon takeover is real. What was actually one of my favorites was, um, day one, how Bezos says that Amazon's philosophy is that it's always day one, which I totally agree with, which is pretty much how we try to live almost at the moment is everything is a learning opportunity. And he attaches at the bottom of— because I read, I made the mistake, I read the 2016 shareholder letter again. I'd already read it, I read it last year, but it was, I think it's good. I think I took notes on both of them actually. Last year wasn't as long, so this one was a bit more full of numbers, I thought. Like, here's what we did. Last year was kind of like, here's 4 points that we think Amazon kind of giving a— or Jeff Bezos and Amazon's insight into what makes a good business. And of course, number 1 was true customer obsession. This is from last year's letter, so I'll just go breeze through these points quickly. Number 1 was true customer obsession. Are you delighting your customers? Number 2 was resist proxies. So that means when big companies get big, they can kind of get into a space where they just think they're fulfilling the process, and they think that the same process they've been doing for so long should keep getting results. Bezos in the letter said, "Do we own the process or does the process own us?" Which I think is a very good point to think about, especially not even on a big company scale, but on any level. Like, does the process— are you just going through the actions? Does the process rule over you or are you innovating on that? So, resist proxies. I'm not entirely sure what, in terms of proxy, that meant, but it'll be more in the shareholder letter. Embrace external trends. Which I thought was massive as well. And one of them was rather than fight against technological trends, make sure you're using them to your advantage because that's what he said. He's like, if you're going against technology and whatnot, you're essentially betting against yourself. And so the two main ones were, of course, artificial intelligence and machine learning. Ride the tailwinds of society. And the fact that they keep struggling with that in mind, they're struggling to keep Echo in stock. So, that's pretty insane. Exciting times. Yeah, exactly. And the last one is high velocity decision making. So, he said if you're 90% sure on what decision you're about to make, you're probably waiting too long. He's like, we like to keep moving fast. And have this saying in-house called disagree and commit. So, he gave the example of a couple of producers pitching him a pilot for Amazon Prime, and he said he didn't really agree with what it was, and he's like, well, it's going to take way too long for you to convince me, so I'm going to disagree and commit and know that you're capable of good work. So I just think that's— imagine having a leader like that. But that's the 2016 letter, which is a lot of value. But my favorite for 2017— what you got?

Just before we move on back to the 2017, maybe you could share your experience with them being so customer-focused with your AWS issue.

Oh yeah, that's a good point. Yeah, so customer-focused. I, last year or so, I've been learning AI technologies and whatnot, using a lot of them. A lot of these technologies require intense compute power. So what that means is essentially you— I use my laptop to build the basis of a program, let's say, and then because my laptop doesn't have enough computing power or horsepower, you need to dial into the cloud to recruit a bigger computer somewhere in the world, somewhere, somewhere else, right? So a lot bigger than what I could store in my room, or actually you could probably store in your room, it's just that you don't have access to it all the time.

And obviously it costs a lot of money to have that.

Yeah, it costs a lot of money to set up this type of hardware, right? So I dial into Amazon's cloud server, which is AWS, and I start up a computer that's somewhere in Sydney Sydney or the US, usually in Sydney because that's the closest to where we live. And I, for some reason, I forgot to shut it down. And this is a problem actually. I think, I think it would be good to get notified if it's running and not being used. Like, surely that's, that's something they could look into doing. And so essentially I had this computer running and it's quite expensive to keep running over a long period of time because it's just number of hours, right? They charge per second. So it's efficient if you're using it for just what you need it for, but not so efficient if you're just letting it run with nothing running, which is what I was doing. I used it for an assignment and then stopped it about, I would say, 2 to 3 weeks later once I received the bill, which was $700 US dollars. So that's almost $1,000 Australian dollars. And I kind of, I was like, far out. I was a student, I'm still a student, it was a lot of money. And I'm like, if I pay for this, I don't have much money at the moment. So I reached out to the AWS support, told them my story, said, hey, I'm a student, I was running this for an assignment, the assignment finished, and for some reason I thought I terminated it. Well, terminate means just pretty much pull the power plug on it. But I didn't, so completely my fault. But they turned around and said, hey, you know what, we can see what you've been doing. Thank you for reaching out to us. Thank you for using AWS. We're wiping the whole bill. That's no small bill, like, either. That's $700 US. So I don't know what their margins are, like what it actually costs them, but that's something that comes back to what he said. He's like, obsess over your customers. And I mean, they may take a hit on that, but I'll be a customer for Amazon for life. Like, I've I bought the Echo the other day and I use AWS almost every second day.

Yes.

Yeah, massive. But 2017 letter, we kind of went on a ramble there in the 2016 letter.

So let's go back to my first point on Prime.

Prime.

So 100 million people, if we look at the metrics, that's $9.9 billion in annual recurring revenue. Just Prime.

That's Prime, $99 US a year.

$99, yeah, $99 US, yeah, I should have said that first. $99 US a year.

100 million customers.

And 100 million customers, so there's $9.9 billion in annual recurring revenue just from that section of their business. Just that one revenue stream. That's just insane.

And that Prime members, I was listening, 2 in 3 households in the US have Prime.

I think I may have read some statistic that more people have Prime than own a gun. In the US. But I think, I don't know how accurate that is, but—

I wouldn't be surprised. That's, it's incredible. I can't believe like the scale of this company. And we were saying, actually, let's talk about something technology-based. So I read in the, we'll get back to that in a second. 2017 annual letter was a big, big thing was Amazon SageMaker or AWS, which is totally another business almost on its own. That's $20 billion revenue.

Yeah, insane.

$20 billion. So Amazon AWS SageMaker, which was released about 6 months ago, but I'm only just learning about it now. I should be more across these sort of things, which is essentially like— how do I explain this? So it's automating the machine learning pipeline, right? So if you're someone like me, true, who's running like data science models or whatnot, I can go to AWS, get into an environment, so like a web application per se, load in my data—

excuse me—

bless you, brother— load in my data into Amazon's databases, and then their machine learning algorithms that they've put in are made to run extremely efficiently. I think they were talking about 10x 10x, up to 10x faster on this dedicated SageMaker service. And it's, it's, I can essentially train my model in one click, train the model on the data, and then it even has an option to scale up to what you want it to do. So how much power you can dedicate. So you can just turn this little knob and say, I want it, I want this to run for X, and I wanted to use this much computing power. So that means it might cost this much at minimum, or it might cost this much at maximum. So that's upfront because a lot of the machine learning at the moment is kind of guesswork. Like you run your model, you adjust your parameters in terms of your settings, like knobs on a microwave or on an oven, and you see how it goes. It's guess and check work. But this thing, you upload it, it does, it works the test, a bunch of different models for you, and it does a bunch of hyperparameter tuning as well, depending on how long you want to let it run. And it's all, all packaged in little modules. So if you want to do the whole pipeline of up-training your data, deploying it to applications, and then whatever— oh, sorry, storing the data online. So storage, training, deployment, you can do it. But if you want to do the storage somewhere else, training on there, deploying somewhere else, you can break it up like that. And that one service had 250,000 customers since launch, and it was announced end of last year, November 29th. And I'm like, that, that is insane. It's making it more accessible. That's what I think is the end game for a lot of these, these cloud-based offerings like Google Cloud or AWS or Azure is okay, they're pretty accessible now if you have the know-how. But for someone like yourself who, and me, who are just getting new into the field and we haven't been developers for 5 to 10 years, I think like SageMaker is definitely very valuable, especially if you're a startup. Like you don't have the resources like a big business has. You could, you could get on there, store your data in one place, train your model in one place, and and deploy it all in one place. It's fascinating. But that's enough from me in terms of the ML side of things. Oh, actually, what else did you find in the letter?

What was your— With AWS?

Yeah, let's go. What do you reckon? AWS, was there anything else from the letter that stood out to you?

How they're really accelerating their pace. Of innovation in the areas of AI and machine learning, obviously, as you just touched on it, but it seems like AI and machine learning within the— within AWS is a very big focus point for them because they obviously see the potential and the size of that market. Yeah. And a lot of people are now starting to integrate machine learning into their organizations. So what's there to run it off?

AWS.

Yeah.

So AWS there, if you look at a graph, I haven't got one for the screen at the moment, but— oh, for the listeners, sorry. AWS, just imagine in the top right, and then all the other offices, imagine them in the bottom left. Like, that's how far ahead they are in terms of numbers, services, etc. But that being said, I went to a Google event a couple weeks ago, and they're ramping up. They're trying to— they're like, hey, I want some of this AWS pie.

And yeah, well, obviously it's 20— they're making $20 billion a year. Well, $20 billion in revenue of AWS alone. So, and it's just crazy to think that when you look at Amazon, they're a retail technology company essentially. Yeah, but in the background they silently became the biggest cloud server provider in the world. Yeah, so it's just crazy.

What were you saying the other day about them being split up?

So if the blunt— if regulation were to hit Amazon, I would just imagine that it would be easier for them just to spin off the AWS side of the business to instead of like categorize, like just take AWS out of the picture and have that as its own entity and then leave the rest of the business as a whole. Because obviously if it's doing $20 billion in revenue, like it's worth nearly $200 billion, that section of the business alone.

What do you mean? Like, so if it's what you just did, a math equation there, so $20 billion means that you're worth $200 billion.

Yeah, it's kind of like I was just kind of throwing out like a ballpark figure of what—

actually, I think I've heard that before. Like if you're doing $100 million RR, recurring revenue, yeah, ARR, you're gonna be worth a billion dollars valuation. That was like with Slack and all those other things. Okay, that's okay. I get that now. So you just kind of 10x it for a ballpark figure. Another massive thing I liked was Whole Foods, which I think is massive. The acquisition of Whole Foods, Amazon's— that's a space I really want to get into, I think. If I was to go over there, like, food is something I love. I'm passionate. Food's like my religion. You know how into it I am. And the fact that I think Bezos said, or in the letter it said, the acquisition of Whole Foods was the fact that first and foremost they wanted to get good quality organic food to everyone as many people as possible, which I think is a massive mission. Like, I totally agree with that, and I think they've got the infrastructure and technology to do it, especially with Prime and Amazon Go. Like, it's just gonna be just a merger.

It also just made sense for them to buy Whole Foods in the first place because they always wanted to go into the fresh food delivery side of things, because grocery— the everything store, as they're often referred to. Yeah, so like Obviously fresh food was their next target after they owned other parts of retail. Yeah, so buying Whole Foods pretty much saved them almost 10 years building out all the distribution channels.

Yeah, well now they've got 100, however many stores it is across the US.

Yeah, so the money—

what did they spend on it, like $16 billion? $13 billion. Yeah, like $13 billion to save them 10 years of building out those distribution centers and Yeah, so now they've got, yeah, like, like a Whole Foods positioned in all different places around America and they can just use that as another hub for people to pick up their Amazon parcels or, or, or do their grocery shopping or not.

And just, I'm, I'm with the Whole Foods acquisition as a whole. I'm surprised no competitor tried to like throw out a counterbid to Amazon's $13 billion thing just to like Even obviously you're not going to be able to outbid them, but yeah, but throw out a counter bid. Who would compete off the top of my head?

Walmart. Yeah, good point.

Yeah, like, one, like, obviously Walmart's got a lot of money as well, but like bid the price up. Yeah, they can pay a bit more because for $13 billion they stole Whole Foods.

And we were saying, imagine if when they come here we've got a So in Australia we've got a supermarket chain called Woolworths and we're just like, Amazon would come here, swallow up Woolworths. Just buy Woolworths. Yeah, and then they don't have to build the distribution centers, they've got stores everywhere. Kind of like what they did with Whole Foods in the US.

I see, I see that happening, but I don't at the same time because Australia, 25 million people population, we are growing very fast.

I'd love to see it happen personally. Yeah, I'd love Alright, little hiccup in episode 1.

We got the camera paused, but we're gonna jump straight back into things.

Thanks for being patient, podcasters. But yeah, where were we? So we were saying, yeah, on-device fingerprinting technique, which is to do with Amazon Alexa. My Alexa's about to go off. So if an ad is playing on TV and it says Alexa, Alexa knows that it's not your voice and won't activate. So when the Super Bowl ad was playing, which was Amazing. Your Alexa won't activate. And then there was a few more things, a new Kindle.

Which skill do you want to enable?

Sorry Alexa, we don't need you. There we go. I thought it was going to come on. Not quite, not quite the anti-activation yet. Just walk out shopping. Just walk out. That's Amazon Go. That's their new terminology for Amazon Go is just walk out. And they cracked half a million employees, 560K.

But you were saying, um, yeah, so it's funny that you say that, that Amazon's a hiring machine right now, but on the flip side, with them hiring a lot of people, at the same time they're patenting and also developing a lot of technology that's going to automate a lot of jobs. Yeah, so they hire one person and But that's why you personally might not get a job. Exactly. Yeah. My— the biggest example I can think of, of what they're doing is Kiva robots. So I don't know if you've heard about them, but Kiva robots. Yeah.

Yeah.

So they're the robots that Amazon uses in its fulfillment centers.

Warehouse.

In their warehouses.

No one working there.

Yeah, it's just like— and a scary photo would be inside of an Amazon warehouse because there is no— like, you don't see them.

You don't see photos in the warehouse.

You don't see any photos or videos of inside an Amazon warehouse because it's all robots and automated. So what happened with Kiva Robots was that was a new startup outside of, I think in Massachusetts. Yeah. Amazon was one of their biggest contractors, obviously one of their biggest buyers of their robot technology. Amazon ended up buying them. Taking them over for about $800 million. And as soon as Amazon bought them out, they cancelled all their other contracts with every other company that they were supplying to, just so that they could own the whole—

they got all the robots. Yeah, no one else is doing robot warehouses, just us. Yeah, that's the great white shark of Amazon. But I think that's enough for the AWS shareholder letter. Um, Facebook. Had its F8 event for 2018. Yeah, did you watch the highlights at all?

I watched a few. Yeah. You want to kick it off with what you thought was good?

Yeah, so let's go for the tech side of things. I think there was one— I don't know whether to be like this or to be scared or to— I'll get your insight and what it might look from a business point of view, but Facebook, they essentially, they're going all in on camera, right? They love cameras. Zuckerberg loves camera. And I'm a bit wary of time here because we want to keep these episodes within 30 to 45 minutes or maybe a bit less. So long story short, they want to grab your old photos. And so old photos on your devices or just photos of, say you took a photo of this point of the room and then we took another photo of that point of the room. And if you make an album with that, Facebook want to use computer vision to reconstruct the room. So you can take a 3D tour of your— of the room and see the photos all in one kind of image, which I think is pretty awesome with— as far as computer vision technology goes. But in terms of what that means in how they understand, like, literally the items in your home, And from an advertising point of view, is this going to be like they know what brand of blender you use now and all of a sudden you're seeing that brand of blender pop up in your newsfeed?

Well, yeah, from a business point of view, just hearing that Mark says that they're obsessed with the camera. Yeah, and being a camera company almost is just another dog shot at Snapchat because what does Snapchat claim it is? A camera company, essentially. So it's just like another reason why— well, another stranglehold placed on Snapchat is that Facebook just continues to go after it and after it and after it. Yeah, kicking Snapchat while it's down almost. Yeah, but I mean, as well, I guess, as you said, like, if they have a model of your room Yeah, they know your kitchen. They have all that details. They know what sort of brands you're using in your kitchen. So that's so much— that data of knowing what's in your kitchen is so valuable because they could go— they now have—

what food you eat.

They have— yeah, they now have more quantifiable data where they can go to like a big retailer. Like say for example, a fridge producer. They can now go to you and say, We know this many people don't have your fridge. We can target ads at these people on our platform. Yeah, so that becomes a lot more valuable to the supplier of fridges, for example. Like, this is just about it. This is a bad example, but you know what I mean?

No, no, no, no, no example at all. I know it's exactly, exactly the kind of example like— and look, we may be overthinking it here because if you just watch the video, I think saw about a minute or 2 minutes of what it actually does. Does. But that being said, with what's happened recently with the Analytica thing coming back, I think a good point on the event I saw was the ability to clear your history on Facebook. I think that's massive almost.

That's been there for ages though.

What do you mean? Like, you can clear like your entire browsing history of apps, products, and external websites? Yeah.

That's been there for a while.

Well, there you go. See, that's— I'm not even on Facebook at the moment.

But I guess obviously on the flip side, people just didn't know about it.

Well, that seemed like it was a new feature that was being announced.

Because after the Congress hearing— There you go. Speaking of that, but like, I'll just quickly— after the Congress hearing, when you logged on, when I first logged on Facebook, I got a notification at the top to say that you can now do that.

Oh, well, that's good. Yeah, well, that's— it's recent. It's fairly recent. Yeah. But they just— we've got enough data, we don't need any more. Let's— how much data do we need? And then there was a couple more things, which was they're hiring 20,000 people to work on manually auditing ads, newsfeed items, content, etc., etc. 20,000.

Yeah. My opinion on that, it's just to avoid regulation. What do you mean? Well, after the con— like, you could— depends on how you take it. The Congress hearing obviously didn't go well for them. Yeah. But like, you look on the flip side, after the Congress hearing, their share price didn't take a hit. Their share price is still on the climb.

So, so they're going to hire 20,000 people to be like—

how good does it look in their perspective to come out and address those points and just go, We're now going to make 20,000 new jobs.

Yeah, that's what I was saying. Like, hey government, we're hiring 20,000 people to curate this stuff so you don't touch anything from us. Yeah, we'll just keep it all, all within the Facebook world.

It's not going to make a dent on their profits either.

Yeah, it's like the 20,000 people, what's that, $50k a year, let's just say, that's a billion dollars. Yeah, but I suppose that's nothing compared to like—

what's spending a billion dollars to protect their monopoly?

Yeah, yeah, that's true.

Nothing.

That's— yeah, to protect their, their complete pipeline control of, of what information people see, which is crazy. Um, what else? Yeah, the F8. There's— otherwise there was Oculus and a few other things. Um, watch the, watch the video if you want. Otherwise it was like group video chat for Instagram, which was, I thought that's gonna come eventually. Group video chat for WhatsApp, group video chat for Messenger, just the same features across, it seems 3 different chat apps almost now with Instagram with a focus on, but I think that's gonna be the next big platform, even bigger than Facebook.

I was about to say, yeah, Instagram's their big play right now and their big money maker. That captures the most attention and has a lot more people.

I use it.

It doesn't have, it has, I don't even think it has half the people on Instagram that does Facebook, but the attention it catches is Yes, keyword. Yeah, yeah, yeah. And I think Instagram is always in the top, either top, either the first position or the second position for the most downloaded iOS app.

Yeah, it's always, always one and two.

Yeah, yeah.

And so where does AI come into this? Well, of course, with the computer vision they're using in the, in the room for the Facebook camera app, they're also using AI to to fix the newsfeed, like trying to detect fake news versus real news, which I think that's a very hard problem because they're— I think AI is still blown out of the water in the fact that it can do some superhuman stuff, which it can, but if a human expert struggles, a machine learning model is also gonna struggle a bit.

'Cause obviously the machine learning models is developed by humans.

Yeah, exactly. So if a human expert on content curation can't tell whether something's fake or not, It's going to be hard. But that being said, Facebook on another AI side released— never underestimate the power of these big tech companies, AI firms and research, because they essentially replicated AlphaGo Zero, which is DeepMind's Go-playing engine. If you haven't heard, DeepMind built an engine which can play Go better than any human player in history and better than its own engine it built a couple of years ago, it beat itself 100 times, and essentially it learned from scratch. Like, they didn't teach anything, they just said, "Here's Go," a few minor settings maybe, and it learned. It played against itself and just got better than anyone will ever be at Go. But Facebook— so DeepMind is Google— Facebook AI Research built their own version called OpenGo, I think. And the reason they called it OpenGo is because they open sourced it. So that the code for it is all online. You can look how they did it. I don't think— I'd love to see their one play AlphaGo Zero, like it'd be a contest. Imagine that. I'd watch that live. Yeah, like I would 100% tune into that. Like, I think someone should show me that.

I think a lot of people would.

Yeah, so they— and the code's online, and they also released it along Maybe the framework's been out called ELF, which is something— E, experimental maybe. I don't know what the E stands for, but Experimental Light Framework, which is a reinforcement learning framework, which is essentially what they used to play AlphaGo. They used this framework. So if some another developer like me and my team, who we've been working on an OpenAI competition to build a super playing Sonic the Hedgehog, we can use Facebook's framework, which is all open source. I haven't dived too much into that, but I think that's really cool for Facebook to go, hey, we've seen DeepMind do this really cool thing, we're going to replicate it with our own engineers and our own scientists and then open source it for everyone. I think that's how everyone wins. It's— if it's open source, Okay, sure, you might not make as much money, whatever, but is that the goal?

You contribute to the development and the progress of the field.

Yeah, right.

By making it open source.

Yeah, exactly. So think about if you're a young developer, you're just getting into the field and you go online and you want to learn this stuff and you jump in, you use Facebook stuff because all I know— where you gonna work later? Yeah, because it's all open source. Where are you gonna work? It's— you're gonna be, I want to work for here because that's who got me a leg up. That's— I think that's, that's like the 10-year rule, a 20-year rule. It's like you should always, always be nice, always whatever, because you never know when it's gonna come back. Exactly. Yeah, but I think otherwise, I think that's enough. They got the Amazon shareholder letter, Facebook F8 announcements, Facebook open sources its own version of AlphaGo. Oh, we did have one more thing, but I think for now we'll leave it. But it's essentially the sleep podcast from Joe Rogan with, I think it's Matthew Walker. I'm not entirely sure. But you listen to it.

I don't know.

What do you think of the sleep podcast with Joe Rogan?

I think everyone should definitely check it out. Yeah.

What was your takeaway from it?

I'd probably say that I really was drawn by the fact that when they changed the school time. So, originally a school in America used to start at 7:30. Their SAT scores or something like that.

Test scores, yeah. They start at 7:30, yeah.

Their average test score was like a 1200 or something, which apparently is still very good. But when they moved the start time to 8:30, their average test score went up to 1500. Which is just insane.

Yeah, and I think that the other one I was blown away with was the doctors doing 30-hour shifts or something, which is inhumane. And it all started from this one doctor back in the day who did coke. No, he was addicted to coke because they used it as a pain treatment and he was testing it on himself. This is about 100 years ago. So essentially the podcast was Joe Rogan, who's an awesome dude, with Michael Walker or Matthew Walker. Daniel Bourke, sorry bro, I forgot your actual name, but he's a PhD researcher in sleep, essentially, for 20 years. He's wrote this book and it's just blown me away. I already value my sleep highly, but this thing, anyway, this doctor, 100 years ago or something, they used cocaine to treat pain. And suddenly he got addicted to cocaine. He would do 30-hour shifts because people were like, this guy's insane, he's working so long. Well, the student doctors need to keep up with me. And the interns, something like 1 in 20, because they're so sleep deprived, 1 in 20 ends up killing someone accidentally. Because 1 in 20.

Yeah.

Doctors, new doctors, student doctors. And 1 in 5 injures someone, which is ridiculous. Or you're 400% or something more likely to make an injury after a certain period of time. Like, it's just the data that they had, and I'm probably mis— mis— misreinterpreting some of these exact numbers, but when you listen to them, they're just ridiculous. Absolutely ridiculous. So you'd— I— if I go for another any surgery, I'm going to be asking the doctor, how much sleep have you had? Because there's something like—

well, major surgery.

Yeah, major surgery. Like, if the doctor hadn't had 'X' amount of sleep. And this is no, like, no malice towards doctors. It's just the one case example. It happens everywhere.

Yeah.

Um, how much sleep have you had? Because something like 170% more likely to make an error if they've had 6 or less hours sleep in the past 24 hours, which is, I think, something so easy to fix. But who knows? Who knows? I'm going to be sleeping more, let's put it that way.

With that being said, it shows the value that AI and robotics and machine learning will bring to the healthcare world. But we'll talk about that next podcast most likely.

Yeah, maybe in a future episode, but that was fun. We had a little bit of business, a little bit of technology, a little bit of health and 3-ish topics, Amazon, Facebook, and health. But otherwise, what do you think the question of the week is? What can we give our viewers to go out and do? It's a good one. Yeah, question of the week.

You caught me off guard here.

Yeah, question of the week is go check out one of the things we talked about today and send us in, send us in something. Actually, that can be challenge, challenge of the week. We're kind of thinking of this on the fly. So, challenge of the week, go through one of the resources we've shared with you and share, share your favorite thing with us. If you want to ask us a question or if you want to talk about a topic, send me an email, daniel@mrdbourke.com, and we'll get it on one of the next shows. If you can send in audio, we'll try to play it on the show and we'll go off that. So if you've got any questions, any advice, whatnot, send it in.

Feedback's appreciated for the first one.

Yeah, actually what worked. This is the first episode. So what worked? What didn't? What do you want to see more of? What do you want to see less of?

What do you want to hear more of?

What do you want to hear more of? More technology? More health? More business? Who knows? We're still working this out. But otherwise, if you want to find out more, all the links will be in the description or the show notes. But otherwise, do we have a sign-off? No. Oh yeah, Will, next episode Will's going to do upside down. He promised me. So that's, that's, he's going to do the whole thing upside down. We'll see you next week. Bro, you're gonna sign off?

I'm outta here, time to go play some Fortnite.