The Vergecast - Macbook Pro review, Lenovo smart display review, and the Data Transfer Project

Episode Date: July 27, 2018

After a week of speculation on whether the new MacBook Pro was throttling performance under heat, Apple released a software fix to address the problem. Dieter’s review for the laptop (post software fix) published on Wednesday so Nilay, Paul, and Dieter discuss what happened with the product and how it performs. Dieter also reviewed the new Lenovo Smart Display, so there's a chunk of the show dedicated to that device, which exceeded expectations. And we’ve got another week of Elon Musk. Transportation reporter Andrew Hawkins gives us an update in what has been happening in the world of Tesla, SpaceX, and the various endeavors of the unpredictable billionaire. We’re going strong with our weekly interviews, and this week we have Brian Willard from Google to discuss the Data Transfer Project — Google’s initiative designed to transfer data from one service to another without downloading and re-uploading. We also can’t forget the segment Paul does every week called “What is it like to be a dolphin?” so keep listening for that. But you know, we’ve got a whole lot more in between that — so listen to it all and you’ll get it all. 02:30 - Apple MacBook Pro review (2018, 15-inch): the heat is on 22:01 - This week in Elon Musk with Andrew Hawkins 26:53 - Lenovo Smart Display review 35:44 - Data Transfer Project with Brian Willard of Google 59:46 - Paul’s weekly segment “What is it like to be a dolphin?” 1:02:30 - Amazon’s Alexa Cast makes it simpler to play music from your phone on Echo speakers 1:10:00 - Facebook growth slows in aftermath of privacy scandals Learn more about your ad choices. Visit podcastchoices.com/adchoices

This episode of The Vergecast is brought to you by ZipRecruiter. Hiring used to be hard. We had multiple job sites, stacks of resumes, a confusing review process. But today, hiring can be easy and you only have to go to one place to get it done, ZipRecruiter. com slash verge. ZipRecruiter sends your job to over 100 of the web's leading job boards, but they don't stop there with their powerful matching technology. ZipRecruiter scans thousands of resumes to find people with the right experience and invites them to apply to your job as applications come in. ZipRecruiter is so effective that 80% of employers you post in ZipRecruiter, get a quality candidate through the site within the first day. Right now, Vergecast listeners can try ZipRecruiter for free at this exclusive

web address. Guess what it is? It's ziprecruiter.com slash verge. Whoa. Oh, that's me. That's your joke. Welcome, Virchcast. I was ready for like a symbol crash or something. I left my symbol at home today. I'm sorry. I'm your buddy, Nilai. Paul is here. Hello. Deeder. What is your relationship to the audience this week, my friend? I'm your good buddy this week. Oh, good. Dieter's on a roll.

A very quiet week in news. We got a good show, though. We got to talk about the MacBooks. Dieter reviewed them. Yes, I did. Well, one of them. An emotional journey for all of us. Deeter also reviewed a new, the new Google Assistant displays, the Lenovo smart speaker.

That's me cool. A bunch of Facebook news, which Deeter is awkwardly going to ignore because it's my works on Facebook. But, like, Facebook experience today, as we're recording, largest stock drop ever. Like, just huge. Billions of dollars wiped off. The biggest, I think you, there's got a story on the verge. I'll read the headline to you. One of the biggest American history.

And this is neat. I'm excited about this interview that we did. Last week, Google announced a new data portability initiative, which sounds very boring. But basically, Google, Microsoft, Twitter, Facebook, all teaming up so that with one click of a button, you can move your data between all the sites. So you got your photos and Google photos. You like, screw it. I hate it.

I want to go to Facebook. One button, they go to Facebook. It's assigned to increase competition, give you, control of your data. I talked to one of the guys behind the projects, a Google engineer named Brian Willard, so we're going to do that in the middle of the show. It's super interesting stuff. Like, we always talk about competition and lock-in. Exciting. This is Google driving forward an industry initiative to make it easier to move your stuff around

without having to download it all from the cloud and re-uploit it somewhere else. And there's some stuff in there that's complicated. So we're interviewed Brian Willard and Miner Show. But let us begin, Dieter, an emotional tale, a man in his MacBook, and whether or not the processor operates or not. Oh, my God. And it ended up, like, we made a video that, like, reflected the emotional journey, and we maybe put too many of our emotions in that video. There are a lot of emotions.

So, you know what? The review is not even the most important part. What I think of what happened is not even the biggest deal here. The biggest deal here is Dave Lee, the YouTuber Dave Lee created a test. that revealed a massive software bug. I don't know, massive, maybe minor, simple something. They revealed a software bug that then Apple finally repeated to verify, repeated his exact workflow to export video.

And then they had to issue a software update that completely changed the way that the computer handles thermals. I think it completely changed it because it changed our tests. So that, like, in the space of a week, they released a laptop and then released a software update that changed the way that it handles heat so that it could actually run at the speeds that it's supposed to run. So let's start at the beginning. So everyone gets the MacBooks for review.

You got one for review. You know, we went to the party with the hot shot teenagers. Everyone felt bad about their careers. Laptops are given to us. We start doing the tests. The tests look good. Other people that we... They looked fine.

They didn't look great, but continue. Dave Lee posts this video. And I'll credit to him. found the bug and he says, hey, when I run Premiere, it was Premiere, right? Yeah, it was a particular kind of export in Premiere and how particular it was and how edge Casey it was is very much open for debate. Sure. So I'm running this export in Premiere and I'm noticing the processor clocks way down.

And this is my favorite part of his video. If I stick the computer in a freezer and do the export while a computer, it does it. So the problem is obviously heat management. Then, Deeter, just correct me if I'm wrong. Then the Reddit sleuthing began, and everyone tried to figure out what's going on. Everyone's running benchmarks. Other people that we know and talk to Jonathan Morrison runs like 500 tests on a live stream. Like he went for it.

You like test can't find the bug. Other people can't reproduce the bug. Other people can reproduce the bug. What happened then, Deeter? Well, so there's been many theories about what precisely this bug was. And the one that seemed most likely, there was a guy on Reddit who started, the voltage regulator module started messing with that. Like that might actually been the source of the heat problem.

Anyway, his fix seemed to work better. So anyway, Apple released a software update, and they also released a statement. And following extensive performance testing under numerous workloads, we've identified there is a, quote, missing digital key. Yeah. In the firmware that impacts the thermal management system and could drive clock speeds down. under heavy thermal loads in the new MacBook Pro, blah, blah, blah. We apologize to any customer who's experienced less than optimal performance.

So here's the thing. I don't know what a missing digital key is. No, it does. Do not know. Here's my completely uninformed theory. Go ahead. Intel specifically accelerates certain video encoding, and Apple forgot to pay the license.

They didn't enter their key. Yeah. That's like, that's the best theory I've heard. I mean, we fell down a hole of like, I was like reading Intel white papers the middle of the night. Yeah. I don't know anything about how to do that. So that was confusing. It's possible that in the past 24 hours someone has actually figured out what it means.

But the problem is that VRM thing I mentioned earlier, they exposed to said, nope, that's not it. And Dave Lee's got that in his updated benchmarking video. Every time I'm like, what does that mean? Does it mean this? No, it doesn't mean that. Well, what does it mean? Digital key. He's like, ah.

So I do not know. I haven't looked in the past 24 hours, but the precise outlines of what they've done seem to be along the lines of what has been discovered, but it's not, like, precisely that. But it is apparently a simple fix. And it does, in fact, work in our testing. We ran a bunch of benchmarks,

and then we ran them again after the software update. We are not seeing the performance increase we would like to see on Adobe, but Adobe Premiere Pro CC. on the Mac is, you know, but we have seen remarkable performance on Final Cut 10, which is not what I expected at all, all of which is to say that benchmarking is a dark art and tiny changes in your setup can have huge changes down the line. If you watch some of the benchmarking videos, you'll see that if you, depending on the initial state of your video, if it's a 4K, what kind of 4K, depending on what you render it to, depending on what bit rate you want that final file to be, you know,

you can just go down the line and you're going to get radically different tests all around. So when Apple says it's, you know, 70% faster, I mean, that's on one particular test. It is, in our experience, anywhere from we saw off like just 10% to almost not much of a percent at all, all the way up to 50% on some of our tests. Let's go back to Digital Key. I'd like to just live in that moment. Apple, okay. How did they make this mistake?

They won't, they're not, no, we're going to tell us. But it's like very, remarkable. Right. Like they must have run tests. And if all they're saying is, we didn't run this one edge case. Then the performance they shipped it with across this other variety of tests must have been acceptable to them, which is great. But then they turned the key. Right.

They did something. Well, they found the key, but they were busy. They located this key. And now the computer is way more performant on certain tests, right? How do they not know that? Like, this is their primary skill set, is shipping computers, right? Like, how do they not see the results they were getting out of this hardware and realize? Well, they only do it every, like, four years.

He's like, I'm always forgetting that key. It's not true because they haven't done the key in a long time. That's the part that I don't, that I think is troublesome. And, Deeter, your review, both the video and the written review, they're centered on the notion of trust, right? The previous MacBook had a trust problem around the keyboard, which, you know, you can believe Apple that it was a small problem or not, but that became the narrative of the thing that you probably can't trust this keyboard. There's a trust issue around the entire concept of the touch bar. Is this a good idea? Do I trust Apple to take me down this road where I'm typing on glass instead of just pushing an escape button?

One positive in this, if you want like a real silver linings of this, this is a problem Apple, fix because Apple is responsible for the whole stack, the whole software stack. Yes. It's a problem Apple can fix. But the question is they fixed the software. Cool. And they maybe did that in collaboration with Intel. I don't know.

Maybe it's an Intel's problem and whatever. But Apple controls more of it than usual. So it was Apple's problem to fix. When the meal is bad, you don't blame the butcher, you blame the chef, you know, when the steak is bad. Okay. Sure. When the steak is bad at the restaurant, you don't blame the butcher. You blame the chef.

If it's rotten meat, you don't get mad at the meat seller. It's the chef whose job it is to create the meal that you eat. Because they're the gatekeeper. And they should stop you from having to experience that rotten meat. So Apple's the chef. The digital key master. Oh, man, I had so many keymaster gatekeeper puns with this missing digital key. So here that, let's just be in this metaphor together as a family.

Here, the steak is the core I-9 chip. Right. That's correct. The six-core core I-9 chip. And under certain eating patterns, over-cooked it. And if you put it in a freezer, it tastes great. Just to be clear. And Apple is the chef, and the chef forgot to season the core I-9 with digital keys.

Am I tracking along here, Dieter? Yeah. I don't think this is obviously not Intel's fault, right? Like the hardware is fine. They just ship, they just give them chips. It's Intel's fault insofar as the only chip that Intel has to sell to Apple is a big, fat, hot, six-core chip. That is not Intel's fault. That's just what they make.

That's what they make. And you could theoretically imagine an alternate future where Intel had shipped a 10-nometer chip, and then this MacBook was designed for it and everything would be fine. But then they forget the key, and everyone's getting lower performance. And it's still Apple's fault. Should Apple have redesigned this computer for better thermals, made it thicker, done something else, spent a bunch of money to re-engineer the thing for this chip? No, let me defend that. I disagree.

Like, that's a good argument. I don't know the answer to that question. I think the answer to question is probably no. They shipped a misconfigured computer, right? So now the computer works as designed inside the case that is able to handle the heat. The computer was misconfigured. I don't think the question is, should they have designed a case that would have accounted for their misconfiguration? That seems backward.

Oh, no, no, no. They should never, not that. But, like, assume now that it's configured correctly, do you still think that this is the right case for this chip? Because even configured correctly, you theoretically are still, you know, running into thermal limits on this. I guess that would be my question for you, the person who reviewed this laptop. Yes. Well, so my answer is... The chef has become the butcher!

I hate to send it back to the kitchen. Who's the butcher now? My answer is probably not. There are many benefits to the Mac. You get Mac OS. You get a thin computer. You get, you know, all the stuff. All the stuff that you get with the Mac.

But the thing that this MacBook was supposed to do was help address a lot of the malaise in Macworld. Where are the mid-range Macs? Where's the Mac Pro? when are they going to regularly update their Macs with the latest processors on a good cadence? Like, what's going on at the touchbar? What's going on with the keyboard? That sense of, uh, we were hoping that this MacBook Pro would be the thing that people could just be genuinely excited about.

The thing that everybody has said to me is, do you remember back in the day when Apple would release a new Mac and you would just be excited about it? And that was the end of the story because you knew it would be good and it was like, I can't wait. Do you remember that? But what is it like using a laptop with like? a core I-9. Like, what is... I mean, it's stupid fast. I don't know what to tell you.

Like, how many Chrome tabs before it slows down? The thing that you can do to slow it down is like just attack it with like serious video exporting or, you know, serious lightroom exporting or that kind of stuff. But to me, like, I honestly think that most people don't need a 15. They want a 13. And the 13 is pretty overpowered for what most people need. What Apple really needs to do is release a good mid-range laptop. and they don't have one right now.

I was talking about this last week. You know, after the software fixed, this thing is great. But should you trust it? Well, the keyboard seems fixed. Seems more trustworthy now. They fix a software bug. That seems more trustworthy now. So, yeah, but that doesn't mean you feel good about it

after they shipped a misconfigured computer and they are lying. They're lying by omission, at least, about why they redesign the keyboard, right? Like, those things don't affect the computer's performance at this point now that they've released the update and whatever. They don't affect the computer,

but they definitely affect the sort of of culture of the computer exists in. I had the first-gen MacBook Air, which I had like, it was like one and a half. It was like not enough cores somehow. Yeah.

Did you say one and a half? I don't know. It was something weird. I just remember that it would throttle down when it would get hot. Yeah. And it would just literally freeze. And it would just sit there and be completely...

I'm not Becky that computer. And I had to get like... Ages ago. Special Thursday. third party software hacky stuff to like keep the temperature. Beck and I weren't married yet, so I did not install third party hacky stuff on our computer. I remember making that choice and being like, I can't support this. I don't know if this is forever.

So I feel like this, it's not like Apple's never like screwed up in this way, and it seems like they've addressed this one better. My worry with this one is that Apple, like Intel has released hotter chips than Apple's hardware can handle. And if the case is just Apple screwed up with software, yeah, that's disconcerting. But at least like it seems like... So when the bug was found, the argument was Apple put hotter chips and can handle in this case because Johnny I have demands that everything be thin. Right. The misconfiguration revealed Apple made a mistake. I think the most revealing thing is the Deeter is getting better performance in final cut.

All these other things that are not edge cases after applying the update. Yeah. Right. And that is like the, they should have noticed that. That should have been the thing that tipped off the misconfiguration, not Dave Lee. And I think the thing, you know, we talk about iPads and future and mobile. And like, if this had been an iPad and Apple's ecosystem was as closed as their other ecosystem, there's a chance of no one would have ever found out, right? Like, you can't run those kinds of benchmarks. There isn't there isn't third, there isn't the Intel power gadget for the 811. chip, people can't write their own kernel extensions to re-regulate the voltage control. Like, this is like a, this is a power of open computing story in like a big way because someone was able to run the tests, match them up to what they thought, actually measure in a way that you cannot measure on some of the more closed platforms. And then other people were able to investigate the working of the computer to identify

a solution. And I think the guy who identified the thermal problem. He told Hym on our team that Apple solved the same problem in a more elegant way. So they didn't do what he did. They went higher up in the system and solved the same problem. But this doesn't happen with phones, right, like in a large sense. Well, except when they explode. You don't need the Intel Power gadget to be like my phone's exploding.

I mean, there is a wide feeling that intent. Well, especially as like Moore's Law, slow down. Intel's done a lot of things to kind of confuse the X-86 instruction set and like optimize very specific use cases. And it is a very, very complex chip. And Arm is obviously a lot simpler. Especially if you design it yourself. Obviously if you're operating system. Yeah.

And so there is a vibe that like how much complexity can we truly sustain. At the same time, that's not, you can't give Apple a pass if everybody else was able to find the missing key. It just sounds like an RPG, man. There was a super interesting Stevensonovsky thread a few weeks ago, several weeks ago, about just sort of what's happening with Intel. And he was like, when I ran Windows, we would go to Intel all the time and tell them what we needed is Microsoft to make the computers thinner or whatever. And Intel's response was, just use our proprietary extension. Yeah. Like over and over and over and over again.

And they consistently said no because they want to preserve their options. And what they couldn't convince until that they should do was build better graphics. And you just go find it, read it. It's super interesting. Like if you're into that stuff, Snofsky's been posting a lot of that stuff lately. And it's just, Intel just like look them in the face. And we're like, we don't need to do graphics. But would you like to use this proprietary extension that speeds up one function in Excel to lock Excel to our chip?

And they were like, no, we would not. So Dieter, end the line. Do you trust the computer? I trust the computer, but I don't love the computer. That was my conclusion. And maybe that's unfair. But it's not just that there were a couple of controversies and things that put a bad taste in your mouth leading up to the software fix.

It's that in general, the grand experiment of this design, only the USBC slash Thunderbolt ports, the touch bar. I don't know. The whole thing is just like it didn't pull off the trick that it. it was supposed to pull off of pushing Macs forward and getting rid of dongles. And like, the whole thing just didn't quite get there. And the brutal truth is it's good. It's a very good laptop.

But I am more excited for the next big redesign of the MacBook now than I was before this thing came out. But, like, it's great. It's the best laptop I've ever used in my whole life. There it is. It is the, like, and the most powerful anyway, the best is probably still the MacBook. Little MacBook, the second generation of that one, even though it's underpowered. 2015 MacBook Pro with discrete graphics living that life. The number of people who have reacted to this review cycle by digging up 2015 MacBook Pro

as I've seen on Twitter is a lot higher than Apple should want it to be. We can talk about USBC all day a long night, but I was just thinking back to the event where the photographers and the videographers and all the fancy people. And they just kept on showing us pro workflows. And almost every single one of them involved reading a card. and all of the pros they showed us for like the card read speeds are off the chart so fast it's like why don't you just put a card reader

like you you're just showing us that what people need is a card reader in their computer more than anything and it's just stuff like that that I think is just leaves a bad taste in everybody's mouth that the industry is not ready the ecosystem of things is not ready for this computer

to only have the USBC ports same bring back Firewire 800 400 400 that's my jam FireWire 400 is like a a remarkable I.O. standard. Let me tell you about this. Firewire 400 is why the iPod existed. Yeah. They couldn't have made the iPod.

Anyway, I'll stop now. It's the best. Okay, I'm going to read an ad, and then we're going to have this week in Elon with Andy Hawkins, another wild week in Elon, and then we're going to talk about these smart speakers. This episode of the podcast brought to you by ZipRecruiter, SmartSuite to Hire. Hiring isn't most people an idea of fun. It can be exhausting. It's not easy, and you know what? It's not supposed to be. If you aren't having tough conversations to make sure you get the right person for the job, you're not doing it right.

But how can you stay focused in finding the most qualified candidate with ZipRecruiter? The powerful matching technology scans millions of resumes across the network to actively find the right people, the right experience, it invites them to apply to your job. As applications come in, it spotlight to top candidates to save you time and make sure you never miss a great match. It's so effective that 80% of employers who post on ZipRecruiter get a top quality candidate through the site within the first day. So if you're hiring, it's time to get smart, Paul. Go to ziprecruiter.com slash verge right now where you can try ziprecruiter for free, the lowest risk price there is. That is some deep econ in the middle of the ad copy. Don't waste another second. Go to ziprecruiter.com slash verge and start putting that technology to work for you. That is

zipprecruiter.com slash verge free the lowest risk price there is. That's not their tagline, but I like ZipRecruiter. Hey, this is Andrew Hawkins, a transportation reporter at the Verge. and this is the week in Elon Musk. Musk calls Tesla Critics' boss to complain about him. The unpredictable billionaire lashed out again this week, contacting the employer of a well-known but anonymous blogger who dared to criticize him online.

Montana Skeptic is a somewhat well-known writer for the financial news and analysis website seeking Alpha, who has spent years tearing down Tesla. Skeptic has long been upfront about the fact that he's a Tesla short, meaning he's betting the company's stock will nosedive at some point. This made him a target for Musk's loyal fans on Twitter, who doxed the blogger and revealed him to be an employee of a firm that apparently has invested in oil companies.

Using this information, Musk allegedly called skeptic's employers and even threatened to sue him. Skeptic subsequently deleted his Twitter account, but not before posting a final message online in which he said, quote, you might say Elon Musk has won this round. He has silenced a critic. but he has many, many critics, and he cannot silence them all, and the truth will be out. We all know that Elon is extremely online and extremely logged in,

so it's too soon to say whether he went too far this time. SpaceX launches and then catches another rocket. This week, SpaceX launched one of its Falcon 9 rockets from the California coast. According to the Verge's own Lauren Grush, the mission was completed in a little over an hour. The rocket's booster managed to land on the company's drone ship, in the Pacific, despite poor weather conditions at the time. There was some confusion at first about whether it survived, thanks to bad lighting on the boat, but SpaceX later got visual confirmation that the booster touched down successfully.

But the same couldn't be said for the rocket's nose cone. The company's recovery boat, Mr. Stephen, saw the structure falling from the sky but was unable to catch it. This was SpaceX's seventh mission for one of its long-standing customers, Arridium. Musk's rocket company has a contract with Aridium to launch. 75 satellites which provide global telecommunications coverage. After this flight, SpaceX only has one more group of satellites to launch, and then its job for Eridium will be complete.

Tesla whistleblower meets with SEC investigators. Martin Tripp, the former Tesla employee and self-described whistleblower, is meeting with the Securities and Exchange Commission to discuss his case. The meeting in itself doesn't mean that the SEC has opened a new investigation into the company, but it could be a sign that Tripp's case is gaining traction. In case you don't remember, Tripp leaked information to reporters about safety violations at Tesla's factory. Musk subsequently portrayed him as a disgruntled employee in Saboteur.

I love that word saboteur. Keep an eye on Musk's tweets as this story progresses. Speaking of, Twitter locking accounts using Musk's name to combat crypto scans. This is a fun one. Twitter has implemented a new method for combating cryptocurrency scammers. It now automatically locks unverified accounts that change their display name to Elon Musk. If you have a non-verified account that is not associated with a phone number, changing your display name to that of the billionaire CEO of SpaceX and Tesla

will result in an immediate lockout. Twitter will then ask you to pass a CAPTCHA test, as well as provide a phone number to regain access. The measure is designed to combat the innumerable bots on Twitter that are imitating Musk's very active Twitter profile, to scam unknowing users in replies to his tweets, to which I say, live and learn. And finally, SpaceX Hyperloop Pod Race breaks new record. Last Sunday, Musk was in Los Angeles at the Hyperloop Pod race put on by his company SpaceX.

20 teams of engineering students competed to see whose pod could go the fastest through the company's nearly mile-long test track. The winner was Varh Hyperloop, a team of 40 students from Munich, Germany. Not only did they comfortably beat all the other teams, but they apparently broke a new Hyperloop record achieving a top speed of 290 miles per hour. But while the event was about Hyperloop pods going really, really fast, it was also a celebration of all things Musk. There were Tesla vehicles, SpaceX capsules, Boring Company pyramids, and tons of merch for sale, as well as food trucks and an 80s cover band. Musk was there, as well as his girlfriend, Canadian pop star Grimes. It was like Bonaroo meets the X Prize if that's your kind of thing.

This has been the week in Elon Musk. For more news, check out Theverge.com. That Elon, man. It's a busy fellow. That special Twitter rule is something else. I'm just happy that our staff responded to it by all changing their display names to Casey Newton. Which is incredible. All right, Teter, you did another big review this week.

I'm still here. A new, are you still our buddy? Yeah. Good buddy. I don't know. That's more threatening than I anticipated. All right, we saw these at CES. The three of us actually played with them together. The Lenovo smart displays.

They're out now. It only took seven months because it's a whole new platform. And you put it in them. Yeah, so the Lenovo's the first. They're called the Lenovo Smart Display. There's going to be other smart displays. It's the Smart Display platform. And, yo, it's really good.

Is it? It's really, really good. It runs Android Things. And what that basically means is it's just a, front end for the Google Assistant. Like, you can't really open up a web browser. You can't really do anything with touch other than, like, respond to stuff you've already asked for with your voice.

There's a couple of things. Like, you can tap on the weather. But it just displays stuff. And, you know, the speakers could be a little bit better. They're, like, average. I maybe called them mediocre. Maybe it was a little mean, but they're me. I think it's a little too expensive at 249.

But the experience of using this thing if you are in the Google ecosystem is head and shoulders above the Echo Show. It is a great kitchen TV. It is very easy to watch video on it. You, like, maps are great on it. Like, all the Google stuff is incredibly good. And there's not much non-Google stuff on it, actually. You can get web snippets, but I don't know, it's interesting.

What's a day in the life? Like, I imagine this, this is a kitchen? This is something you put in your kitchen? Yeah, you definitely put this in your kitchen. With the Lenovo Smart Display. Like, how often would you actually engage with it? I mean, about as often as you would engage with, with like a Google Home or an Echo,

except that the addition of the screen means you get a couple more things that you wouldn't do. It's like the news briefings show video if you have video in your news briefings. If you're a YouTube TV subscriber, you're just like YouTube, asking it to pull up a YouTube video is great, and then just having a video playing in the kitchen

in the background really cook is really nice. You know, it doesn't give you beyond those two things like that much more than a speaker without a display. You get video, you get timers that stay on the screen. The only and I can argue about this, but the echo show is garbage because the timers disappear off to screen too quickly. And you, I don't know, like the recipe thing is neat, although it sort of depends on like how much do you want to use, you know, recipes at Google Fun on the web and study your own. But it is like a really clever workflow.

I'm actually most curious, it redesigns the web pages when you call up a recipe. Yep. I don't, I got feels, right? This is breaking the web. The websites give them permission to do it. They form a partnership. They're not just scraping rando recipe sites. Yeah, Neelizerblog.com.

Where's my ad revenue? You've got to adapt to the API. You've got to get some structured data in there. I just think it's wild that they form the deals. I'm sure they're getting their recipes right. But it's one of those things where you take the web and then you build something next to it that's much more like a CD-ROM. You know, like it's just a weird customer. some thing on the side.

I think the thing is, the web is a mouse interface. And I feel like, think about web browsing on your phone. Yeah. It's not as good. Yeah. Yeah.

You can't have regular, you can have Sir Taron Bernersley's original web vision on every device. Which actually would have worked because it was just a list of like hypertext links. That's true. Does this thing have a browser deeter? No, I was unable to.

I don't want to say definitively no, because it's clearly built off browser tech and it may be, possible if you really try to get a web page to pop up. But I was unable to do it. If you ask it a question, it shows and plays like the snippet of text. And you can tap a button or ask for more information on whatever. You call up a map. You can like scroll through pictures. It all feels like an app interface for Google's like stuff and the way that Google sees the web. You basically look when you go to like Google search on your computer on your phone and like there's all like the fancy results instead of, you know, the 10 blue links at the bottom,

that I rapidly scroll past. Take all the fancy results you get from Google and then make a kitchen device out of it. And that's what this thing is. Here's an interesting thing. I assumed that I just had weird Wi-Fi problems because I couldn't get cast to work from Netflix. It turns out that this is a cast display. You can cast video to it, but you cannot cast everything to it. Netflix will not cast to this display.

What? Oh. Correct. It's not clear in my YouTube video because I successfully cast it. Hulu, and I sort of used it in the same sentence as Netflix, and it was a bad edit in the video, so if you're going to comment on that, I'm sorry, I screwed that up. It's right in the article, and I link to Android Central, which has the list of services that aren't supported by cast on this thing, but it's real dumb that Netflix can cast

some stuff, but not other stuff. No, because cast is like a weird protocol, right? So when you cast it, it's built on this thing called Dial, which is like another thing, and the first version of Chromecast was dial, but the handler for everything was Chrome. So, Cass sends the URL, it sends the off token. The device says, oh, I have a handler for that. I will pass this URL and it's authentic into the handler.

And then for the Chromecast, in particular, the handler was always Chrome. So it could do any web video. Okay. But I bet this thing... So you're saying that this thing doesn't have a handler for playing Netflix video. Right.

And that leads me to believe it doesn't have the full Chrome underneath it because they'd have to do Chrome. Right. So Netflix needs to build. a handler of some kind. And they probably have... It supports YouTube, YouTube TV, Facebook video, Google Play Movies, HBO, Go, or Now,

Hulu, PlayStation View, Spotify, Connect, Tablo, and Vimeo. Which is a pretty comprehensive list. Yeah, but it doesn't have Netflix on it. There's one thing that makes your list comprehensive. Yeah. Like, if they were like, we don't have Hulu and they were like, we have Netflix, we're like, yeah, Hulu come.

That'll be fine. But I bet they need to make a deal with Netflix. So I think that answers that sort of like underlying web question because it was just a Chrome cast. Like it would just work. But they need a custom handler for this. Remember a couple years ago I wrote about the internet bundle and I was afraid about smart assistants putting us in Wild Gardens and all the things, we only get to see

what the things that they made deals for. It's happening. So I watch your video and at the end of it you're like, I hate saying, okay, Google. And that's the thing that's the thing that stops me. Seriously. That's the thing I would, I'm not like, I got a bunch of echoes. They're fine. I do like the idea.

I mean, to be honest, I've been considering, like, making a full-on, like, get rid of Google move in my life. Yeah. Try to, like, get off of Google. Like, find replacements for all Google services, see if I can go a little indie. But right now, like, having Google's calendar, Google has, like, all my email accounts, like, all that kind of stuff. That seems like so much more convenient and potentially so much more valuable than what Amazon can provide. Yes. And it's really hard to move that stuff.

This is like a great segue into this interview that I did with the Google engineer. Oh. Let's do that. Wait, I had one more question about the, I wanted to call it the Echo Show. The Log Novo Smart Display. What's the app story? So people can make skills for it, basically? It supports the same, like, Google actions that are on the Google Home.

They can do some display stuff, but it's. it's pretty clear that they are not as close to where they should be and where they want to be in terms of third-party support for things. There's no app store, of course. So there's more to come. And by the time this airs with any luck, my exclusive interview with three of the engineers who worked on this product will be up on the website. Ooh. Got a day two story for you all. Look forward to that.

So you're confirming that it will be up on the website? Well, it's going to have to be now. Perfect. So Paul, you were just talking about how hard it is to leave Google. They've got your email, they've got your photos. Yes. They've got a bunch of other information as well. It's actually quite terrifying.

A tight embrace. A tight, very. But the same is true for like Twitter and Facebook. So Google and Twitter and Facebook and Microsoft, a bunch of other companies, are rolling out this new thing called the Data Transfer Project, which is super interesting. So as you probably know, Google had Google takeout for a while. Now, there's all kinds of ways to get your stuff out of Google. But you have to physically download it.

They've got to put it somewhere else. That's like hard. So there's an engineer called Brian Willard, who is working on this thing called the Data Transfer Project. Joined us. We talked about what the goals are, about competition in the space, about enabling competition. Super interesting. Check this out. We'll be back with Paul's segment.

Hey, everybody. So we have Brian Willard from Google with us. Hey, Brian. How are you? I am great. So last week you had some big news sort of in the Google world, in the tech world. We talk on the VertraS all the time about vendor lock-in, proprietary protocols. You've been working on this idea of data portability for a long time.

And last week you announced a huge new project being led by Google with some big partners to make it easier to move things around. What is that? Yeah, that's right. So we announced the data transfer project, which is a collaboration between Google, Microsoft, Facebook, and Twitter, to really make it much easier for users to be able to transfer their data on the internet. So it's an open source project so anybody can contribute. It's not just those partners.

And we're trying to build kind of the shared foundations to make it so that that vendor lock-in doesn't happen because we agree here at Google that that's not something that we want to support. And so Google's been working on this for a long time. The first version of this was called the data liberation front. It's like a renegade project inside of Google that became a real thing. And that became takeout, I think it was called for a while. And now is it migrating into this?

Is this the growth of it or is just something new? This is like the next generation. So they'll both coexist. So that's right. the takeout from the deliberation front in 2011. Brian Fitzpatrick was kind of instrumental in getting that started. It's been excited to see that kind of grow, both inside Google and then other companies kind of add that same functionality over time.

And so the data transfer project is more about doing service-to-service transfer directly, instead of kind of takeout is about downloading your data and getting a copy of it. So right now, you know, you can go to takeout and you can download everything that you've ever uploaded to Google. You can go to Facebook and export a bunch of stuff. You can go to Twitter and download all your tweets, although just delete those. You don't need those. But what you're saying is, okay, then you have them.

What are you going to do with them? This is more you can, inside of Google, you're going to be able to send it directly to Facebook. That's right. That's exactly right. And so when we look at kind of how our users are using takeout and how the Internet's going, we see a couple of trends that made us want to work on this project. And so the first one, mobile only usage is growing. So people with only a mobile device.

And then Internet usage is growing in a lot of places where connectivity is either metered or it's not so great. And so for both those cases, downloading your data isn't really a tenable solution to that problem. And so with Takeout, we added a bunch of functionality

so you could directly move the output of your export in the cloud. So you can move it to Dropbox, you can move it to Box, you can move it to One Drive. And so that kind of solves that problem, but you're still left with a zip file in the cloud. And so you have to be a little bit knowledgeable

to know how to do something useful with that. And so the data transfer project tries to solve all those problems. It transfers your data right in the cloud directly to another service provider. So you don't have to figure out how to make use of an inbox file that you downloaded on your phone.

So that's really interesting because I look at some of the exports I've downloaded over the years. And sometimes people ship you basically an entire fake website, right? Like I think the example that comes to my mind is Twitter. You download all your old tweets. You open them. You basically get a Twitter website running locally off some cash data. But now, and then you have to like do something with that. And maybe that's useful to you.

Maybe it's not. You can't really send it anywhere else easily. How are you going to manage the translation between you know, one service to another. Is that, is that the problem that you're solving? Like, you've got a bunch of data in Facebook and you want to send it to some Google service. It's obviously not a one-to-one. How do you manage that translation? Yeah, that's right. And so this gets a little bit technical, but we're breaking it down. And so you mentioned moving your Twitter data to Google. And so that's a good one to start with. And so when we think about that,

instead of thinking like, we're going to move all the data Twitter has to Google, because if you imagine the opposite of that, you can't really move your shopping history from Google to Twitter. And so instead of thinking about company-to-company-wise, we break it down into the kind of data types, which we call verticals. And so instead of talking Twitter to Google, you talk about, okay, what do we do with your photos on Twitter? What do you about your context list on Twitter? What do we do about your profile data? And so it makes the problem a little more tractable to think about these verticals. So then we need vertical, then we run into the same problem, right? And so this depends a little bit. And so for some things like photos, right, it's super easy.

You just transfer it. Everybody knows the same formats for photos. There's JPEG, there's pings, all that. But for some of it, it gets a little more complicated. So context lists, there are a number of different formats. And so it becomes just to your question, how do you transfer those? And so the way the data transfer project tackles that is by building adapters. And so Twitter will build some adapters to translate their proprietary APIs into a common format. And Google will build some adapters that translate our APIs into a common format. And so Twitter doesn't really have to know or care what Google's doing. And Google doesn't have to know or care what Twitter's doing. But these adapters translate their APIs into a common format,

and then anybody can kind of join and build adapters to their APIs and get access to this information. So this is something I know you've been working on for a long time. What keeps everybody honest in that ecosystem, right? Like Google seems, obviously you guys are pushing it forward. You evolved your project many times over the years. Besides just a sense of good feeling and camaraderie on the internet, what keeps all the players in that ecosystem honest that they're going to keep their adapters up to date. They're going to, you know, tell everybody what they're doing. They're going to make sure that their exports are clean. How do you keep that in balance? Yeah, that's a really good question. And so some of it is just being a good player. Like, I think

there's going to be a lot of different motivations for different companies. Google has been committed to kind of doing the right thing by our users around portability for a long time. And so we're intrinsically motivated to do that. I think another interesting aspect to this is that you can look at, let's say, startups. Instead of talking about big companies, let's talk about a small company. And so a small company wants to grow, they want to acquire users. Usually you need to get some information, right? They might be the next photo sharing service. They might be the next Twitter or something.

And so they need your contact list or they need your photos, right? So in the past, they had to go spend a lot of engineering effort, which is pretty precious to these startups, to go figure out how to import this data. And so now there's going to be the data transfer project where they can just plug in and get access to all this user information if the users choose to go to their service. And so there's a big net win to companies to be able to allow users to transfer into them if they're offering a new or cooler service. And so we think by tying import and export together

and kind of saying, hey, we have this ecosystem. When you build adapters, you really need to build an importer and an exporter. And so companies have a, if they're not motivated like Google, by just doing the right thing, they have kind of a business reason to be part of this ecosystem because they want to be able to build the next cool service and then make it easier for users to come use that service. And so we think by tying import and export together,

you solve some of these long-term problems. Like how do you make people spend as much effort maintaining their export capability as they do their import capability. So I want to talk to you about competition in detail. But let's set that aside for one second. When you're announcing with big players, you're announcing with Microsoft, with Facebook, with Twitter, what were those conversations like when you went to them and said, hey, we're expanding the notion of data portability, we're going to build this new thing, we want you to be involved.

Were they like, yes, we've been waiting for this? Do you have to cajole them? What was the shape of those conversations? I'll say it's varied. And I want name names, but I think some players were fairly excited. And some, it took a little convincing. But one of the cool things about how the data transfer project is architected is we weren't trying to build kind of new API types or new standards.

And so it's pretty lightweight to be able to go plug in a company to this. And so one of the things we had a lot of success with is because a lot of these companies have open APIs for some of their data. We went and built adapters for them to kind of demonstrate to them kind of the power of the system because it's kind of hard to talk about abstractly. And so we went and used their open APIs, got developer keys from them, and kind of built demos to show them what this would enable and kind of how cool this would be for users. And once we had, we were past the word stage and started having some code that was working, I think the conversations got a lot better. And everybody was pretty excited to be involved. That's cool.

I mean, so walking through how it will work. I know we can't, it's a, you know, it's a radio show. We can't show people. But let's try. So you've got your mobile phone. You download a new app, you know, Neelize Photoshop. app, and there's a button that pops up that says, do you want to import your Facebook photos? Is that the interface you expect most people to see? Or are you going to go to

data transferproject.com and manage it all from a central dashboard? No, the first scenario is right. So I think kind of the open source project supports a lot of different use cases so it could support both of those. We think that for a lot of reasons, that first one, where you go to, you find your new app and you say, okay, I want to import my data and it shows you a whole bunch of places you can import your data from is the solution we're going for. And so you'd pick, let's say you wanted to pull your Google photos. You'd select Google from that list. You'd be presented with a Google OAuth screen where we use Google's standard kind of login mechanism in account security to make sure you're really the right

person. You'd complete that OAuth flow, and then you'd transfer the photos. And then one of the cool things about this project is we're trying to set it up with security and privacy first. And so the way the framework works is as soon as it's done doing that transfer, it'll then destroy those auth credentials. And so we won't have like long lasting connections to your data. This is kind of a, the user wants to do this, they do it one time, it's done. If they want to do it again, they can, but trying to do nothing tricky with users, just have them really make informed choices with where they want their data to go and then make that super easy for them. And this isn't, obviously, you work at Google. We're talking about Google Photos, but this would be the data transfer is

initiated by the app you're using, right? There's no central cloud service managing this. That's correct. There's no central cloud service managing it. You can initiate it from the new service like we were talking about, or we envision, and this isn't built yet, but being able to be into Google product and be like, I don't want my photos here anymore or I want to go try this new service and initiating it from kind of, we call it the source service, but the place the data is already at because maybe that's where you already have the relationship from. And so you can do it from either way. And so this brings to competition, right? And I think this is probably where you have a number of stories from the past because most companies, I include Google

in this. I know you're saying Google has good intentions, but Google's still a company. They want you to stay in the ecosystem, right? So what is the incentive to get everybody to participate? beyond, you know, making the playing field more level for new competitors. If you are running Google Photos, which is very good, does the PM of Google Photos really want to build an export function to send people to Flickr, RIP? Or is there some other value to that where you're saying, well, if we build the export function, it's more likely that other people will participate and you're the best product. So your import function will be better served. Yeah. So I think when you look holistically at Google, and you can go back to Eric Schmidt, kind of when we started the deliberation front, he's kind of set the culture down. and it's maintained. And we want to win because we have the best products, not because we lock users in.

And so sometimes doing what's right for the user as a company is more important than any short-sighted product-specific decision that might get made. And so I think we've had that history and have that culture. And so we're very excited about letting users make that choice. And we really, really want to win by having the best product. And I think having tools like this is one of the things that incentivizes products to make sure that they keep innovating and don't get stagnant. If you have a walled garden, you can get stagnant because you users don't have a choice. But if your users do have a choice, it kind of forces you to keep improving that product to make users be excited about interacting with it. And so I think tools like this are really beneficial to making sure that Google keeps innovating and has the best products.

So you brought up Walgarten. So I got to ask, have you talked to Apple about, you know, ICloud and imported export and data transfer to I cloud? We've had conversations with a lot of people, and I don't want to get into kind of the internal things where they're ongoing. So I wouldn't be surprised to see a lot of new people join this project now that there's a lot more publicity around it. But it's hard to predict exactly who and or exactly when. How long you've been working on this? I've been working at portability here at Google for six years. We've been working on kind of versions of this for maybe four or five years.

Had some failures. But we've been working on the Data Transfer Project for a little over a year now, a year and a half. Tell me about some of these failures. Where has this gone wrong? So an interesting feature that we built that you can still use so you can go see it is we thought we could do this with structured downloads. So instead of doing the service-to-service thing, we thought, and this is three or four years ago, wouldn't it be cool to be able to just instantly upload

these structured exports from the other companies to drive? And so we started building going down this road. And so if you go to drive today and you have a zip file and drive, you can see the contents of that zip file. So we'll unpack it for you and kind of show it to you. And so that was our first attempt to do this. And we built about a bunch of infrastructure for how to par zips and then started to parse data out of the zips

and decided that that wasn't a road that would go forward. We would have to know too much about other companies' data formats. And when they change those data formats, it would be on us to go update our import logic to figure out how to make sense of that. And so it's interesting that you can kind of still see the artifacts of that in the drive UI today. But we learned that that wasn't going to be a way to go. And so that's why one of the things that informed our kind of design decisions around the data transfer project is that Google kind of only has to care about Google APIs. And Facebook only has to care about Facebook APIs.

And when either of us changes them, it's only on our company to make those changes, which I think is a much better dynamic than putting it, that burden on the importer to go figure out every time you make a change. Yeah, it seems like this is also only truly enabled by the two things you're talking about. One, which is the overwhelming prevalence of mobile, a little bit of bandwidth constraint, but by how much work is now moving to the cloud in general and the user expectation that all this will happen in the cloud versus on their local machine. Are you thinking about user expectation as you make these choices, or are you just trying to make something, anything happen? No, we're definitely thinking about user expectations and kind of user use cases.

And so we're mostly focused on the cloud stuff, but an interesting use case is you can pick your favorite encrypted mail provider that that does end-end encryption. And so a product like this doesn't really work for that use case because the messages are always encrypted. And you definitely don't want to give your key to us to do the encryption and then re-import it somewhere else. And so an interesting approach to this kind of thinking about user cases, is for scenarios like that you can download the data transfer project and run it locally and kind of do that encryption so it still maintains end-end encryption on your machine and then re-loaded it. a new encrypted form to your new provider. And so when we think about portability, there's a lot of different use cases. One is just making it easy for the mass markets because kind of downloading this code and running it via Docker or compiling it is probably out of reach for a lot of people.

But there is a subset of people that that is appealing to and kind of knowing that they're running exactly the code that they see at GitHub is available to them, which I think is really promising for this approach, that it kind of suits a lot of different use cases. So again, I was looking at your Twitter. And obviously, big companies are the news all over the place. and even you have been tweeting about how data portability will enable more competition. So if you're fed up with Facebook, but there's a huge network effect or you're fed up with Gmail, but there's a huge network effect of how Gmail apps and services, you can go and

that's better for Google's team because they have to build a better product. Do you think that's actually going to play out in this sort of startup ecosystem now, or is this really it's going to get slightly easier to move from Gmail to hotmail? That's a really good question, right? When I look at the ecosystem, I think there's a lot of, lot of work to do on some of the smaller distributed web stuff. And a little bit of it, it's, I shouldn't make the call, right? Like, we want to provide these choices to users.

And if people can build the products that they find compelling, we should make it really easy for them to move to them. But I think it's really hard to predict. And I'm probably the wrong person to predict, like, whether that succeeds or not. But, like, I'm really passionate about making sure the users have the choice. And we give them the ability to succeed if they can. So right now, obviously this is basically just a GitHub project. It's not shipping, right?

That's correct. Today it's not shipping. We really hope to kind of build it into the Google UIs that we give users the ability to do that in the future. We can't really control where other people put it in their UIs, but I really hope to see it cropping up there. So it's not shipping. You've got to work into Google.

You've got these deals with your partners. Are you committed to, you know, across Google? It's going to be not just Gmail and photos, but YouTube as well. And are they committing to it's not just, you know, Facebook's blue app, but Instagram and WhatsApp as well? Are you just sort of, they're signed on, we're going to see how the ecosystem develops. Yeah, we're all signed on to the principles. And as with all forward-looking statements, right?

Nobody wants to commit to things in stone. So I don't want to talk about any specific product. But I think my plan is definitely to keep expanding this to more and more data types to make it more useful. But, yeah, I would hesitate to give dates on when things will make it into the product. Because a lot of that depends on kind of like the user feedback we get. If people are really cramming for a data type or there's a really cool new place to take that data, I think we'll prioritize those. But we wouldn't want to share the internal roadmap too much because it's likely to change. I'm just curious.

What's sort of the most interesting problem you've had to solve is you've been working on this for six years at a company like Google? Or the hardest problem, rather. Yeah, the hardest problem. I think it's like just the number of different data types and kind of there's not one solution that fits them all. So transferring your contacts is different than transferring your mail, which is different than transferring your photos. And so kind of trying to come up with solutions that allow the flexibility to meet all the different data types. and also have a little bit of foresight, because you can imagine, and we hope, the data transfer project is used for data-leg loyalty cards, where Google is not in the market.

And so we don't want to, we want to leave it open enough so that the players that are in each of the markets that we're enabling get to kind of decide how it's run. But hopefully we can provide the tools to make it a lot easier for them. And so kind of making sure we have the right solutions to the right spaces is, I think, an important issue. Are you running into any of the sort of larger policy concerns as you architect the tool? Like, obviously there's the GDPR in Europe. Obviously, there's a huge amount of scrutiny in this country.

Are you building around that stuff? Are you saying, you know, this is the user's data. We're letting them do what they want with their own data. They've clearly already opted in because they're making the choice, and we don't have to worry about that. Yeah, I think we're starting from the principle of approach, as we've had a long history at Google about letting users take their own data. I think we are trying to be mindful about kind of these external events that are going on, right? And so how the slider moves between how easy it is to transfer your data and how, and how much we inform users about exactly what's happening is, I think, an interesting one where we'll

continue to kind of move how our design decisions are based on external events, as I think that's an interesting conversations that's having on the web. Yeah, I mean, to me, just honestly, it seems like if you're more in charge of your data and you know that you're taking the action on it this directly, then that's probably a good thing. It at least fulfills the spirit of what everyone wants you to do. But on the flip side, it seems like a bunch of new apps that make it super. easy to import your data that you may not have a long relationship with. It's probably great in the spirit of competition, but now suddenly it's that much easier to send your data everywhere

in a way, especially if it's happening in the cloud and you're not seeing the sort of tangibility of the data transfer happen on your machine. It seems a little dicey there too. Yeah, that's exactly right. And this goes back to kind of my earlier point about there's not being like one size, fits all solutions. And so kind of the way we're taking the approach on the open source project is that we're going to build the tools, but we're not actually doing the transfers because each individual provider kind of knows their data types, knows their users best,

and should be get to make the final decision on, you know, how strict do we want to be with how we vet these companies, right? Google shouldn't be the person deciding that for Facebook. Facebook shouldn't be this person deciding that for Google. You should know your users. An interesting example of that is that Google, we have kind of a product called the Advanced Protection,

which adds some extra security guards to your account. And so because we have these separate set of users that are in this space, it probably makes sense to offer different sets of protections to them that maybe make it a little bit harder to transfer their data than it does to an end user or a user without these additional security features that they're opting into. And so we have that other companies have that same kind of thing. And so it's really important that companies are able to make nuanced decisions about this because there isn't one answer that fits every situation, every company, every data type, right? The protections that you use for your medical records are probably different than the protections that you should use for your URLs that you shortened, let's say. Yeah, just thinking about it now, it seems it's your data.

You should be able to make whatever mistakes you want with it. But there's like a user education process about, hey, we've never seen this app before and it's asking for all of your Google photos. Maybe you should think twice. Like, are you trying to build something like that in where it's like, you know, some scammer is going to make an app that just starts sucking in data? And, you know, at the end of the day, it's the user's data, but they made a mistake and Google might have some obligation to call it out.

No, that's exactly right. And we're trying to debate that, like, how much we protect users themselves. And kind of what we keep coming back for, we don't have set answers for that yet. I think there's a lot of different moving pieces there. But kind of one of the fundamental things we think is going to play a critical role in this is transparency, right? Like, we need to be very transparent, right? How many users have moved to this service? How long has the service been around?

And really highlight these signals that will help users make good choices. And then also have some backstop where we know there are some legit bad actors out there. And so really highlighting those and using the data sources that we have from a lot of other places, to feed into these kind of reputational decisions. But you're right at the end of the day, it is the user's data. And so they should get the final choice, but we need to make sure that we're giving them the information to make that informed a choice. I mean, to me that the story of the Internet in the past two years is the sudden recognition

that every tool has to be constructed with bad actors in mind. Whereas, you know, I'm a certain six years ago when you started, the hope was that users was they'd have the ability to download their data. That would be great. And now there's an entire other class of participants that it seems like it constrains everybody's ability in a pretty serious way. Yeah, I think that's true. I think the other side of that, though, is when I talk to my friends that work in payments, right, they've been having these problems with risk assessments and kind of thinking about bad actors for a long time. And so I think

learning some lessons from them will also be interesting. From back in the day, it's been interesting to, we've had these conversations for a long time. So it used to be people inside of Google brought up the argument that the data was much more secure on Google server than it was on your home machine, right? Because it's much more likely that you have a virus that was going to rip up all this data from your machine. I think this is just an evolution of that conversation that now we have to worry about the security of other services and not just the security of a machine. But it feels more like an evolution than kind of a revolution. Yeah, that makes sense. So what is next for this tool? What's the next, it's obviously been announced. People are signing up. What's the next big step

that you foresee? Yeah, so two things. One, in terms of the open space world, we really want to make sure that we get a lot more partners involved, right? The partners that we had at launch day, we're all pretty big. And we want to make sure that we have a representative sample of the the companies they're going to be using it. So continuing our outreach and making sure lots of different stakeholders are involved is kind of goal number one for DTP. On the Google side, we've been working to kind of integrate into the product, make it in the UI, so users can actually use this functionality. And that involves tackling a lot of those questions that you just answered around, you know, how do we provide transparency? How do we make sure that

we're informing the users the right way? But we really are dedicated to trying out making this a product as much as we can and not just having it be an open-source project that results in some good publicity for a second, but really turning it into a useful tool. for users. So are you going to manage the adapters or is the data transfer tool going to manage the sort of the adapter store or is that all sort of up to the companies? I think there's probably a two-pronged approach for that. One, because we're in such early days, we don't have a defined governance model established and I think it's going to be important to have kind of that variety of stakeholders I talked about before, involved before we kind of solidify what's the right

decision. So it's not just these big companies making the decision, but we make decisions that are right for everybody. The other side of that is I imagine there's a world where there some vetting that happens in kind of if you can be an official adapter that's in the GitHub, right? We need to have some controls over who can contribute stuff to the GitHub just so you're not, you know, contributing viruses left and right. But the other side of that is we're really wanted to do this as an open source project because we want people to go fork it. And so if they want to go build their own adapter, but for every reason, DTP isn't a good fit is the place to put that code, they can go fork it and build their own adapter and kind of get this functionality off on

their own without there being any central gatekeeper. Yeah, I think that the gatekeeping question and the sort of user trust question are they're inextricable, right? And the question, I mean, it's, I'm very excited to see where you go with it because that question is, it gets at the heart of it. Yeah. And we've been wrestling with it and don't have finalized answers yet. But I think a lot of it goes around transparency and making sure that we make all these, this data as transparent to the user as possible.

Because the users are fairly savvy. And as long as you present the data in a right way, I have faith that for the most part, they'll make the right decision. Awesome. Well, Brian, thank you so much for joining our show. I know we went a little bit over, but I appreciate it. very exciting project. We're eager to see what happens next. Brian Willard, thank you so much. No problem. Thanks for having me. Big fan of the podcast.

Okay, Paul. You. Every week, my man. You do the thing. Yeah, it's called, what is it like to be a dolphin? The best. Yeah. I bet it's great being a dolphin. That I really think about it. Don't just write down the name of my segment. It would be great to be a dolphin. Except for the whole tuna net thing. I'm sure that's like, I mean, you know, sometimes people like, sometimes people get hurt. It's not all roses being a human.

Most dolphins are having a great time. But if you want to know what it's like to be a dolphin, you can check out the Dolphin View headset, which is, what's exciting about this is that LIDAR is so inexpensive now. And it's like, it's bonkers how it's changed over like the past, I would say, three years. So for about $100, you can put a LiDAR sensor on the front of these pair of glasses. This is just basically an open source project. And then you wire that up to a chip that has some very simple software on it. And then it sends like signals to like bone conduction headphones. So basically you're not, it's not like echolocation, but in a sense you are having a sound experience for what you are quote unquote seeing around.

you. So the guy says like... Technologically enabled synesthesia. Yes. Ooh. Yeah. And so you can close your eyes and he says like you can detect if like a door is open or closed, that kind of stuff.

So I don't know. I always love this kind of stuff like, you know, there's like sensors. You can put like an interface on your tongue so that you can like sense stuff. There's people have like sensors on stuff on their backs. There's like people who have augmented like senses of direction. I love this kind of sensor augmentation kind of stuff. So this is... Can you get it?

Is this available? It's a project. You go on GitHub, you follow the instructions for about $100. Yeah. Which is crazy. Are you going to build this thing? No. Because I assigned Deeter a story just now.

I have the power, my friend. I think you should spend a week as a dolphin. It's basically... A dolphin week? Dolphin week. It's our much more confused answer to shark week. But instead of like providing shark... or dolphin programming.

It's just you being a dolphin for one week. Dolphin Week at the verge. We'll work on that. All right. It's going to be really bad with him on the Vergecast during Dolphin Week because he'll only be able to communicate and clicks and whistles. That's right. I mean, at this point with the Verge cast,

potentially an improvement. There's a bunch of echo stuff that I did want to talk about, but that was just a good segue before. New Echo Dot is revealed and leaked photos. They're adding an equalizer to the speakers. Alexa, Cass. Alexa cast. That seems a mess.

And then they added some interface stuff to the Echo Show. I'm the only person in the world who likes the Echo Show. Maybe I'll try it out. But I just like it because I can see what it's doing. It's not a computer the way that I think the smart display fashions itself as a computer. So Alexa cast. It only works inside Alexa music is my understanding. But it basically works in the same way that other cast things do where you're listening

to it on your phone and you're like, no, I want to send a speaker and then the speaker starts playing it instead, right? So a little bit like Spotify can act, a little bit like Google Cast. Nothing at all like AirPlay, I suppose. Why wouldn't they just support Google Cast? This actually makes sense to me. Because they are in a lot of advantage here.

A beef. There's a serious Amazon Google Beef. But there's no, there's no, it's, it's, okay, they put out AlexaCast. It only works on Amazon music. Right. Which everybody uses. My favorite music service. So it's like they don't have a huge,

they're not like taking their huge user base in saying you can no longer use these speakers. They're saying to their huge user base, you can, if they just use Google Cast, then Spotify people would be able to just like use their Alexa devices. Right. If they just use Google Cast, you could shoot Netflix to a fire TV. Like, Cast is like open-ish, right, in that Google way.

Yeah, why does everybody hate Cast? Like, it's actually confusing. Is it not as open as we think? And like, does it support Google too much? Or is it just not that good, but Google makes it good with like special Google magic, fairy dust. It is called Google Cast. Yeah.

Right, but the underlying, it's this thing called Dial, which is an underlying open standard. Yeah. So like when the Chromecast first. Fundamentally, all it is is like, hey, play this thing at this URL. Tell that thing to play the thing at this URL. That's at its core. That's all cast is. It's like, hey, you hear me?

Play that thing. And then it's like, cool. I'll play that thing. That's it. Here's a URL. And everyone has to make their own implementation of it. And then Apple's is extra special because it sometimes does that, but most of the time it doesn't. Just like straight.

Apple's AirPlay actually streams from the device sometimes. Yeah. So Dial was developed by Netflix and YouTube. Okay. And I think Google cast is something different now. So I was looking. I think they've embraced and extended. Hmm.

Yeah. Yeah. Yeah. So, yeah, what's the solution? Is this, I mean, does Amazon just have to do this because of Google? Google has so much power and Google Cast is their thing. Well, probably, but like, imagine if you're Spotify, right? And you're like, you're the music service.

You already have the AirPlay button. No, it's all, Spotify just like, you know, they abstract all of the protocols, of course. But you're Spotify's developer and you've got to support AirPlay, AirPlay 2, right. Bluetooth, Google Cast, Spotify Connect. Alexa cast Roku Connect We had Anthony Wood

On the show last week He's talking about Roku Connect Like suddenly you're Literally the fragmentation of services About you're going to have a bunch of stuff in your house You want your apps to talk to it Yeah There's something mistaken here

Like so We're making some kind of mistake here When Spotify has to spend X more hours Of engineering time supporting one more Cast protocol as opposed to making its app better. Because they're all just there.

You should just use them. Here's the thing. I tweeted some of this effect. I was like, if Wi-Fi were invented today, every company that makes these different casting standards would have made a different version of proprietary, different version of Wi-Fi. And you'll never guess who responded.

The former CEO of Sonos, John McFarlane, who writes, they all benefit from the Internet and Wi-Fi standards and cannot but do proprietary in return. They all suck, too. How is their behavior different from stupid consumer electronics who worked to make universal remotes impossible? Wow. There's just every, it's like every trigger for Vergecast discussions.

You've got universal remotes. You've got IR blasters. You've got Wi-Fi standards. You've got Sonos. It's all there. I think these casting standards are, they're like sort of a lever of power for these companies. You know what's funny about that is that Sonos ran a proprietary white. If I riff.

Like, that's Sonosnet. Like, that's what that is. It's a proprietary version of A2 to 11B. Now they run a Wi-Fi a little bit. But you can still buy, like, a Sonos boost and then have a second proprietary. It's true. It's a fact. I don't know what the moat is for proprietary casting.

Like, it's not like Netflix is going to give up and just pick one. Hmm. Right? I mean, the moat is you buy speakers that are compatible with the phone that you have. or if you happen to have speakers that work with your phone, you're less likely to switch to another phone because it'll no longer work with your speakers. Yeah, but how does, okay, I'll give you all of that.

Now put Alexa Cast in that framework. If you are the one Amazon music customer in the world who's like, thank God Alexa Cast is here because now I'm only going to be in the Amazon ecosystem. Alexa Cass is just straight up spite. It's just straight up spite for Google. It's leverage. It's there so that when they're negotiating with Google

to allow the Lenovo smart display in the Amazon store, which guess what? It's not there. That they'll be like, tell you what, we'll switch away from a Luxacast if, you know, give us something on this deal here. Microsoft needs one of these. Microsoft just uses Mirrorcast. Which is, by the way, another standard in this mix.

Oh, that is an actual, but that's streaming, that's not just a URL. Yeah. Because I was just thinking, because, like, obviously it's bonkers to think of you buy a speaker that's specifically designed for your phone or computer. But like if you...

Is it bonkers? Because I have a home pod to sell you. I know, I know. No, I was thinking like, you know, I grew up very close to a Best Buy, like walking distance from a Best Buy. So I've been to Best Buy very many times in Circuit City. And was it Future Shop? What was that other one? Future Store.

Whatever. Future store. This explains a lot, actually. I grew up going to Radio Shack, so you're, you know. I'd go to Best Buy and hang out at like the cell. cell phone kiosk and try to like upsell customers on like Bluetooth. Anyways, if you think about what Best Buy's layout was was there's a computer section, there was a phone section, there was like a home audio section, there's a TV.

Now if you go to Best Buy, it's there's the Samsung section, there's the Microsoft section, there's the Apple section, you know, there's the it's. Well, yeah, Best Buy's turning into, so two things. Best Buy is turning into these collections of stores and their growth, like they're succeeding, not necessarily because they're selling stuff, but because they sell tech support. Yeah. Because there's a really good profile. It's not just the geek squad.

They have a whole tech support division that just like helps you out. I'm sure encourages you to buy any Best Buy things. But that's their growth because everything's so complicated. American teens need to get on the job because Best Buy is turning that into money. If somebody knows of an open version, open alternative to Google Cast, please let me. I think it's... This can't be that hard. Figure it out.

All right. Real quick, we got to end this. Dieter, do you want to do your disclosure? My wife works for Oculus, which is a division of Facebook. Facebook is having a very bad time. Yes. It's, you know, we don't usually talk about Facebook on the show, but I think it's important. After basically a year of scandals now, they posted their quarterly results.

The actual numbers in the quarterly results were not that bad. They posted 49% year-over-year revenue growth. and they only missed their earnings targets by like 0.8%. Like literally cents. It was like they were 8 cents off. Like literally that was not very much. And the big headline was that user growth is slowing. User growth in North America and the EU is slowing because they have run out of people.

Like that's the story. And their margin guidance, they're going from 44% problem. margin is somewhat lower because they're investing in privacy, security, getting bad actors. So they're investing in all this stuff. Their stock took a nosedive, largest one-day drop for any company in history today off their market cap, like $110 billion, like crazy numbers. The one thing that caught my eye was that they said the revenue from ads and Instagram stories and stories on Facebook, those ads are cheaper than ads in the feed, but users are moving to stories. So they're in this, like, bad feedback loop. I don't think that's the entire reason for the drive.

Wait, explain that one more time. So, you know, they're putting lots and lots of ads in Instagram stories now. Right. Those ads are cheaper than ads in the Instagram feed. They're cheaper than ads in the Facebook feed. But the users are moving to stories. The more people, stories are a runaway success story from a user perspective. Right.

But monetizing that behavior is less lucrative than monetizing the feed. So that's, like, one forward guidance. They don't know if they're going to make it out. They roll out a new metric. They're still showing you each individual app, but now they're showing the family of apps because they want to show consistent growth, and the family will keep growing even at Facebook stalls. So just like a lot of bad stuff coming at Facebook, and the stock market killed them for it today. We don't really talk about stock prices on the show.

We don't really talk about stock prices on The Verge. But I think it's important to point out only because the question has been, will Facebook have consequences for its behavior, for its lack of behavior, for whatever? for whatever. And it seemed, today is the day where the consequences have shown up. And like, I, you know, obviously you should read Casey's newsletter of the interface. We're now publishing it in the morning as we're heading to midterms. You know, Casey writes a whole newsletter about social media and democracy. So we're putting on the site as we get closer to midterms as important. So there's a lot more color there and a lot more, you know, Casey has a lot of insights to read that. But I just wanted to bring it up on the Vergecast and talk about it for two minutes because

we're at the point now where the things that are happening to Twitter and Facebook and YouTube and the other big social networks, it's very clear that they won't just have like a public perception impact. They're going to start to have corporate financial impacts as well. And I think if anything is going to shock these companies into accountability, it's literally the vertical line on Facebook's stock price chart today. It's like clipping along, like, you know, stock price up and down. And then it's like, vertical line.

Hey, what would you do if you were Facebook? You know, I read, um... Because everybody's mad at them, both sides. I read a tweet storm from this dude, Ryan Lawler, who used to be at TechCrunch, you know, I think it works for Samsung. But he's like, Facebook has a huge problem where they don't make anything. They can't, it's not like Samsung where like the phone blew up and they're like, the next one will be better. Forget the old one. It's just Facebook.

They can't roll out a new Facebook. And even if they did, it's still just Facebook. So they can't, they're not going to benefit from the cycle of improvement or like a version numbers or whatever. They just have to try to fix that thing. And, like, is there only downside there? Like, are they just old? Are they ever going to be cool again? I don't know the answer.

I feel like this was a question, like, years and years ago we talked about maybe back in gadget. It was, like, Facebook permanent. Like, MySpace died. Is Facebook unkillable? And I think it's interesting. It's like it seemed like I really don't know how the stock market works at all. Yeah.

But it felt like this was like some. sort of sign that Facebook might not be unkillable? I mean, we'll see, there's still Facebook. There's nothing out there that can compete with it, right? There's nothing out there that competes with Instagram in that way. None of this is rubbing off on Instagram. So it's a lot. Facebook's CFO on the call yesterday said, our total revenue growth rates will

continue to distillate in the second half of 2018, and we expect our growth rates decline by high single digit percentages from prior quarters in both Q3 and Q4. So they're telling the market, like, this is slowing down. The question is whether there's lots of businesses that don't grow with the rates of Facebook, right? They just like mint money. If that's going to be their business, that's fine, but that is not the story they've told people. Anyhow, not to end on a dour Facebook note, but it's important. It's out there.

Read the interface with Casey Newton. That's what I do. That's what you should do. And also last week you recommended that I listened to the Keraswisher interview with Mark Zuckerberg. It's very good. On recode. That was... Very good.

Intense. I wish Zuckerberg was more of a leader. It's like a... You get the sense that he doesn't want to make the decision. And like, this is a moment for him to make a lot of decisions about what he wants to do. Anyway, if you haven't listened to it, go listen to it. Speaking of listening to stuff, you should listen to Recode decode with Karras-Wisher. You should listen to Recode Media with Peter Kafka.

That's very good. Vox.com is doing a show on Netflix called Explained. the new one is about weed narrated by Kevin Smith it's so good it's so good and the cryptocurrency one is narrated by

Kristen Slater which I just have to have to put that one back out there you can also listen to speaking in Casey Newton oh wait no I'm sorry one more the cricket one is narrated by Asif Manvi it's also a very good episode

I understand cricket now do you I do I watch it too but I had to watch there's like an animated segment where they explain the rules and I watched it twice

not gonna lie All of this very good. You can also speak in Casey Newton, listen to Converge with Casey Newton. That show's great. Both seasons, Why you Push That Button are sitting there waiting for you to binge them. We had a meeting about season three, watch you push that button. Caitlin was in the office, Ashley, Andrew, our great producer. That was a fun meeting.

That show's coming back very soon. And you can obviously watch us on YouTube and go on Instagram, skip the ads and watch our stuff on the stories. All right. You talk to us. I'm Reckless on Twitter. Paul's Future Paul, Dieter's Backlund. That's the show. We'll see you next week. Thank you very much.

Rock and roll. Paul. promo code.