[email protected]
Episode 69 June 13, 2017 | 41min

What We Chat About When We Chat About Chat

Our co-founders chat about chat, bots, privacy, and the internet of things.

Show Notes

This week Paul Ford and Rich Ziade embark on a wide-ranging conversation about innovation and change in tech — and its impact on our daily lives. Topics covered include connected devices, machine learning, the future of medical apps, technologies and superpowers, and whether it would be fun to go to a bar with Siri. (Spoiler: it wouldn’t.)

[Intro music]

Paul Ford Hi! You’re listening to Track Changes, the podcast of Postlight, a digital products studio at 101 5th Avenue in New York City. My name is Paul Ford, I’m the co-founder of Postlight [music fades out].

Rich Ziade And I’m Rich Ziade, the other co-founder of Postlight.

PF And today we thought we would talk about—

RZ Well! Tell me what Postlight is, Paul.

PF Ah thanks for catching me out there. I—I always forget.

RZ Really quickly. We don’t wanna bore everyone.

PF We build web things and internet thing and uh platforms, things that you can build on [yes]. So you come to us and you say, “I wanna build an app,” and we go, “Alright, well let’s talk about that.” And then we figure out that you really actually need to build something a little different.

RZ Yeah! We strategically put it together for you. We design, and we think about user experience, and then we go ahead and build the damn thing [that’s right] which is unusual to have all of those talents in one place.

PF Yeah you often will never have a single PowerPoint during your entire relationship with us.

RZ That’s true. We won’t do that.

PF And yet it’s still a good relationship where all parties are happy.

RZ Yup.

PF Yeah.

RZ It’s all alive! It’s a living document, the entire project.

PF Code! Code! Is alive.

RZ Design is alive, too. You shortchange design when you explain us, Paul.

PF That’s probably very true. I need to stop doing that.

[1:25]

RZ We’ve got some very talented designers.

PF Alright, enough with this very—

RZ Enough about us!

PF Yeah—ambiguous advertisement [yes!]. Let’s—we were talking, at lunch, about voice . . . Alexa, Siri, all that stuff.

RZ Yes.

PF And bots.

RZ Yes.

PF And I thought it’d be interesting to kind of bring everybody into the crazy world of just us talking about products.

RZ Yes.

PF And take them on our journey.

RZ Products beyond point and click.

PF Yeah! Just like—

RZ I think that’s—that’s what we wanna explore here today.

PF Because the reality is that for all we try to make things concrete, and put them in people’s hands, and make them clickable. We live in a world of complete abstraction in which everything is fungible and it’s very confusing.

RZ Yes.

PF So I wanna bring people into that world.

RZ Ok. Let’s do it.

PF Alright. Do you use a lot of voice stuff?

RZ No. Not a lot. I wouldn’t say a lot. I don’t use Siri—it’s turned on on my phone for some reason. Every so often this British guy says, [in British accent]: “What’s wrong, Rich?”

PF You know, can we—

[2:23]

RZ Something like that.

PF Actually it’s been many years of Siri, and it’s not good.

RZ I just uh I never latched onto it. I wonder—I wonder if there are people who like can’t live without it. I’ve not met anyone.

PF I’m sure—I’m sure there are. I have five—Maybe if we drove more we’d have more of a relationship with it. I have five-year-old kids and they’re bored of Siri.

RZ Mmm.

PF They—they wanted to talk to it for awhile and now they’re like, “Eh—”

RZ Well, it’s very safe, right? I mean if you had to meet Siri for drinks, right? At that speak easy a few blocks from work. That is gonna be a really lame hour and a half.

PF That’s right.

RZ Really lame.

PF It’s terrible.

RZ So tell me what’s really going on, man. After the second drink, tell me what’s going on. And he just wants to tell you tomorrow’s forecast.

PF Yeah, some Siri is just um—

RZ It’s a boring human being!

PF “Sorry, Rich, I’m afraid I didn’t get that!”

RZ Right. You know the fact that Siri is tied to a software license agreement pretty much sums up the Siris of the world [yeah], and the Alexas of the world. They cannot possibly carry any liability with any advice they give. Right?

[3:24]

PF Right.

RZ Like if they tell someone, “Uh you don’t need to wear slippers,” and then someone goes and cuts their foot, Amazon will be sued because Alexa gave bad advice. Like things like that drastically limit what they can do, right? Because they have to think about all the different sort of points of liability with this thing giving you a piece of advi—cuz it’s closer to human. You could do it in an app, right? But for some reason, because it’s talking to you, you could tell the old lady—like if the thing gets unplugged by mistake—

PF Back—back this out for me [yeah]. What are you talking about with liability? I didn’t notice—I haven’t thought about this aspect.

RZ I think what Siri can do, right? And the things it can say are quite limited because of legal concerns. For example, if I tell Siri, “Siri, I’m really down today. What do you think I should do?” And Siri actually gives advice there.

PF Like—like Siri says, “Kill yourself.”

RZ Not—not even. Or Siri says, “Hey, why don’t you take a shot of whiskey.” Or—

PF Or Siri says, like, “A nice hot bath and a razor blade.”

RZ [Laughs] Whatever.

PF Or Siri asks you what floor you work on—

RZ And then there’s actually—

PF Then suddenly, like, does the equation for mass and velocity to let you know how hard you’d hit the ground if you jumped from the building.

RZ Yeah.

PF Yeah. That’d be bad.

RZ I mean there’s actual advice. So it has to play it safe. Siri has to like play it safe [yeah], essentially. Also, it’s hard. It’s a hard problem. Like they don’t have the machine-learning/AI to walk someone through—

PF The thing is that voice recognition is pretty good. Overall: pretty good.

[4:58]

RZ I have to say—

PF It’s gotten real good.

RZ And [stammers] the—I don’t know what they call them, near field microphones [listen, listen!], they’re incredible.

PF I was around for the days of drag and dictate, this is a different world.

RZ Oh no. The way—it can hear you from across the house.

PF Yeah.

RZ And you could—you could be really low volume. It’s actually incredibly impressive.

PF So what’s fascinating is that the computer has figured out what we’re saying. It’s been able to—it can translate, you know, phonemes into text from across the room [right] but it can’t figure out the meaning. It can’t do it, right? [Right] And so like there’s still no semantic model that’s like just exactly right where it can go, “Oh I know what you want.”

RZ Yup.

PF So my kids can ask for like, “Show me a picture of a pink unicorn.”

RZ Right.

PF But I’d rather just like I don’t have a relationship with Siri. Now, do you use Alexa?

RZ We use it [stammers] I got it wired to Spotify which is convenient. You could put music on.

PF What do you say?

RZ You say, “Ech—” I use Echo cuz my little girl’s name is Alex.

PF Oh, so you had to rename it?

RZ You could name it Echo or Alexa.

PF Those are some—what choices.

[5:54]

RZ So we say, “Echo, play uh “Daniel” by Elton John.”

PF How—how often do you say that?

RZ “Daniel”?

PF Yeah.

RZ A little too often, strangely.

PF Yeah, yeah, probably pretty often.

RZ It’s a very good breakfast song, for some reason.

PF It’s great. It’s a nice song. Nice to wake up to Elton John.

RZ So the plug-ins! Like I could get it wired to my thermostat. So I can actually say, “Echo, set the thermostat to 74,” and it’ll actually turn on the heat. It’s pretty cool.

PF Ok. So [um] you’re—so we’re consumers of voice products. So you—you more than me. You’ve allowed it to control your home and life.

RZ We could go days without talking to the thing.

PF It has a relationship with your children. Ok.

RZ But it’s handy.

PF Ok.

RZ Um there was an article recently about how it’s doing a poor job of teaching kids manners . . . You don’t say please and thank you to Echo. And the kids are talkin’ to this thing and they’re thinking this is the way the world works: you just tell people to do things.

PF You know kids—you know what kids aren’t that great at, mostly? Manners.

RZ Generally speaking.

PF Yeah. I mean I’m sure that this is a real endemic problem on our society that kids [yeah] aren’t saying please and thank you to robots enough.

[6:53]

RZ Right.

PF But it’s like if I can get my kids to say thank you without reminding them [yeah], I’m pretty happy. My children have a reputation for politeness and they’re terrible at politeness.

RZ Mine too.

PF Right? Other people are like—

RZ You’re very diligent about it.

PF Yeah, other people are like, “Your kids are polite.” And 99 percent of the time they’re not. So.

RZ Well, they’ll sort this out. [Yeah I don’t—I don’t—] They’re four-year-olds and we need them—So the way you get the—it to learn other things, the Alexa, the Echo product, is you add what are called skills. Like outta the box it didn’t know how to connect with my thermostat. So you would actually auth into the thermostat app and link up the Echo with the thermostat app. So—

PF Just even saying that you know how screwed up that entire process is.

RZ Exactly. It’s a messy—it’s a messy process.

PF “Lemme auth into my thermostat app.”

RZ That’s right.

PF [Sighs.]

RZ That’s right. Now, what would be interesting—

PF “So I can talk to my thermostat.”

RZ Imagine a . . . keep the marriage together skill. Like, Echo detects that you haven’t spoken to your wife in two days.

PF Or that you’re speaking constantly.

RZ [Chuckling] Or the volume level [is very high] coming out of the kitchen is a little too high.

[8:00]

RZ And it says, “Why don’t you settle down and have a cup of tea together?” Or something.

PF “I’m calling the marriage counsellor!”

RZ [Chuckles] It just dials one up!

PF Dials the marriage—“Hi, guys!”

RZ What’s another skill that would be really—

PF Oh Alexa could do really well with you opening the fridge. “Hey, whatcha doin’?”

RZ Yeah, right, right, right.

PF You know Alexa could—

RZ “Again?”

PF “Again?”

RZ “Again, Paul?”

PF “Huh, dinner was just an hour ago.” [RZ laughs] That’s a definite Alexa skill [yeah]. Uh Alexa could—well you know Amazon has put a camera on Alexa now. And they’re gonna look at your clothes and tell you if you’re wearing—if you’re dressed.

RZ See we’re getting to a weird place. Right? We’re gettin’—“Eh! You’ve put on a couple of pounds there! I dunno if those pants are gonna work for you anymore.”

PF “Yeah, why don’t we get you some bigger pants?”

RZ [Laughs] I mean it’s—it’s funny. It’s interesting. The—technically it’s fascinating. There’ve been articles written about how they spent months just reducing, cutting a second or two off the response time [sure]. It’s really, really impressive stuff.

PF Amazon can do that. That’s what Amazon does.

RZ It’s what Amazon does and they’ve done an amazing job and everybody’s on it because they think it’s the future so Microsoft has one. I think it’s called Cortana or something. Anyway, they’re interesting. Um there—there’s a part of me—I think there’s a part of us, generally, that subconsciously draws lines around where tech is allowed to go. I—I’m always been fascinated with the way Bluetooth headsets hit a wall.

[9:32]

PF Well cuz they looked ridiculous. The people—[hesitates/stutters] things get over that wall. People who used cell phones, the first group of people who used cell phones, were terrible.

RZ They were—like the big brick? The Gordon Gekko brick?

PF Or even the—like the original Motorola StarTACs?

RZ Yeah. It’s—

PF That was cell phones.

RZ But there’s—I think Bluetooth lost. The Bluetooth headset lo—I don’t think it’s gonna turn. Like people still wear it. Like the sales guy still wears it.

PF Yeah, I know, the little like—the little tag in the ear. The problem, first of all, Bluetooth sucks.

RZ [Stammers] This is not a technical issue, Paul.

PF I wanna talk about Bluetooth for a sec.

RZ No, no. We can’t veer off into Bluetooth.

PF No, really? Cuz why don’t we take 45 minutes and try to connect to a speaker.

RZ This is an interesting psychological, social, technological—

PF I can’t make a clear phone call and I can’t connect to Bluetooth speakers.

RZ Jesus. Here we go. Here we go.

PF So this is exhausting.

RZ Connecting a Bluetooth speaker is something you have to negotiate. It is a very painful process.

PF You know what else? Any of the free public wi-fi services? Where the ones that are connected to like your cable provider in the city?

RZ Yeah.

PF Also a nightmare.

RZ Yeah.

[10:30]

PF You just—I leave my house, I turn off wi-fi immediately [yeah] because someone’s gonna try to help me and it’s gonna like ruin everything.

RZ Yeah, and then there’s that—there’s that router that always reaches outside. It’s like the DG6.

PF Oh yeah.

RZ What is that router?

PF Oh there’s a couple of routers.

RZ It’s just—it’s the same thing over and over again and it asks you the password when you’re just walking down the street.

PF It is so hard to connect things to other things. Little tiny worlds, unto themselves, they’re ok. But when you decide, you’re like, “Oh I’m gonna auth into my thermostat.” Or “I’m gonna connect to this speaker.” You can see that’s where the whole system breaks down [yeah]. Everything that we talk about, everything we believe, all the product work we love, it’s all great as long as you’re in a box.

RZ Yeah.

PF And then you try to draw that line between the two boxes and you’re utterly screwed.

RZ Apple works real, real—like AirPlay is pretty damn elegant. Apple works real [you know what I’ll say—] to try to make that easier.

PF They’re the ones who take it most seriously.

RZ They do.

PF Like the—there was—this is getting a little old school but like at one point Microsoft got very into the idea of plug and play. That was like a big feature [yeah] in—in Windows and so—cuz that was when you couldn’t plug anything into anything [no]. You know you’d buy a printer and then they’d update Windows and that was it for your printer.

[11:41]

RZ You have to go get drivers [yeah]. Nobody understands what’s going on.

PF God! That’s another thing to talk about: one day I would love to do a show that’s just about like HP printer websites.

RZ Oh it’s unbelievable.

PF Oh and you’re like, “I have to download like a four or 500 megabyte driver.”

RZ They’re huge! [Yeah] It’s—it’s pretty amazing.

PF Um anyway—

RZ I just wanted to round this out by saying I think there is an instinctual reaction we have when technology goes a little too far. We’re losing that battle with phones. Like you get on the train and everybody’s head down on that phone.

PF But here’s what I would say: what else should they be doing?

RZ Look around you!

PF No! Why? Who are you gonna look at? Some guy who’s—do you really wanna make eye contact with someone on the Q?

RZ That’s fair. That’s fair.

PF Like I mean this is the—we judge people for that, there’s two ways to—

RZ I’m really talking more about I’m meeting someone for drinks. And we end up on our phones like six minutes in which is so bad. Like, keep your phone off.

PF I think most people—I mean I go out with you all the time. I think mo—I think we’ve kinda stopped it.

RZ We did.

PF Most people have stop—you just, honestly, there’s a little bit of an obligation to kinda keep the conversation up on both ends too.

RZ Yeah [stammers] I mean specifically you can bring it to a halt sometimes.

[12:47]

PF Now there are some people who you see their hand kind of constantly moving towards their pocket.

RZ Yeah, yeah.

PF And they’re trying—and it’s like and their rarely keeping their end of the bargain up because they just wanna get back on that phone.

RZ You know when it happens? Like I’m hitting the rest room and boom, within a second [oh!], you’re up—that phone is up and running.

PF Yeah, of course.

RZ Yeah. So I do think thought that we are going to keep an arm’s length for certain things that go a little too far like these voice products, like the Bluetooth headset, like Google Glass which was preposterous looking.

PF Alright so we have some boundaries there.

RZ I don’t think we think about ‘em too. I think we just say, “Alright, look, this is getting silly [oh it’s revulsion!], I need to go kick a soccer ball in the backyard.”

PF Well this is what’s tricky, right? Google Glass produced a feeling of revulsion. That’s very hard to overcome.

RZ It’s true.

PF Even the motive force of Google’s incredible access to marketing and capital was not able to overcome the situation that it created for itself with Google Glass [yeah] and, you know, I think that’s really, really hard when you’re making stuff like that because, look, if Google had brought me in the room and been like, “What do you think about this new Google Glass product that will augment reality?” Almost invariably I would’ve been like, “Wow, that seems really interesting [chuckles].”

RZ Yeah, well it is interesting! Right? [Yeah] They were overlaying Google Maps and you could all kinds of—it was technically interesting, right?

PF And you bury the part of yourself when you’re building and assessing this stuff, you bury that part of yourself [yeah] that would look at it and go, “This is a terrible idea. People look stupid wearing it and it feels invasive and creepy.”

[14:17]

RZ Eyes and eye contact are a very primal, meaningful thing.

PF Right. Did anyone stand up in a meeting and say, you know, “Sergay, [chuckling] don’t do this.”

RZ Yeah. Yeah. I mean it was—

PF Probably not.

RZ No, it ended up in a lab [yeah] setting where it was a fascinating puzzle, right?

PF Exactly.

RZ And I think—I think the psychologists and the sociologists weren’t brought in and I think they learned a lot once they—they went outside. So.

PF So I think, you know, voice is funny because it’s—it’s sort of found its place. People are very excited by it. It feels to me—I’ll tell you what I think about voice is that it’s just another thing that’s gonna get absorbed into the larger computing substrate that surrounds us at all times [yeah]. Like Alexa right now is a big enough circle with a little speaker in it but it’s gonna be either thumb size or built into your phone-u-later.

RZ It’s in your la—lamp.

PF Yeah. Or like your headphones [yeah] or whatever [yeah]. Just this moment where like voice will just kind of get layered into everything. And I think it probably has some, just some hard limit. It’s like another—it’s like another mouse.

RZ Yeah, I think that’s right. And I think we don’t want it to really—like there won’t be that demand to go beyond that limit.

PF You know what’s interesting is that the voice interface has made people suddenly conscious of how much computers perceive, when they are just on and watching and listening [sure]. Now the reality is: that’s been going on for awhile now, it’s just that you bought a disc home from the [yeah]—you know like a circle home from the circle store [yeah] and you’re looking at it, and it has a little glow right on it [that’s right], and it’s a black device, and it says things, so you’re like, “Woah! Wait a minute! It’s been listening all this time.” But that’s—you know in a non-auditory capacity, that’s like what’s been happening on the web when you move your mouse around on most screens [RZ chuckles right] like it’s just—

RZ The listening has been happening for a very long time.

[16:01]

PF Yeah and everybody’s like, “Oh my god! Am I gonna give [chuckles] Amazon all the data about what I’m saying in my home?”

RZ Yeah. You already have.

PF You kinda have like when you use your computer, somebody’s mostly watching that [sure]. [Stammers] You’re under observation at all times.

RZ Yeah I think that’s a great observation. People just don’t—it’s just not been this obvious.

PF Now, things like this come along. Let’s change tack a little bit, things like this come along, where, “Hey! Big, new interface!” First of all, everybody gets excited. Everybody gets like outta their mind. “Oh voice! Voice is the future!”

RZ Yeah.

PF “Voice is—oh my god. This is it. No one’s ever gonna use a screen anymore, people are just gonna—their eyes are gonna atrophy, and their ears are gonna grow to twice—”

RZ This was the same sentiment that rose from bots, right?

PF Right, right, right. Conversation interfaces—things come and go in our fad driven VC funded industry [yeah]. So like voice gets big or Amazon does something with voice and then suddenly someone’s like, “I’m gonna create a 20 million dollar fund that will invest in Amazon skills developers.”

RZ Right.

PF Right. Or inbots.

RZ Right.

PF Or whatever. And, you know, a lot of times that is—that 20 million dollar check doesn’t get written but like a 30 thousand dollar check gets written out of the 20 million [mm hmm] and people are kinda like floating around. Bitcoin! Was another one. Where like, do you remember all the—all the Bitcoin startups like where did they go?

RZ Yeah. Very well funded, some of them.

[17:15]

PF Oh my god, yeah! And I’m sure you know 18 months, 24 months in now and it’s like, “Huh, wonder if it’s the bold future?” There’s a couple things: there’s Etherium—

RZ It’s funny how it trips up, right? It’s not a bug [yeah]. It’s a social bug. And it’s a—it’s a psych—it’s a collective psychological bug.

PF That’s right cuz the tech—the technology is basically fine but then people start to over promise around it.

RZ Yup.

PF So like voice will just kind of eventually nestle back into the computing universe that will surround us at all times [yeah]. Bots are ok but, you know, they have this—they actually have the same problem voice has which is they can only recognize so much.

RZ It’s just—also—

PF I think people saw Slack and were just like, “We gotta ride the wave.”

RZ Slack I don’t view as a bot. It’s—

PF No, but I mean it—it created an environment for bot creation and consumption.

RZ But there’s nothing—like nothing hit it out of the park. Here’s the thing: let’s walk through it, right?

PF Well, what could a bot do?

RZ Right [stammers] a very logical case for a bot, for example, I type in a stock ticker and I get the stock info, right? [Mm hmm] That sounds great. But, guess what? I like the control of having a web browser going where I want to go and finding the information myself. Like it’s not—the leap you’re giving me isn’t a leap. Like it’s all a click away. It’s not that big a deal. The added value isn’t there. It’s just not there. Like the stock ticker is four character presses away. I don’t need it to be a chat interface.

PF There’s an interesting thing underneath chat which is that really advanced chat interfaces end up being these sort of declarative logic decision tree kind of platforms where—

RZ Those are interesting, yeah.

[18:49]

PF Right? Except that—I think that ultimately they are better expressed visually.

RZ Right.

PF Here’s what’s interesting about bots and chats in this world and where I think that from a product perspective it gets kind of cool to go uh which is that thinking about chat interfaces forces you to think about sequencing of user actions and gathering input and data from users [mm hmm] and that there becomes a kind of um . . . and let me explain this from declarative logic around how you’re building the program. So a lot of programming is—is you’re kind of hanging around waiting for people to tell you what’s gonna go here and what’s gonna go there. And when we’re talking about something that’s more declarative, you’re talking about kind of building a database, and then maybe like the user interface is adapting to input as it goes along. And that is where chat is very interesting cuz I say, “I like dogs,” and suddenly it’s off in a whole new direction.

RZ Yeah. And that’s—that can be interesting, that can be kind of curious but—

PF What’s interesting to me is UX doesn’t do that as much. User experience, visually, doesn’t adapt to information as much. It tends to be these sort of more holistic, larger boxes, right? That people put stuff into. You get a—you, it’s like, “Oh, you like dogs, well this will be the animal window and you put the animal in there.” [Yeah] But it’s not—suddenly the browser doesn’t become dog browser when you tell it that you like dogs.

RZ Yeah.

PF What I’m getting at here, and what I’m getting at is that chat and UX and some of the thinking that’s going into building the chat interfaces could start to have a visual expression and a sonic expression as well. Like you could start to—things could start adapting to users more [yeah]. That model could be there. Cuz right now the overall design aesthetic is you create a very simple interface that’s very, very predictable and logical and that it works exactly the same way every time. That’s a good thing. On apps. Like it cuts chaos.

RZ Yeah.

PF But when you think about—I’m thinking about certain clients that we have where they have like hundreds of thousands of rules that drive their business logic [mm hmm] and you do weird things with the UX in order to make that—make it possible for [yeah], you know, people to fill out forms—

RZ Yeah. And they’re incredibly complex forms.

[20:57]

PF That—you could build a bot from that data.

RZ You could but the thing is . . . that falls on its face, right? I need to see that infor—that information needs to persist for me. It’s not a conversation. When I do x, a whole other world of interface kicks in and you can’t ramble that off to me. Like it’s just, it’s too information rich, right? Like I don’t think it’s solvable. I think of wizards when I think of bots. Like, “Hey, let’s you get you going. [Oh! Not—] Here’s seven steps.”

PF Not wizards with hats but the like, “Hey!—]

RZ Yeah.

PF “Enter your name.”

RZ “We’re gonna go through six steps and get you to here.” And it’s uh it’s very linear, you can’t really veer off, they’re set—they’re setting you up. Right? “We need this information to get you going.”

PF Yeah, I think that’s the thing like computers can do a lot more than that.

RZ Of course they can!

PF Right? It’s just the tool—the tooling isn’t there. To do very adaptive things with how you react to people and gather data.

RZ No, I—I—exactly. Adaptive is the word here. Google is the best bot in the world. The URL box has been appropriated by Google [yeah, that’s as close to—]. If I type in “Venice”, [yeah] you don’t just get search results, you get the top of search [no, and it’s—] summary.

PF [Stammers] It’s as close to old school AI as I’ve ever seen.

RZ Exactly.

PF You put in words and it knows all about all the things and it’s very good at predicting intention.

RZ It’s very good at it.

PF It has a very good statistical model of mass aggregate human behavior and desire around knowledge, right?

RZ Yeah, but it’s one interaction is the thing, right? Like—

[22:30]

PF Oh yeah, no, no. I mean but you think there’s probably more energy and time spent on that text box than on any other like UX—

RZ Massive!

PF Yeah.

RZ Massive. Utterly massive. Like if I type “drug side effects” it’s gonna actually go and pluck the side effects and I don’t have to click through [yeah]. Like that’s incredible. And it goes on and on, right?

PF No, you’re looking at I mean ten, 15 years of just utter focus. At this point more.

RZ Yeah.

PF Yeah.

RZ That, to me, is the most interesting. I don’t wanna go back and forth. I don’t wanna go say, “Hey, ok well that’s interesting but can you tell me about the really serious side effects?” I don’t want to. I’ve got it.

PF Here’s the tricky thing though is that you—

RZ I wanna go read the thing.

PF Yeah, Google’s got you trained, though, you know kind of—we all have a good internal model of how Google works, and how to use it, and take advantage of it. But with new knowledge domains it requires a kind of interaction. Like, you know, Google isn’t great at gathering complex data . . . from people.

RZ No. It almost doesn’t want to—

PF No, I mean—

RZ—to go too far there.

PF That’s not its strength. Right? [No] It’s the same liability issue that we were talking about with Alexa. Like you’re not—not even necessarily legal liability but just it can only promise so much. And actually, I will say this one thing: Google has gotten much better in its core product in sort of not over-promising.

[23:40]

RZ Yeah.

PF Like in the product there’s a lot of little clues and—and things that communicate some of its own imperfections, and when it oversteps people flip out [yeah]. Right? Like when it gets the wrong picture connected to a person’s name [right], or—or like when people are able to sort of like push it to, you know, when a racist picture shows up when somebody searches for something like [right]—like it—it has learned the hard way.

RZ Is there a bot you can’t live without?

PF [In deep voice:] Nooo.

RZ Is there a bot that you really think is cool and you love?

PF I hate to say it cuz I love conversational interfaces. They’re fun. But, no.

RZ Nobody’s—nobody’s nailed it.

PF No.

RZ So far.

PF I think like um. . . no. There isn’t. Bots can be fun in a chat room, you know, and you can ask them to do things, and they—and everyone can kind of talk to them. Like they’re kind of fun as Tamagotchis [yeah] and virtual pets.

RZ You know what’s funny is that some—there are businesses out there—I’ll give you an example uh where they’re almo—they’re using humans . . . to simulate the bot. They’re essentially trying to cut out—like, for example, there is an app out there that is effectively a pharmacy doctor app. I did it. I had like a sinus infection and I—

PF Oh, tell us about that! That’s fascinating.

RZ Are—are you being funny?

PF No, I’m being funny.

RZ Oh ok.

PF Yeah.

[24:52]

RZ Um I thought I had a—I had a terrible like face headache and I—I said, “I don’t feel like going to a doc—it’s a whole process. I—I just need my—I like—every few years it gets bad enough that I need antibiotics. So I install this app, it asks me like 12 questions, essentially a wizard [mm hmm]: what’s wrong? What’s going on? And it definitely had a tree logic to it [ok]. And then they said, “Ok, we’re gonna a video call with in about an hour.” And this guy [chuckles] comes on and . . . he’s a doctor. He’s been—essentially what he does is he has a conversation with you about how you’re doing because they can’t just write a prescription after like 12 questions [mm hmm], like a doctor needs to sign the prescription [sure]. And I have the conversation with him and he’s like, “Alright, we’re gonna put you on a five day regiment.” And that was that. And—and it was great! It was really convenient. But, you know, the days of—first off: is a computer ever gonna be able to write a prescription? Holy hell, right? Like—

PF Honestly, within our lifetime, very—very possibly.

RZ See that’s insanity.

PF Why not—why not just a sensible decision tree? It’s gonna be able to look at your skin, and look at—like—like there’s a portion of medical care that is essentially where the doctor functions as a—as a robot.

RZ Are you gonna be cool with letting Alexa check the rash on your leg by essentially taking Alexa and pressing it up against your leg?

PF I, you know, I don’t think if it’s cool or not. I mean I think it’s like there’ll be a point where, you know, the Journal of American Medicine says, “This works ok.”

RZ I don’t wanna walk in on you and find Alexa pressed up against your leg.

PF But I think—here’s the thing, right? Like it—there might [RZ chuckles]—- there’ll be ten years, it’s like self-driving cars. There’ll be ten years where doctors have to review the leg’s photo.

RZ They have to sign off.

PF Yeah.

RZ Right, right, right, yeah.

PF But it’ll get less and less.

[26:41]

RZ Yeah. There’s a—Warby Parker just came out with a—an app that will determine your prescription. It’s a whole process you go through. Um it’s still very early days but you go through, again, a wizard. And then they make you put the phone at a certain distance and do a eye test. It’s pretty crazy, and I don’t know how they’re getting the answers. Like I couldn’t go through it because if you’re over 40, they won’t let you do it for—you also, you have to tell them what state you’re in because certain states are not allowing it [mm hmm]. Um but what they’re trying to do is cut the doctor process—cuz they—they don’t have doctors at all at their retail location [sure]. So here we go, right? Like again is an app going to determine my—my eye prescription?

PF Yeah.

RZ See that’s—that’s pretty crazy. I think.

PF It’ll be decision trees and it’ll just, I don’t know, doctors have all kinds of diagnostic software they use now.

RZ Yeah, I guess.

PF I don’t see a way around that, right? Like we have—it’s a—all the market forces align [yeah] to make that happen, whether it’s good for humanity or not, that’s happening.

RZ Yeah.

PF It is in everybody’s best interest given how completely we screwed the pooch on health care in our country to create [interesting] automatic systems.

RZ Yeah, no, I agree with that. I mean—

PF And so if you can get that cost down to ten dollars, then the insurers will give it to you for free. Like, ATNA is going to end up investing in apps for this kind of monitoring and compliance.

RZ Just to bring costs down.

PF Absolutely, right?

RZ Yeah.

PF They’re gonna end up—you’re gonna end up with like ATNA owning startups [yeah] around optimol—you know?

RZ Skin detection . . .

[28:22]

PF Yeah! That’s right, “Send me a picture of the skin tag.”

RZ Like blood pressure app—like they have them today. There’s a blood pressure tool that just—it does Bluetooth to your phone [right] of what—how you’re doing. And it logs it all and you can take it to your doctor and—

PF And I mean, you know, we tend to articulate and talk about this stuff mostly in terms of privacy in the world of technology.

RZ Yeah.

PF But I think when you think about it in the larger global sense, it’s—nobody’s at ATNA is thinking about it in terms of privacy.

RZ Right.

PF Right?

RZ Actually, you know, we’re also—we’re in a first world country.

PF That’s right.

RZ Like in places where they would never get this kind of care, to actually bring the cost down by 300, you know, 400 percent is a big deal.

PF Well we have a system of health here where it’s just so completely credential driven. But like if you think—I remember there was during—it might have been the sixties in China. There was a book called The Barefoot Doctor’s Manual. And you would get that, and it would tell you, it had pictures like “How to Deliver a Baby”. [RZ chuckles] Because there weren’t enough doctors [right], and there were a lot of people. So you had to get the information out [sure] and certain people would take on responsibility [right]. And I think I’ve got right, like I might have that story off. But I remember going and looking for it. I had it as a kid. For some reason I had a copy of this book and it was very hard to find any information about it but I remember it very clearly. So you’re gonna transfer the information and if you think about, you know, the blood pressure monitor that tells you how to adjust your life uh in an Android app that runs in all countries, it’s an incredible life giving tool.

RZ Yeah. Without a doubt.

[29:54]

PF So if you get the blood pressure monitor down to below a dollar and it runs on an Android smartphone that everybody already has, and talks, god forgive me, through Bluetooth [yeah], you’re probably going to save hundreds of thousands of lives.

RZ Probably.

PF You know and then on the flip side: we, here in the first world, we tend to talk about how scary it is when ATNA which is a terrifying company that doesn’t know how to build a good website that provides my health care here at—at Postlight. We worry about what it’s going to do with that data. Cuz it gets to track us our entire lives and we—there’s no trust or transparency with who has our health data. I’ve got health data scattered all around the world, right? It’s probably stored in, you know—

RZ Yeah, I mean, every time you have to claim coverage [right], they have the story.

PF There is stuff about me and my body in databases in—all around America. I’m sure, right? So—

RZ Yeah, I mean, that’s good, no?

PF It is and it isn’t! Sometimes you think—but if you think about them taking away your health coverage because they’re able to aggregate three or four data sources, and do some data mining. It’s not—it’s not good, it’s really bad.

RZ Data mining meaning you’re too risky, we want you off.

PF Yeah, right. And, you know, we’re covered by our—our corporate care here which gives us a buffer but I didn’t have that for—honestly this is one of the first times in my life except for when I was an editor at Harper’s that I’ve had stable health insurance [right]. Like I had to start a company to get it.

RZ Well, you didn’t have to start company to get it but—

PF No I might—

RZ But point taken.

PF There were other ways I could’ve gone. Look, I [stammers] the point I’m getting at here is just that it’s—it’s such a big subject and so . . . like there’s no way to even find out half the time. Like there’s—there’s probably somebody working on that blood pressure [oh there’s no doubt] compliance tool.

[31:33]

RZ They’re starting to head in that direction cuz they see opportunity there. No doubt about it.

PF What we hear about are the ones that are purely money driven. Right? We hear about the ones that get 20 million dollars in its own Crunchbase and—

RZ Well they are money driven!

PF Some of them aren’t. Some of them—I’m sure there’s somebody actually thinking about the cheapest way to get—

RZ Like the Bates’ Foundation Grant uh—

PF That’s the thing like what are they thinking about blood pressure in Android, right?

RZ No, that’s true. I mean . . . I think that’s ok. I think good money driven motivations lead to tons of innovations or—

PF No, I’m not arguing against capitalism—

RZ No, it’ll happen both ways.

PF What I’m saying is like there’s no—we get the news about these things, sort of either filtered through a tech culture sphere lens. Like, “Oh my god! You know this is gonna be bad for privacy!” [Right] Or through a money lens, like, “Oh! They got 80 million dollars in funding, this is a big deal.” [Right] There’s a whole nother world where stuff is happening, and people are using tools, and coming up with new ideas that doesn’t actually filter out [right] because it’s done with a, you know, with a 200,000 dollar grant from a medical science foundation.

RZ Right. All of that’s fine! I mean—

PF That’s just fine, I just like I wish I knew more about it.

RZ Mmm.

PF Cuz that’s the real world. Like, those things are gonna come from behind too.

RZ Yeah.

PF That’s probably where more disruption could happen then, you know, another VC funded juice company.

[32:47]

RZ No doubt. At the very least raising awareness so that, “Hey, you should get to a doctor.” Like it’s not gonna [right] cure—it’s not gonna treat you at that moment. But it’s like, “Hey, you know what? This test is a little iffy, get to a doctor.”

PF Yeah, your little—your little phone buddy [yeah] who keeps an eye on your health.

RZ Yeah. And that could be a bot. That could be Dr. Feel Good. Dr. Feel Good.

PF No. Conversational interfaces probably are pretty good especially like for kids [see that’s interesting, right?] and beginners. Like so the health science and connected internet of things devices ecosystem is just a beast, man. Just a beast. Every time I go poking into it, I’m like, “There’s another 50 million things I need to learn.”

RZ Yeah. There’s a lot of playing around happening right now.

PF I think it’s—Well I think at some level like you gotta be mindful: we’re in this world, right? Where we’ve got the graphical user interface and we create documents and we send Word Docs back and forth or Google Docs and we make websites and people put data into the websites and so on and so forth. And it’s like we’ve kind of had this like information processing productivity interfaces run the world and control our culture thing going for 20, 30 years now. And I’m wondering as the next 10, 15 years unfold, if the core computing experience isn’t just like, what’s in your phone but like the little thing you wear on your finger, talks to your phone [yeah], tells you kinda what—how much you need to—what you need to eat later in the day—um—

RZ Yeah, too much sugar, today.

PF You need a haircut.

RZ Yeah.

PF Like it becomes—

RZ Are you cool with all that?

PF It doesn’t matter if I’m cool or not. I’m here to watch it.

RZ No, would you—would you jump onto it? I guess what I’m asking, more generally, is will there be a resistance? The point I made earlier was we tend to resist at certain points.

[34:28]

PF But, again, you’re thinking about this as a cultural movement, right? What will happen is that these people will offer these tools. The like, you know, [yeah] the haircut watcher device [yeah] and it’ll be something you plug into the wall next to your thermostat and it’ll auth into your haircut watcher and every two weeks it’ll be like, “Hey! Looks like you need a haircut.” [Yeah] Which, you know, you could’ve done a recurring event but you spent 180 dollars and it was made by the, you know, it had—the same guy who did Nest [yeah]. So it tells you, “You got a haircut, great!”

RZ My—my brother will buy any of these things.

PF Yes, your brother will buy any of these things.

RZ I mean we—we had lunch with my brother and he got a phone call on his watch from a guy who he was working with and the amount of beeps and—

PF Your brother has an infrared camera that he plugs into his Android phone [yeah] and it had some construction capacity that I didn’t understand. He’s like, “Oh no, I need this!”

RZ Yeah, and he has—he has a pen with an HD cam—like a video camera.

PF Yeah! He’s got a video pen. So, wait, look, look—

RZ [Laughs] Some will embrace these things!

PF It’s not just that. It’s that it’ll be . . . certain people and certain temperaments are fine with having a machine tell them they need a haircut every two weeks but human being get bored and tired of everything and, I mean, I’m a quantified self veteran and I can tell you that certain—things just stop working. Unless you’re motivated. If it’s your little buddy going—and then you’re like, “Oh! I would’ve forgotten that otherwise.” Like if it’s saying, oh I need to go get a haircut. “Yeah, boy, I always forget. That’s great. Lemme go get that haircut.” But if you just don’t want a haircut, you’re gonna ignore the hell outta that thing. And then what’s the point of it?

RZ Not only are you gonna ignore the hell out of it, you’re gonna throw it out the window.

PF You’re eventually just gonna be like, “Well I stopped using—The batteries ran out. I didn’t replace them.”

RZ There’s a list of stuff that’s been tried [yeah]. Like the juicer story. You know there’s juicer products—

PF Juicero.

[36:15]

RZ Juicero. Right? It was like—people, it turns out—it squeezed juice out of these packets but it turned out humans could squeeze the same amount of juice.

PF There’s a big article in Bloomberg Businessweek around this [right] and it just showed people squeezing the juice [yeah] and it turned out to be much—but, you know, honestly, it looked kinda good. It was a nice looking juicer product.

RZ I mean this is the overreach, right? Like it’s like—

PF Somebody took it apart, it’s all aluminum inside.

RZ There’s a scale, there’s a cooking scale that’s like a thousand bucks where it like—it’s tied to an app so when you’re weighing the different ingredie—it’s reached a ridiculous place, right?

PF I think there’s a—there are fantasies around design. I think people—Apple has really changed what is acceptable in the high end of the market.

RZ Right.

PF And so it’s not just that you can make a good looking product but you have to have a narrative around it and everything has to be the very best [yeah] and uh and people also just sort of vanish into their own asses. That happens a lot. And so—

RZ Yeah, I think, my prediction is a haircutting tool would fail. I think people think—assume there are conveniences that people want and they actually don’t want and in fact will find them very annoying. And—and I think you’ll see that a lot.

PF Well and most people just want like good health, and power, and authority over their environment.

RZ Yeah, it’s gotta be a serious leap!

PF Right. [Right?] So if you think—you know my big question when we’re talking about product, I always frame it like what superpower does this give me?

RZ Yeah, great way to put it.

[37:37]

PF The haircutting device would never give you a superpower. Cell phones give a superpower: they connect you to all of humanity at all time. All the information is wherever you are, it’s just there.

RZ It’s insanity! That window is—

PF That’s an actual superpower. Like if you had—if you created a comic book hero in 1955 [right] called, you know, Mr. Smart-O and he could [yeah] see everything, and talk to anyone, at any time, [right] and all the phone lines connected to him through a special wave that went into his brain [right], it would be like, “That’s a cool comic book! What’s Superman gonna do to fight him?!”

RZ Right.

PF Right? And we have that in our pocket. So that is like a 1950s superpower is a good metric for if something’s gonna explode or not.

RZ Yeah.

PF Alright well we nailed it down—

RZ I think we just solved internet of things.

PF Really came to a lot—

RZ Right here in 40 minutes.

PF—a lot of great conclusions.

RZ Yeah.

PF This is a—this show is a great way for people to see kinda how the sausage is made [snickers].

RZ Yeah.

PF This is what—what it looks like when the wheel [slowly:] turns and turns.

RZ Yeah I mean this is the pain of product strategy, right?

PF Yeah!

RZ I mean this is what you have to think through.

PF Just have this conversation for 30, 40,000 hours and so um [RZ chuckles] [music fades in] my name is Paul Ford.

RZ I’m Rich Ziade.

PF You’ve been listening to Track Changes, the podcast of Postlight, a digital products studio at 101 5th Avenue in New York City. Uh if you need us, god, how do you get in touch with Postlight, Rich?

RZ Uh [email protected] You don’t even have to need us. We love talking to people.

PF It’s true.

RZ So feel free to email us questions, thoughts, comments, uh inquiries, anything.

PF Alright, let’s get back to work.

RZ Take care, guys!

PF Bye! [Music ramps up, plays for eight seconds, ramps down to end.]