Episode 187 - Coding from Natural Language, NYT Tech Walkout, Off Predictions
Today's news update covers OpenAI's Demo where they showed how a natural language interface could be used to write code. What does this mean, and where will it be useful? We also cover the walkout of tech journalists at the New York Times, and people talking about their failed tech predictions on Twitter.
Links
OpenAI Codex Live Demo
MSN: New York Times Tech Staff Walkout In Growing Union Fight
Vice: “Socialist” Current Affairs Magazine Staff Walks Out
@btaylor on Twitter: Technology Predictions that Turned out to be Wrong
Related Episodes
Episode 177 with Further Discussion on the Last Prediction Panel
Episode 134 on OpenAI’s GPT3
Transcript
Max Sklar: You're listening to The Local Maximum Episode 187.
Time to expand your perspective. Welcome to The Local Maximum. Now, here's your host, Max Sklar.
Max: Welcome, everyone. Welcome. You have reached another Local Maximum. I'm pretty sure that it's September by the time this comes out. Is that good or bad that it's September? I feel September is a bad luck month but I don't...
Aaron Bell: I like September, but it definitely makes me feel like, "What have I wasted the summer not doing?"
Max: Yeah, that's true. That's true. All right.
Aaron: We're talking to you from the past into the future here.
Max: It's not that long. It's not that long, hopefully, unless we save this episode for a long way. Let's actually get into the meat of it. I don't think people want to hear us talk about this. All right so we have a few tech and AI things talk about as a news update. Let's start out with this story about The New York Times tech staff. What's happening here with that?
Aaron: Yeah so over the great lady at The New York Times, apparently, the news Guild of New York, they're attempting to unionize the tech workers there. Specifically, looking at folks like data analysts, designers, software engineers, and the like.
Max: Now, sorry to interrupt you but as someone who was part of The New York tech scene, it used to be a pretty big deal to get a job at The New York Times, as a data scientist, in particular. They had a pretty sophisticated operation over there. Does this include that group?
Aaron: It certainly seems like it. I would expect that this also includes some folks on the IT professional side, but that wasn't explicitly called out in the article. There's especially some interest in who does and does not qualify here, which I think is the point of contention. Specifically, New York Times management wants to exclude data analysts and designers from the group eligible for this vote.
I guess the way it works is that there's been a petition circulated and they've gotten enough signatures and then there's going to be a vote of the eligible people of: "Do we or do we not want to unionize?" They're trying to negotiate which people actually get to participate in that vote and as a result, be part of this group unionizing.
Max: Software engineers would vote for it, and I don't say that in a loving way to software engineers. I know that if that was on a ballot, everyone would just click yes and not think about it.
Aaron: Yeah, and they're not the only ones going through this. There's been a lot of talk in the last couple of years of Google and Amazon and Microsoft potentially unionizing. I don't think any of them have, at least on the high tech worker side, that many of those companies may have portions of their workforce that are unionized but that are in the more traditional unionized work groups that we're used to seeing. Traditionally, there's a big divide between blue versus white collar and the hourly versus salaried or exempt versus non-exempt.
Max: Right, so this is different. You don't normally think of software engineers as being unionized, and yet here, they are trying to unionize in The New York Times. The headline is that the tech staff walked out so it's not simply a unionization fight. Are they losing these people?
Aaron: I think the walkout was not like, "We're quitting" walkout. It was like a mini strike to display their displeasure with how things are going. To be fair...
Max: My theory, can I just say before... Sorry to interrupt again.
Aaron: Go ahead, lay it out.
Max: I have a theory: As a software engineer, you're better off doing a permanent walkout because they don't realize how screwed they are until you actually leave. Sorry, that's just my opinion of things most of the time. But in The New York Times, I would think... Well, look, a lot of these are engineers who want to use their prowess for propaganda purposes so they really want to be there.
Aaron: Some of that may be going on, and I think it's fair to expect that the people who are most strongly advocating for the unionizing are going to tend to be aligned with those kinds of views already. Potentially, at least the way that the News Guild of New York wants to see it, this is a group of 600 employees that would fall under the new unionized guild. The pushback from management trying to exclude the data analysts and designers would knock a third of those out so I guess knocking that down close to 400.
This negotiation between the two groups of who should and should not be included in the vote is being mediated by the National Labor Relations Board, which I believe is kind of the standard approach that they act as an intermediary to make sure that this all runs smoothly. Apparently, the union went public with some aspects of this negotiation in a way that management either wasn't expecting or feels was not playing fairly.
Max: What's the fallout from this?
Aaron: Well, it's unclear where it's coming down yet; the vote hasn't happened. I think it's something to keep an eye on because how this goes here may feed into what we see at, like I said, places Google and Amazon. There were a couple of things that I wanted to point out related to it.
Max: What does it mean for reporting at The New York Times? Are they going to have... Because they've had if I'm going to say something positive about The New York Times, let me take a deep breath here, they have had some pretty good data projects in the election when it was like, "What's the probability that each candidate wins in real-time?" That was pretty cool. They have had some really good data widgets and data sets that are available, and I'm worried if they...
Aaron: I would be shocked if this unionizing movement is successful, if that dramatically changes what we see there. I don't have any special insight into what's going on with management in the editorial board of The New York Times, certainly. I don't foresee a major realignment as the result of this. I think this is a lot more about benefits and finances, necessarily, than it is about the editorial tilt or mission of The Times.
Max: Gotcha. Oh, man. So The Times is not collapsing just yet?
Aaron: I don't think so. There were a couple of things that I did want to point out that are kind of related to this. We talked about how there's kind of this white-collar, blue-collar divide traditionally, but there have been notable exceptions. A traditional one is pilots are generally unionized, and most people don't think of that as a blue-collar profession. Although, if you realign your view of them as the bus drivers of the sky, then maybe that makes a little bit more sense. Also, Boeing engineers have been unionized since, I had to look this up before the show, since 1946. They are kind of the standout exception in the field of, certainly, in defense engineering, which is where I had worked for many years. None of the engineers at your Raytheons, your Lockheed, are in an equivalent union to the society of, for professional engineering employees and aerospace which is an AFL CIO-affiliated organization.
Max: Yeah and we're also going to talk about current affairs, right?
Aaron: Yes. Yeah so this pops, I think, the last week. Current affairs, I'm not labeling them. They self-identify as a socialist magazine but the...
Max: It's like The New York Times except they actually labeled appropriately.
Aaron: I'd say they're even further to the left. The two poll quotes there: The first is that they fired most of their staff for trying to start a worker co-op. I think this article was actually from Vox so not a right-wing take on this either. Directly quoting one of the workers in question, they said, "We were fired by the editor in chief of a socialist magazine for trying to start a worker co-op." Five fired staffers wrote in a letter posted on Tuesday. At the end of the day, even a organization that has socialist objectives doesn't like being... They don't like it when the employees are trying to usurp power out from management.
Max: Yeah, of course.
Max: There's absolutely a little bit of schadenfreude to this, but it raises some interesting questions about people who spout these views and opinions, how much do they actually believe it when the rubber meets the road when it comes to affecting them as opposed to telling other people what to do?
Max: Yeah. Well, just to make things fair and balanced here, I could say the same about certain business owners or business interests who claim to be pro-capitalism. Although, how many of them do these days is sort of a... They claim to be pro-free market, but then want to use government to regulate in their favor and make it more difficult for competitors to come in. In this case, just to be fair, I'm willing to say that there is a lot of hypocrisy on both sides.
Aaron: Sure. I think the takeaway shouldn't be: "At heart, everybody's really a capitalist." It should be: "At heart, nobody likes to be on the short end of the stick." Everybody has an inherent interest in their own personal advantage to a greater or lesser extent. I already threw out a comment there about a little bit of schadenfreude here that seeing people posted it on their own petard is, I may take a little bit of guilty pleasure in that.
But my more general take would be that if folks want to unionize, that should be something that they're able to do. If they can get the votes together and they can collectively bargain, then more power to them. However, I think, certainly, the positions that I've worked in, you're basically working on an EdWel contract so if your employer decides that, "Yeah, we don't want to keep working with you for…” pretty much literally any reason, then they can terminate that contract.
People attempting to unionize should not be surprised if management decides, "Yeah, you can do that but it's no longer worth the hassle of dealing with you. You're not so valuable that we can't replace you." That's a risk you take when you do this.
Max: That's the risk you take for a lot of things. I've been promoting the idea that you should, at times, take risks of being a little insubordinate in order to, if you have an ethical concern or if you have something like that is going on, if you're kind of being caught in a management power struggle and you don't necessarily want to help one side. There's all sorts of things that happen that allowed you to be pushed back, and you don't always get fired. We've seen very small minorities and companies be able to pressure management.
Aaron: It's a much stronger position if you are actually willing and able to walk away. I think a lot of people in these discussions are posturing that way but are not actually willing and able to walk away/be fired as a result. That could lead to some rude awakenings. Again, I don't know what the nature of the particular individuals here, of their current contracts are, whether there's stipulations that make that different than what I'm imagining here. That very well may be the case.
Max: We live in a world and the rise of the big corporation is a big part of this, which has been caused by all sorts of interventions into the free market but don't get me started. It's just the fact that it's difficult for people to walk away is a really big problem. I would recommend people make it easy for yourself to walk away from your job or from a client, if you can. These days, it's not something that everyone can do, but it's definitely something to look for if you can.
Aaron: Moving on from tech workers who could be let go for how they're advocating on the job to technology that could go on to replace them, they got this OpenAI Codex live demo that you brought up.
Max: This will never replace an engineer but it's pretty cool technology. I don't want to say never because then, somebody's going to wrap this around my neck and 20 years.
Aaron: Certainly not there yet.
Max: No, but this is a really cool demo that I found from OpenAI. And we've talked about OpenAI before when we did the episode on GPT 3. Oh, let me get the let me get that episode number out for GPT 3. I should always have the archive up. GPT 3. What was it? I can't find it. Anyway, all right. I'm sure I'll find it, but anyway, we talked about this before. These language models that are very sophisticated and OpenAI is the organization that's working on it. They have a solution now or some software now where you tell it what you want the code to do in plain English and then it writes the code.
The first part of their demo was just the simple demo, which is not as impressive because they say, "hello world" and then it writes code that prints: "hello world." Okay, not a big deal. Then he goes, "Say hello world with empathy" and it literally just types: "hello world with empathy" so you think it doesn't change the font or anything. You're like, "Okay, it's just copying.” And then, they had to do a bunch of work to ask for spaces and to serve the page.
It's a good demo, but I was very impressed with the second part where they were asking it to print graphics on a website and to find the rules and to say, "Hey, move this graphic to the right when someone hits the right arrow. Move it to the left on someone hits the left arrow." Essentially, you say that, and then it writes the code into, in this case, JavaScript. I can't tell you how... Especially when you're dealing with a new language. When it comes to taking a keypress and having it do something on to a graphic for example, if you're designing a game, that should not be a complex operation in terms of writing code and usually, it isn't. It usually does take forever to figure out how the heck I do this in whatever language I'm using, particularly if it's JavaScript.
It looks really useful if you could say, "Hey, write me some code that does this." And then it does it, and then you're like, "Okay, now I could edit the code myself." A lot of their follow ups when they're like, "Have the code do this and have the code do that." It sounds like, "Okay, now that I see how it's done, I could easily change it myself." But it looks like, first of all, a really great technology to fix all of these Stack Overflow lookups, which I'm looking forward to. Who knows? Maybe I'd be interested in trying it out, seeing if you want it to do something a little more complicated, like what would it do?
Aaron: Yeah, it struck me very much as... I don't consider myself a software engineer although I do occasionally write some code. I've heard it described that there are two general types of people that you find in software engineer roles. There are people who went to school for computer science and there are people who went to a coding boot camp and learned to code.
It sounds like this kind of tool, maybe it would replace some of those people who went to the coding boot camp. They've learned how to do a thing as a skill, whereas it would be very much like a collaborator with someone who has a computer science background because they can focus not so much on the nuance of this particular language and implementation. They can talk about kind of the architecture and the underlying concepts and ideas and how that's being done.
Certainly, for those coming from the academic computer science background, that's the much more interesting thing to them anyway, and it lets you take a shortcut through a lot of the "busy work" around the actual implementation of those ideas. I think that gets to a really cool area of, I guess you could call this AI, where it's not about replacing humans. It's about building collaborative environments where the human and the machine can play off each other's strengths.
There's been a lot of talk about how no longer can a human beat an AI in chess but I believe, and I have to go back and check if this is still the case, but at least there was a period of time where a human paired with an AI could still beat an AI chess player. That there's... What's the word I'm looking for? It's not synthesis. It's a corporate buzzword.
Max: Synthesize? Collaboration? Synergy?
Aaron: Synergy! That's the word. Yeah. There's just there's a synergy there between the different skill sets and the ways our minds work.
Max: By the way, this is why doing a solo show is so hard because I keep forgetting the word.
Aaron: Oh, yeah. I never would have gotten to synergy without that. I agree that the kind of "hello world" part in the beginning was not super impressive but they needed to lay out kind of some building blocks for what they're doing. Otherwise, it would just look kind of black magic when they were doing the latter part.
Max: It reminded me a lot of... When I was an undergrad at Yale, I took a class in functional programming with the late Paul Hudak, who sadly passed away very, not super young. He was a professor, but 60 years old, he was still involved in all that stuff. He, I believe, was part of the creation of Haskell, a functional language and he was showing us how you could write a game in Haskell, a game of Pong in 17 lines.
I remember being at the beginning of that lecture being like, "There's no way he's writing a game in 17 lines." And literally, the 17 lines of code looked like: "Define left wall. This is how the left wall works. Define top wall. This is how the top wall works. Define right wall. This is how the right wall works. Define paddle. Connect the paddle to the right, left buttons. This is how the paddle works. This is how the ball works." Lo and behold, in 17, I think you said under 20, but it ended up being 17. You literally create this paddle and ball game in code.
It didn't keep score yet. There'd be a lot to add to actually turn it into a shippable game, but it was playable. I thought that was so cool and they did something very similar in this case where you could actually define a game. You can define physics, almost, very easily and with just phrases and in JavaScript. I've done some animations in JavaScript. I don't recommend it but the fact that you could do it just by voice command and having it...
Essentially, what it does is it's using statistical tricks with a deep learning neural net. Think of it as it doesn't really understand what it's doing but it's statistical tricks plus, I would describe it as. Bayesian stats plus some extra so it's kind of in between statistical programming and understanding. It's kind of in that gray area where it can get really good at figuring out what you mean when you say something.
Now, I wonder if you want something really complicated. If you're describing a game or you're describing Foursquare and you're like, "Well, I want someone to create this venue but it could be a duplicate so warn them if it might be a duplicate." Then, they're like, well, I don't know what. There might be some very complicated things that don't work, in my experience.
Aaron: I was wondering kind of what the level of complexity it can handle.
Max: I was just going to say like a game with graphics, books that are very sophisticated but under the hood, it might not be as sophisticated. Again, I think this is incredible technology. I think it could be used, like you said, to really improve on ramping to new languages. I would be skeptical that it's quite there yet as a: "We're going to code up the whole thing for you." It's a complex data system.
Aaron: Yeah, I'm very curious kind of a: how it works with optimization and what kind of assumptions it's making. You could have a, kind of classic example, if you have some sort of recursive operation that needs to happen. There are several different types of functions you could use to do that. Do you do with a while loop or a for loop? Or do you use some sort of cases. If you just give it the high-level instruction, what's it going to do? Whereas if you were designing this from raw code yourself, you would conceivably have a specific outlook for how you wanted to handle that, and ideally, you would kind of optimize it for your approach.
Max: One of the things you can do with this is it does it the way it wants but then you could go back and say, "Now, do it with a for loop," and it'll do that. You can actually talk about the code as well.
Aaron: I was also wondering, will it be able to self-optimize? Or is this kind of a case where you could use this AI to build the code, build your application, and then, there'll be a separate AI agent that you can run on after the fact that will look for ways to optimize so that they're not... They need to be specialized in their own way rather than trying to solve a much more general problem.
Max: Yeah, it depends on how you want to optimize it. There's already code to do that that's not so much machine learning. Maybe it's AI, but it's more just algorithmic optimization. Yeah, I don't see why you couldn't do that. I guess I would ask, "Write a function that finds the optimal traveling salesman route between this graph," and go see how it does the NP-complete. Is that NP-complete or just NP? I think it's NP-complete. See how it does the very difficult operation, which is for those of you who don't know, the idea that you have a bunch of cities to hit and the order of the cities that you're going to hit was the most optimal.
Well, look, if they're all in a circle, you just follow the circles, easy. If it's small number of cities, it's easy, but it turns out from computational theory, from computer science, if you have a large number of cities and they're all jumbled up, it's very difficult to find the optimal solution. Although, possible to find a statistically very good solution.
Aaron: That's actually been floating around in the back of my head because the most recent chapter of the Neil Stevenson book I've been reading involves a version of the Traveling Salesman Problem. They use a quantum computing in many worlds model to resolve, which is a little bit of a different take on it but it's also in a not Earth universe.
Max: Yeah, I think in practice, it's really overkill because... Well, I guess our Google Maps instructions don't do it because we tell them the order.
Aaron: That's always been a function I've wanted to optimize for. If I say, "I've got to go to these five places today. Find me the most efficient route."
Max: It's not that it can't do it; it should be able to do it. It's just that it would take an incredibly long time to guarantee that it founds you the absolute best route but a computer scientist could get in there and be like, "I could design something pretty fast that's going to find you probably the optimal route most of the time and the times it doesn't, a very good one."
Aaron: This is exactly the type of problem that I have to work with sometimes in my job that we have a scheduler tool for networks that...to make a long story short, we set a bunch of requirements out to it and it has to solve it. You could have an algorithm that finds an optimal solution, but we don't really care about an optimal solution. We just care about a solution that meets all of our requirements, all of our constraints. You just want to find any viable solution within that space rather than waste computing time and other resources on getting to an optimal solution. If you really needed that optimal solution, then you would have had more demanding requirements from the beginning.
Max: Right, right, and you almost always don't. Okay. I want to move on from this, but definitely check it out. I'm going to post the OpenAI link, demo link, which is from YouTube, on the show notes page, some very exciting stuff. Should I name the developers? I don't know who they are.
Aaron: OpenAI Codex, but I'm not sure who they're affiliated with.
Max: I don't know if they have the people's names who are doing it. I always like to cite the actual people in the video.
Aaron: It'll be in the show notes for sure.
Max: Yeah. Okay, cool. All right, great. It's an interesting question that was posted on Twitter a few weeks ago that's kind of a question that we like to ask here on The Local Maximum and in my tech retreats a lot, that caught my attention which is from Brett Taylor, @btaylor. He tweeted, "What is a technology prediction in which you had a strong conviction that turns out to be wrong?" It went viral. There are a lot of really interesting answers to this so I just wanted to point that.
Many people cited self-driving cars. We've talked about that but our predictions on the commercial availability are still a few years in the future. 2030 is my date where if nothing has happened, I'm willing to say, "Wow, I was off." Since a lot of other people were off, and they thought it would happen right now, it's certainly one that seems deceptively difficult. That was a big one. That was the one that the original author cited himself and then many other people agreed.
Aaron: Is that mostly being driven by Elon Musk and Tesla? Or a decade ago, was there somebody else who was pushing the self-driving cars on the horizon?
Max: Well, there was.
Aaron: My sense of time is so mixed up that I can't remember how vocal he was about this that far back.
Max: No, no but in 2005, there were competitions and academic research going into it. I guess people assumed that, "Hey, in 10 years, maybe this will be out on the market," but turned out to be a lot more difficult than that. On the other hand, that's a very different situation where we're at now where we have trillion-dollar companies putting in billions of dollars to get this thing working.
We have Waymo that we've been following, even though it's just one company, but we've been following that directly how they keep expanding their area of service from first Arizona to parts of California. Then, it's like, "Well, what's next?" You're really thinking, "Nine years from now, they won't have expanded from that given that they're expanding a little bit every year?" We'll see. I could be wrong though because, again, it's deceived other people.
Aaron: While we were talking, I looked it up and it looks like the founding of Tesla was all the way back in 2003, 2004. Although, I don't know if they were talking much about self-driving at that point or if they were exclusively focused on the electric car piece of it at that point. Either way, I lost track of how far back their legacy actually goes there.
Max: Right, right. Another person cited Edtech, education technology and I'm not really sure. Well, I guess they had predictions where information and feedback loops would be, I guess, more than they were. I was working on this stuff and this caught my eye because I was working on this stuff in wireless generation back in 2006 to 2009, I was there. I get to say 2009 because I held out in that job until January 9th or whatever. That was on my resume.
They had a vision at that company where it was like, "Okay, everything you do in school is going to be fed into a machine and the algorithm is going to tell you, 'This is your learning style and this is what you need to do next.'" I guess it's no surprise that it hasn't quite happened even though certainly, assessment and feedback is a standard part of schooling these days. Although, I haven't been, I don't know, what are they doing in school these days? I don't know. I guess it's not that. That was just one that caught my eye.
Aaron: Yeah, I think a big piece of that is inertia, that there's a huge amount of inertia in the public school system. That even when you've found a technology that works and there's an improvement, rolling that out across the country and integrating it into the existing curriculum and pedagogy is not a quick and easy thing to do.
Max: Right. Another one put in was 3D printing, which I'm surprised we have not discussed 3D printing on the show. I looked into it and I couldn't find it. I don't think we ever discussed 3D printing.
Aaron: Yeah, if we have, it's only been in passing.
Max: Right, right. Apparently, this person says there was a lot of hype around it in the early 2010's. I went to the 3D printing store on Lafayette street in Manhattan.
Aaron: I think we just went through that together at one point. Yeah.
Max: Oh, yeah. I brought you. Okay, you were visiting and I was probably... Okay so that's pretty cool. That's not there anymore and in fact...
Aaron: But there are a lot of maker spaces, which are kind of the community-organized version of that.
Max: Right, and it's available. I know it's available in the Westport Library in Connecticut, and a lot of schools have it now so I wouldn't say that... And the Foursquare office has it and they actually use it to build little tidbits that they need around the office like iPad holders and things like that. It has been turned on since March 2020. I guess they thought the 3D printed world would be a lot bigger, like you'd actually have a real printer in your...
Aaron: I think there was a lot of expectations that it would be more like the Star Trek replicator. That there would be a huge array of things that you would no longer buy that maybe you would buy like the plans for it and you would just print it out at home, and you could do that. You can do that but for the vast majority of things, that doesn't make sense because of supply chain effects that...
Max: You just order on Amazon.
Aaron: Yeah, 3D printing is still very useful for prototyping but not for production. If you want something that's not a one-off, then you can probably get it better made and cheaper buying it through someone else rather than printing it yourself.
Max: I wonder if we could do an episode on this because it's a really interesting space of emerging technology.
Aaron: Yeah, it's an area that I'm very interested to get more involved in but I don't have that much first-hand experience with, unfortunately. Definitely, something I'd be interested in doing some more research on and talking about in a future episode.
Max: Here's another answer that grabbed my attention because I had forgotten about this. The Twitter poster, he puts his name as: "Steve is fully vaxxed." Okay. I don't know. I bet. I bet he puts a lot of things on his profile that are the... Okay, I'm already judging.
Aaron: Let's leave Steve alone.
Max: Let's leave Steve alone but his name isn't Steve, his name is Steve is fully vaxxed. Okay. He said, "When I first saw Kinect…” That was, if you remember, was coming with the Xbox. I really thought motion was aiming would be the...
Aaron: Was that the original Xbox or the 360 that had the Kinect?
Max: I think it was 360 but “....when Xbox wants Kinect even launched and they said it can't detect your heartbeat stress in the living room and the games you play will adapt to become harder because of…” I just think when that came out, it was really exciting. That came out when I was in grad school and people at NYU used to build, especially in the ITP program, people used to build all sorts of games off of the Xbox Kinect.
I took emerging technology in 2009 at NYU Stern and we mentioned gesture but the main part of our final paper was on multi-touch technology, like how to use your iPhone, which turned out to have a fairly large future with it. That was actually pretty... Some of what we said there was spot on but I think we... I'd have to pull it up. I'm sure we had some paragraphs on gesture control, which have not panned out.
Aaron: Yeah, the Kinect or whatever its equivalent feature was more or less phased out, I believe, of the latest versions of the Xbox; they're no longer using that capability. An early pioneer in that space was the Nintendo Wii. How much does the latest generation of the Nintendo technology really use? I guess it's not using the camera aspect of it. That's more motion sensor stuff.
Max: The current Nintendo Switches I have does rely heavily on the sensor on the gyroscope and all that but no gesture technology system.
Aaron: No cameras observing you.
Max: It kind of seems good enough.
Aaron: Although, all of the VR systems are at least I think the leading ones often have multiple cameras observing you as part of their system but that's kind of branching in a different direction than the mainstream gaming systems.
Max: Do you have a Kinect?
Aaron: Yeah, I still have my old 360, which occasionally gets turned on and has the Kinect.
Max: Maybe I'll give it a try next time.
Aaron: I haven't used that aspect of it for anything for quite some time.
Max: Okay, okay. Just turn it on, see how it works. Okay, another few bullet points that I found here where one person said they thought remote work would... Essentially, they said we thought remote work would suck less even though a lot of people are on it.
Aaron: We'd be even further behind on that front if we hadn't had a pandemic in 2020.
Max: Right. Right. Someone said they thought the AirPods would flop. A lot has been said on chat bots, which I could go into at one point, Google Glass, and the last one I want to mention, which I made a prediction about this a couple of shows ago, identity management. They were like, they said I predicted 20 years ago, we would have better identity management online and we're still typing in passwords into... And now, you have to type in a second password with multiple...
Aaron: Don't even get me started. I've got two-factor but there's so many nested logins that have requirements built off of each other that it feels every time I log into something for work, I have to get my face scanned three times, and my fingerprint, and then put in a pin, and then a password. Now, I can't read my email.
Max: Same here. Two episodes so a few episodes ago, Episode 184, I talked about the idea of having an identity management device on you.
Aaron: A dedicated device.
Max: That can handle this pretty well but the fact that the industry hasn't moved in that direction for 20 years, makes me scared about something like this being, or innovation actually being coming to the forefront or something like this coming to the forefront. Although, the rise of the crypto economy again could bring something like this.
Aaron: The biggest move that we have seen in that direction is kind of unifying under a single tech behemoth that if you're willing to commit to, for example, to Google, there are many, many services that you can get that all fall under that single login. You can reap some benefits of single sign-on. For a while, you could sign into almost anything using your Facebook identity but I think we're becoming more fractured now, if anything.
Max: No, yeah, because we've learned you don't want to be reliant on Google and Facebook.
Aaron: Yeah, part of it is trust and part of it is access to that even if you did trust Facebook or Google or whoever, that they can't necessarily provide all of the services that you need so you're going to have multiple identities to manage regardless.
Max: Yeahm and I think I mentioned that someone thought Google Glass would be a game-changer. I think the story on that is still out. Obviously, it wasn't a game-changer instantaneously.
Aaron: We have seen the Alexa Glasses, which I don't think they actually have an augmented reality piece to them. I think it's mostly just a portable device to allow you to interface with the Alexa, the verbal piece of the assistant. I could certainly see those ideas being merged in a way that becomes reborn from the ashes, so to speak.
Max: Right, right. I feel like it should be something a lot simpler than what they were trying to do. Don't have the camera because that’d freak people out and basically, just have a heads up display for some very simple things. The types of things that Alexa would do like if you have a... Do you really want those notifications? I don't think the story is over. I think this decade, people are going to try it. Companies like Apple are going to try to build something like...
Aaron: This is verging into sci-fi dystopia territory, but some sort of implants, whether that's implants on maybe high tech contacts or an actual implant on your eye or something that interfaces with another Elon Musk product, the neural link.
Max: Elon Musk can put his tweaks directly into your brain. You don’t have to go anywhere. All right, cool. I think that's about it. That just wraps it up.
I want to ask you out there in the audience, what's the technology prediction that you had? Or what's the future that you thought would happen that never really happened? That turned out to be completely wrong? Go on locals... Oh, Shoot. I made the order wrong. Just go to maximum.locals.com to log into our locals group and sign up and let us know, or email me at localmaxradio@gmail.com. That's about it for today.
Aaron, you have any final words before we wrap up?
Aaron: Yeah, I'm looking forward to seeing some of those predictions that didn't pan out from the audience. Maybe we can talk about some of them on the show in the future. While you're at it, send in a prediction for the next 10 years so when we get to our... What episode number would that be a decade from now? We can evaluate how those panned out.
Max: Now, I'm on the hook to do this for a decade, the 1,000th episode.
Aaron: Maybe it'll be a special reunion episode. Who knows?
Max: Yeah, yeah. All right, sounds good. Okay, have a great week, everyone.
That's the show. To support The Local Maximum, sign up for exclusive content and their online community at maximum.locals.com. The Local Maximum is available wherever podcasts are found. If you want to keep up, remember to subscribe on your podcast app. Also, check out the website with show notes and additional materials at localmaxradio.com. If you want to contact me, the host, send an email to localmaxradio@gmail.com. Have a great week.
