EP6 - How Autonomous Cars Point to the Future of Online Learning

In this episode, John and Jason talk about Jason’s first trip in a self-driving taxi, and how this might point to the future of online learning.

Links & Resources

• Join Our LinkedIn Group - Online Learning Podcast
• Jason’s YouTube video of his first trip in an autonomous car.
  • Waymo One - Autonomous ride-hailing service
  • Speedy Street Tacos, Phoenix
• Sal Khan talks about Khanmigo 
• 6 Ways ChatGPT Save Teachers Time: Edutopia
• John’s strawberry smoothie recipe
  • 8 fl oz unsweetened almond milk (30 cal)
  • ¾ cup (170 grams) Chobani Zero Sugar Vanilla Yogurt (70 cal)
  • ~150-200 grams frozen strawberries from your grocer’s freezer section (56-75 cal)
  • Blend relentlessly until smooth (John uses a Ninja brand smoothie blender)
• How Might We Use Design Thinking to Humanize Online Education? April 19, 2023 at OLC Innovate in Nashville, TN

Theme Music: Pumped by RoccoW is licensed under a Attribution-NonCommercial License.

 

Transcript:

We use a combination of computer-generated transcriptions and human editing. Please check with the recorded file before quoting anything. Please check with us if you have any questions or can help with any corrections!

False Start

[00:00:00] Jason Johnston: What's the snack today? Is that a Shamrock shake?

[00:00:02] John Nash: It's pink. It's yogurt and frozen strawberries and almond milk.

[00:00:08] Jason Johnston: Is almond milking a profession now?

[00:00:11] John Nash: Is it a profession?

[00:00:12] Jason Johnston: Yeah. Almond milking.

[00:00:16] John Nash: Yes. Yes. They train mice with their tiny, tiny little paws to milk the almond.

[00:00:23] Jason Johnston: That's amazing.

[00:00:24] John Nash: Now, that has not reached automation by AI.

[00:00:28] Jason Johnston: No, not yet. They can't make the robots small enough, I don't think.

[00:00:31] John Nash: Well, the nanotechnology, maybe it's, yeah.

[00:00:33] Jason Johnston: Maybe with becoming nanotechnology, they'll be able to have their little robot milkers.

Intro

[00:00:40] John Nash: John Nash. I'm here with Jason Johns. Hey, John. Hey everyone. And this is Online Learning in the Second Half, the Online Learning podcast. Yes, it is. We are doing this podcast to let you in on a conversation we've been having for the last two years about online education, and it basically goes like this: online learning's had its chance to be great, and some of it is, but a lot of it still just isn't. So, how are we going to get to the next stage?

[00:01:08] Jason Johnston: That is a great question. How about we do a podcast and talk about it?

[00:01:11] John Nash: That's perfect. What do you want to talk about today?

Self-Driving Car Story

[00:01:14] Jason Johnston: Well, John, I think you know this, but we are living in the future. It's not quite flying cars yet, but not long ago I was picked up at the airport by a self-driving car.

[00:01:28] John Nash: You were in a self-driving car?

[00:01:30] Jason Johnston: Yes.

[00:01:31] John Nash: What, what airport was this?

[00:01:33] Jason Johnston: Uh, it flew into Phoenix Airport. This company called Waymo, and you can look it up, you can download the app, has two areas that it operates out of. One is Phoenix and one is San Francisco. And they have very limited areas within the city that they operate out of, but it's a fully autonomous, without any human co-pilot or anything, car that will come and pick you up at your location.

[00:02:01] John Nash: So, no humans. I thought humans had to be in these cars.

[00:02:04] Jason Johnston: Apparently not.

[00:02:05] John Nash: Does it depend on the jurisdiction or the state laws?

[00:02:09] Jason Johnston: I think it is in a test round right now. And because of that, and you, I might have clicked on some disclaimers as I went through. I'm not sure.

[00:02:17] John Nash: I'm sorry, I'm laughing. It's like, well, because it's in a test round, we're not going to put any humans as a safety measure in the car.

[00:02:26] Jason Johnston: Right. Well, I was a human in the car. I guess that was the test.

[00:02:31] John Nash: Wow. Okay. So, so can you, it goes to limited area. So, you didn't go very far in this thing, or it took you where you wanted to go, or?

[00:02:39] Jason Johnston: I didn't, I couldn't quite get to my hotel with it, so it works like an Uber for anybody's, uh, that is done on Uber or Lyft. You have an app; you tell it where you want it to go or where you want to get picked up. And they had a location, uh, basically the Uber location for getting picked up at the Phoenix Airport. And so, I had it pick me up there, told it where I wanted to go. And so, I couldn't quite get to my hotel. So instead, I was hungry and I went to get some street tacos. Just probably a couple miles from the Phoenix Airport. So—

[00:03:09] John Nash: What kind of car picks you up?

[00:03:12] Jason Johnston: Oh, it was, it was cool because I didn't realize until I was at it, but it was actually a Jaguar, like a little Jaguar SUV crossover. So, it was a very nice car. Jaguar. Yeah. Bye, a Jag.

[00:03:25] John Nash: Wow. Yeah. So, wow. Okay. Test round in a Jaguar, you just get into the doors open for you or what's what? Walk me through this.

[00:03:36] Jason Johnston: So, I did the app thing, had to come and pick me up, and you could watch it coming around the corner, via the map on the app, just like an Uber. And then it just pulls up. Uh, what was interesting was it pulled up basically right in front of me and up at the top, it had this kind of like, I guess was part of the lidar, people can't see my fingers going into circular motion, but it was this like lidar at the top, but it also had a little screen on it that had my initials, JJ, on the top of the car.

[00:04:09] John Nash: Oh, like a little hologram kind of thing?

[00:04:11] Jason Johnston: Yeah, almost like a little hologram kind of thing on the top. And it pulled up, had my initials, and then the way it works is in the app you can't pull on any of the handles. They're recessed, and so you have to hit a button on the app in order for the handle to pop out so that you can open up the door. I had a suitcase and I walked around the car just to check it out, but also to figure out if there was a way I could get into the trunk. But there didn't seem to be any way I could get into the trunk, so I awkwardly loaded this suitcase into, and I wasn't sure if I should get into the front seat or not. And so, I awkwardly loaded this suitcase into the backseat and then got into the backseat with my suitcase. And then once inside it had a screen for the backseat where you could, uh, click for it to start. And I asked you to close the door and it had this very ethereal, calming music inside and was speaking to me saying, welcome. And then I could click on a button for it to go and start the trip. And honestly, this is something, John, that I've been watching for, feels like a long time. And I, like you, probably a little bit of a nerd, and so I've been dreaming about my first self-driving car moment, and it didn't disappoint. It was really a very smooth, luxurious, easy ride in the self-driving car.

[00:05:30] John Nash: Was it, it pulls into traffic and then, uh, presumably you're on a two-way street or something. You're at an airport and so there's oncoming traffic. You're not nervous or it doesn't do anything untoward or...

[00:05:42] Jason Johnston: You know, because we weren't going that fast and so I didn't feel super uncomfortable from that standpoint. Maybe I would've felt different if we were pulling onto a highway. It was very cautious. It stopped everywhere it's supposed to stop. It did its blinkers. One thing that surprised me though, one part of getting out of the, uh, airport was that it came to a roundabout, and I know a lot of humans that struggle with roundabouts. And so it was, it was interesting to watch it very just easily and slowly navigate, uh, the roundabout. But this is where it was a little strange and I did, I thought this was the case when it was going, but it looked to be that it didn't quite get me to the taco place. Like it just fell short by about a block. And I remember I was trying to get it to go right where the taco place was, but it wouldn't quite go. And so, it dropped me off and it pulled into this, this kind of dark industrial looking street way that was a little abandoned and maybe somebody was walking around, and it seemed a little odd to me. This...

[00:06:50] John Nash: Seems a little, yeah, because if you're putting yourself into the hands of a self-driving automobile that's supposed to get you to your destination in an area you don't know, and it doesn't do that.

[00:07:02] Jason Johnston: Yeah. Yeah. Now, I will say that there was a help button I could have done if I was uncomfortable. I have this video on YouTube, and we'll post a link so you can see my experience. And, uh, the, one of the programmers actually from Waymo replied to my YouTube and told me that I can always edit my drop-off point even after I've come to a stop if I wasn't happy with it. But it's not something I knew. It was probably in those disclaimers at the front or something that I didn't really read and just said OKAY to. And it basically stopped and I was like, oh, I guess that is it. So, I opened up the door and pulled up my suitcase, closed up the door, and then it just took off.

[00:07:47] John Nash: Just left you.

[00:07:48] Jason Johnston: Just left me.

[00:07:48] John Nash: Good luck.

[00:07:49] Jason Johnston: And so, I opened up my map app to make sure I was going in the right direction to the street taco place, and it was just about a block away. And I don't feel super nervous in these kinds of situations, but I did notice though as I was walking towards the street taco place, that it basically dropped me off just outside of like an adult emporium. You know, I kept on walking of course, and went onto the street taco place, but it just was one of those streets. And so, it was a little, that part of it was a little strange.

[00:08:19] John Nash: And, and you could have, yeah. And then you find out later, well, I could have amended that by staying in the car.

[00:08:24] Jason Johnston: Right.

[00:08:25] John Nash: Crazy. Wow. Well, what do you take from this?

[00:08:28] Jason Johnston: I've thought about it, of course, while I was there and so I was curious to see what your kind of first take as you were listening to the story and how this either applies to our own experience with, uh, new technologies coming in, or I guess directly I really want to be thinking about how technologies apply to online learning in the future of online learning.

[00:08:50] John Nash: I'm thinking about the fact still that you got in a car with no other humans in it during a test phase, and it seems in, I don't know why that seems a little insane to me. It's, you know, a 3,000-pound vehicle that you have, you sit in the back of that, nobody is in control. Right. That does seem pretty amazing.

[00:09:13] Jason Johnston: When I've told this story I've had a lot of people, they're divided about whether or not they would do that at this point, mm-hmm, in the stage of the development, whether or not they would get into a self-driving car.

[00:09:26] John Nash: I mean, I couldn't get in the backseat of my SUV and let other family members drive so I could take a nap. I was too, you know...

[00:09:35] Jason Johnston: Do you think you would feel, knowing what you have heard about my self-driving car experience, would you feel more or less comfortable with, uh, with family members versus Waymo? A Waymo Jaguar.

[00:09:49] John Nash: I think, well, firstly, I should make a big disclaimer that any proclivities I have about other people driving, uh, family members or not are my problem and not their problem. Right, right. I would probably not take a nap in that Jaguar. Maybe after some time of, uh, getting used to it, I would be more relaxed, but I think I would be pretty vigilant about what it was doing. Ah, that's interesting. I think it's interesting to think about if we leap to our considerations of online learning and what AI is doing. I can't help but think of the news this week, as we record this on the first day of spring, that Khan Academy is going to integrate GPT-4 into its modeling to create an intelligent tutor for the material that it presents.

[00:10:44] Jason Johnston: Instead of Conmigo.

Yeah, Conmigo, which I think is really neat, but also, it's this adorable little chat icon with big eyes.

[00:10:52] John Nash: Yes. Yeah. And so, I'm thinking, and I have to confess there. So I went out onto ChatGPT and I said, well, talk to me a little bit about self-driving cars and talk to me a little bit about online learning and personalized learning, mm-hmm, and it had some interesting ideas, but I think one is that, you know, when we think about a self-driving car taking over driving responsibilities fully and automatically, mm-hmm, that reminds me of what Khan Academy's trying to do, which is trying to bring some guidance and support just when the student needs it. I suppose that we could start to see this kind of technology into our own cars where we, it already does. I mean, we have passive radar systems and things for right, monitoring speed and doing cruise control. Yeah. But—

[00:11:40] Jason Johnston: Driver assisted, I think they call those, yes. A driver assistant versus self-driving.

[00:11:46] John Nash: Right. So how much self-driving will come along when you need it, or will it, you know, an adaptive system like what Conmigo will do, will give you explanations or feedback or additional resources. How could that be thought about for the cars and also vice versa? Yeah. Speaking of cars, and this will date me a little bit, but around the year 2000 or so, uh, when I worked in another place, I was on a business trip to Munich, and we were at BMW headquarters. And when we had some off time with the engineers we were invited to test BMW's first prototype of a passive radar cruise control system. Amazing. And I got to drive this car on the Autobahn and test this out, and, which doesn't seem that fantastical, but what makes the story fun is that the entire apparatus was operated by a Pentium 386 desktop machine in the backseat of this BMW three series, and the engineer who ran it was rebooting it as we were going down the road, but that's what they had this thing tested on, but that was pretty— Oh my. Well, hopefully it wasn't like Windows ME or something like that. Clippy comes up and tells me I need to slow down. Like no Clippy, Clippy. I already, I already muted you, Clippy. But what do you, what do you think about this and what does it relate to what we want to do with online learning.

[00:13:21] Jason Johnston: Yeah. And I think that's really interesting what you said there about the, uh, BMW because it speaks to how long we've been playing with this technology and how long this has been a kind of a dream of inventors and of drivers to be able to have some either computer assist driving or fully computer controlled driving. And I feel the same way about myself when it comes to online learning. I've thought for years about what if every learner comes into their degree or whatever, into high school or where, whatever you want to think about in terms of your kind of goalposts for learning. And instead of being slotted into the same classes with everybody else, you are taken on a personalized journey of learning, mm-hmm, that begins where you are and takes you at the speed that you can go, works with you in whatever your dispositions are and so on. And then gets you to your end goal when you want to get there, you know? And so, I've been thinking about this for a while, and even with self-driving cars for a decade I've been thinking about the fact that my kids who are approaching now coming to a driving age. I remember thinking when they were five and six, it's like, well, they may not even need driver's license. Right. We may all be in self-driving cars, and I won't have to take them to school. They'll just hop in the electric VW self-driving van, and it'll take him to school and then pop back whenever I need to head to work, but that's not the case. You know, we're quite a way off from that and I feel the same way about online learning. I feel like this idea of adaptive personalized learning is, is still a little bit of a little bit of a pipe dream or something that isn't quite yet actualized.

[00:15:16] John Nash: I keep thinking of you telling Siri to go pick up your kids while you sit and have coffee at your kitchen table. Well, when you talked about when AI might be part of a dialogue with a learner on their own, I'm reminded a little bit of my recent interactions with Bing's AI implementation and the way in which the AI chatbot closes out the conversations by asking me a reflective question, right, when I thought I was in the driver's seat asking it questions to give me data back, and I was of two minds there. One mind was, "Excuse me, I'm asking you the questions. I need some information." And then the other mind of me was saying, "Well, actually that's nice that you're asking me to reflect on why the heck I'm asking you this question," uh, because that sort of iterative reflection helps me think about whether I'm on the right track or what my ultimate goals are. Usually, I've formed a goal in my head about why I want to ask the AI bot a question, and I don't question that goal, right, and what Bing does is actually ask me to reflect on whether I'm even chasing the right thing.

[00:16:34] Jason Johnston: Yeah. Yeah. I think that's good. Had a couple thoughts here while you were talking. One is that in those early stages of Bing trial, they didn't have any limitations on how long your conversation could be, and this is, this is why a lot of people were getting into really strange conversations, I think, with Bing, because you could, you could have a conversation for hours with Bing, it would just keep on going. They would ask you a reflective question, you would answer back. It would ask another question, and it would really be like a conversation. And then I think in the limited, another limited rollout, I think they were doing maybe five or eight responses. Yeah. Mine is up to 15 now. I can do 15 responses. Yeah. I—

[00:17:15] John Nash: Thought it went, I had 20 the other day, then I dialed it back. So, it's 15 now.

[00:17:20] Jason Johnston: Yeah. So they're figuring out that kind of level to which they, you know, how long can we let this go on before it starts a little strange, or it starts getting to a place where Bing could be hacked a little easier in terms of going down directions, mm-hmm, we don't want it to go.

[00:17:37] John Nash: I got the feeling that Microsoft started to understand that in terms of prompt engineering, good prompts have to be scaffolded over a series of prompts, five, six, seven to get where you want to go. So, you have to feed it some things, ask it the things, tell it to make some assumptions about some things. Then you can ultimately ask on the eighth prompt. You can ask it what you want it to do.

[00:18:00] Jason Johnston: Yeah. Wherever it was at like five or eight, it felt too short for some things for me. Like I felt like I was just getting to it and then it would close off the conversation. Then I would have to start from scratch and kind of re-engineer my prompt knowing what I knew about it later on, but it wasn't as fluid. Yeah. Yeah, I agree. But I like this idea of this kind of reflective asking questions. It's not just there as a help but actually guiding students like in a Conmigo kind of way. Where the learner just isn't thinking about where they want to go, but the computer or learning system is also thinking about where it's trying to take the learner in terms of goals. And this is the possibility thinking of AI is when, and this is what we've talked about before, when AI is able to do our laundry for us, what can we do next with that time and energy? What can we build upon? And I think about that with math learning that once we got the calculator, I felt like we started to build on top of that, versus really focusing on a lot of those rudimentary aspects of math. But maybe math teachers would argue with me that that was a step back. I'm not sure. I—

[00:19:20] John Nash: Don't think that AI-induced laundry machines will make me want to study laundry more. However, I do think that as AI begins to take on tasks like the calculator did, teachers are going to begin to have time to rethink the concepts they want to teach, mm-hmm, and how they want to teach them and change the way they assess the attainment of these sort of higher order goals.

[00:19:54] Jason Johnston: Yeah. I think it's very possible that we are on the edge of another renaissance here that will be pushed forward by, by— I mean, we won't know until this time in history has passed, mm-hmm. But it's very possible. I was thinking about my other experience that night of driving from then my street taco place, which was excellent. I'll also put the link in the notes for that street taco place if you're in Phoenix, to my hotel that I used an Uber and some of the, just the contrast of the two experiences. I had an Uber driver who was a local Phoenix person. There was a big golf game in town coming up and some, and the Super Bowl was actually going to be in the next week. And so, we're chit chatting about that. It was a much more human experience for sure. I was not asked to tip the robot driver, so my per mile, the robot car was cheaper than taking the Uber. And so, I don't know if that's just a test, a test thing or if it is the way that things would be. But I don't know. I don't know where I'm at, to be honest. I don't know which one I would prefer. If we could have a world, I don't think I want a world with all robot taxis. You know what I mean? Yeah. Although it was a novel, it was a great experience. I don't know that I want that. I wish for a world where we replace even those really casual opportunities to connect with people.

[00:21:26] John Nash: I agree with you and it's also complicated because this touches on labor markets. It touches on— I mean, look how Uber and Lyft and ride sharing in general disrupted taxis, right, uh, and yet the labor practices of those companies that disrupted the taxi system are not exactly lauded. They are seen as an escape hatch for people, but those companies are not seen as being the most rewarding folks towards their employees. Yeah. So, it's, it's complicated, isn't it?

[00:22:01] Jason Johnston: It is. And, and tying it back to education, there seems to be a world teacher shortage. I know that for public school teachers, they don't feel like they have the time to individually address all the concerns of all their students and to be able to guide them individually. Most of them would love to be able to sit down and take them on a guided tour through the intricacies of figuring out the quadratic equation or whatever. But, uh, they just, you know, that's just not the case. And so, we've got a place like Khan coming in and being able to give a self-guided tour. You got 30 kids in your classroom. It's right there. It'll be available for all of them at their own pace, at whatever questions they have at any point. So—

[00:22:49] John Nash: Does the teacher become a human tutor on top of the AI tutor? I know that teachers today that are trying ChatGPT, even at 3.5 or the, and when the 4.0 becomes faster, are finding affordances that allow them more time with students because they can take care of the routine things that can take hours. Lesson planning, rubric creation, mm-hmm, uh, other proforma activities that districts require due to state requirements that are mostly paperwork but need to be done, mm-hmm, mm-hmm. Now they're done in a fraction of the time. If you can get the prompting right and then you have time to work with your kids.

[00:23:31] Jason Johnston: Yeah, so I think there's, I think there's lots of opportunity here for AI, the self-driving course to help augment students' learning without necessarily replacing, uh, the teachers that are there. But there may be some places that teachers do get replaced, I'm not sure. And there may be some teachers that is not, there's no love lost, but I don't know that we'd be looking back nostalgically at our Conmigo in the fourth grade and about how much Conmigo meant to us. Excellent point. No, but at the same time we were talking about Clippy earlier, you know, and so there may be some nostalgia with Conmigo.

[00:24:08] John Nash: Maybe. Well, I don't know. Uh, we, we speak about Clippy with nostalgic derision.

[00:24:13] Jason Johnston: Right, exactly. Yep. Yeah, it was a, it was different. It was a different place than our, than our fourth-grade teacher that actually took us by the hand and cared for us. So, yeah. Yes.

[00:24:25] John Nash: Yeah, I can name all my teachers from kindergarten through fifth grade at my elementary school. Huh. I don't know that I would look back and say, wow, I sure loved fourth grade with Conmigo. I would rather think about Mr. Dragget in third grade or Ms. Stewart.

[00:24:43] Jason Johnston: Yeah. Well, what do you think, wrapping this up today, John, what are your closing thoughts?

[00:24:47] John Nash: Uh, I'm impressed that we were able to make as many analogies as we could to online learning from your short ride to a taco stand in Phoenix. Yeah, it's fascinating.

[00:25:01] Jason Johnston: But yeah, that's good. And it also reminds me that we want to hear from other people as well. We'd love to continue this conversation. I'd be interested to hear about other people's experience in self-driving cars. Are they on the side where they would get into a self-driving car or they wouldn't, would they get into a self-driving course or they wouldn't, what their preference is. And maybe there's certain, uh, circumstances that people would or and wouldn't do it, I'm not sure. So, if you want to chime in, uh, find us on LinkedIn, we've got a group there. It's Online Learning Podcast at LinkedIn. Or you can go to our website onlinelearningpodcast.com and find all the shows there as well as a link through to our LinkedIn. And yeah, please like, subscribe. Let us know what you think.

[00:25:47] John Nash: Yeah, definitely. And quick shout out, join us at OLC Innovate on April 20th in Nashville. We're going to have a meet and greet and a recording on the 21st too. So, we're doing design thinking and online learning on the 20th and doing some remote podcasting on the 21st. And then this week, Jason, you are co-moderating a panel on March 23rd at UT Knoxville. I'm a guest on that panel. We're going to be talking about AI and its role and implications in higher ed.

[00:26:21] Jason Johnston: Yeah, we've got a bunch of amazing people on that panel. I'm really looking forward to the conversation and so I'm sure we'll loop back on that. I don't think this podcast will be out before that forum happens. So, yeah, it'd be interesting to do a little loop back on that. And I think there's a few people on that forum as well that we'll probably get in as guests in the coming months. Yeah, I hope. Yeah. Well, great talking to you, John. This was, uh, really interesting. Thanks for being interested in my self-driving car experience. Yeah.

[00:26:50] John Nash: Definitely. I'll have to, uh, have to check that out next time I'm in a community that has that, you know, I got to say, I was born and raised in the San Francisco Bay Area. SFO is my home airport. If I had to name one, I can't see getting in a self-driving car and going out of SFO. I just don't feel it. It's, it's, it's, yeah. It's interesting. People who've been in and out of SFO probably know what I'm talking about. I won't go into the details, but yeah. That, that seems, that feels different. Your ride sounded very kind of side-streety and calm. I don't know.

[00:27:22] Jason Johnston: Yeah. Yeah. It was much calmer than some, a lot of airports, which may be why they chose that as one of the places, but yeah. Can I tell you my self-driving dream car?

[00:27:32] John Nash: Yes.

[00:27:34] Jason Johnston: It's actually a self-driving RV. To me, it feels like that will be the goal. That's where we get there, where we can load the family into the self-driving RV. Yep. We go to sleep that night, and we wake up in, I don't know, St. Louis to check out the arch. We go to sleep the next night, and we wake up somewhere in the Tetons and do a little hike and then we wake up the next night and we're on the beach or whatever. So, it would have to be electric though, of course, too. Of course. Uh, I think we need some test cases. I think, you know, like, uh, traveling bands might test this first. That's right. In fact, I think Willie Nelson would be an ideal candidate. Oh yeah. Willie Nelson. Neil Young might even make one actually. He might be into making one. That's excellent. I think that's a good idea.

Yeah. Yeah.

Well, great. Great talking to you, John.

Yeah. See you later. Yeah.