Jon Krohn: 00:05
This is Five-Minute Friday on how to get kids excited about STEM subjects.
In the preceding three Five-Minute Fridays, we were joined by Ben Taylor, and this week we are joined by him yet once more for the final time. In the preceding weeks, we talked about how to sell a multimillion dollar AI contract, why CEOs care about AI especially above most other kinds of technologies, and we talked about what we should expect from an AI platform company 10 years from now.
So, super interesting questions. We’re getting more and more forward-looking as we go with these Five-Minute Fridays, and so my final question here for you in this final Ben Taylor special Five-Minute Friday is how do we get kids excited about STEM in general? Maybe AI specifically, but just kind of quantitative skills in general, science, technology, engineering, and mathematics, how do we get kids to be excited about those fields, so that, like we talked about last week, one of those can be the first trillionaire when they create this AI system that kind of gobbles up all of the niche AI systems?
Ben Taylor: 01:27
So, this is also a great question, and hopefully they’ll have an easier time than you and I did, because I think there were too many key learning objectives in front of our careers that were terrible. Like pick your least favorite math class or something. My daughter is actually here in New York right now at a STEM entrepreneur camp for a week, where they’re playing with robots, trying to program them and different things. So that is helpful to try to get some hands-on, but I think ultimately, it will be AI teaching these lessons.
And so, enjoying storytelling more now, I laugh when I go back in time and think has a teacher ever yelled at you and said, “Pay attention”? Yes, everyone.
But now I’d love to go back in time and say in a witty, snarky way that, “You have failed to earn my attention,” which is an impossible task for any adult to do if you have 30 kids, but it’s not impossible if you have an AI system teaching through VR or some type of immersive technology, where it knows if you’re paying attention, it knows if you’re getting fatigued, it knows if you’re getting distracted. And so I think education will be reformed because of AI, where it’s able to teach us all sorts of topics at a much more engaging and accelerated rate. And actually, my sister, she’s a social influencer. Her name is Science Mom, and her, her goal is to teach kids chemistry and biology in a way that is engaging and interesting.
Jon Krohn: 03:01
Cool. I didn’t know that.
Ben Taylor: 03:02
Yeah. Those fields are interesting, but they typically are taught in a way that’s… Think of chemistry. Oh my goodness, could they make it any more boring, when it’s really interesting? And so I think that will give kids the foundation, but I think we’re desperately needing that foundation early. High schools, they don’t require programming as a core foundation, and a lot of these students are clueless about AI capabilities. Like DALL·E 2 is so fun and engaging, why aren’t high school kids playing with it? Or even better, why aren’t they enriching it or extending it, where they’re putting their hands into these models and doing real things?
But the barrier to entry keeps going down, so I’m not that concerned about the new generation being inspired by it, and I think AI is, it’s such a fun career because it’s changing so quickly, and it’s one of those things, it is never boring, but it also is a career that it intersects all other domains. Neuroscience and philosophy, psychology. There’s so many domains hidden in AI, so I feel like it’s the ultimate thing to pursue.
Jon Krohn: 04:14
I mean, we’re probably preaching to the choir a bit here with the SuperDataScience audience, but it is definitely, in my opinion, the most fascinating thing you can be dealing with in our time.
Ben Taylor: 04:23
Yeah.
Jon Krohn: 04:24
Because yeah, it’s rapidly changing, there’s endless things you can be learning in it. You will never be bored in this career, and yeah, it touches on not only different kinds of subject areas, like you mentioned, philosophy and engineering and so on, but it also touches on any imaginable industrial vertical. So whether you’re interested in climate change or transport, or energy, whatever, it doesn’t matter, with AI, you can be making an impact in that particular industrial area.
Ben Taylor: 04:56
Absolutely, and I think that’s why it should be a requirement, because if you want to be a lawyer or a doctor, there’s no job you can come up with, with even a designer. Like let’s say if you wanted to go and truly just be an artist, now with DALL·E 2 and Imagen, some of these things, you would want to use AI for inspiration, you’d want to use it as a creative accelerator. There’s no domain you would go into where you wouldn’t benefit from being familiar with it. So I do see it as a STEM, as a foundation. The last thought is, the other thing that’s hidden in this AI journey is the meaning of life after the singularity, and the answer is 42. So maybe we will get there in our lifetime.
Jon Krohn: 05:42
We already know the answers. It was written by Douglas Adams years ago. Well, so this has all been a very interesting answer to my question, Ben. I’ve loved this conversation that we’ve been having about how AI could be changing education in the future and why AI is obviously the best job ever. But back to my original question about how we can get kids excited about STEM today. So we have things like your daughter is in a STEM camp, that’s cool, that’s obviously, that kind of option, getting kids immersed in STEM. Things like Science Mom, your sister, which I didn’t know about. So there’s these kinds of content creators that kids can be checking out online. Do you have any other, like are there particular games that we can get kids playing? I don’t know. I mean, you have three kids, what else can we do with our kids to get them excited about STEM today?
Ben Taylor: 06:36
I don’t have the best answer for this. I think the listeners would probably have… It’d be interesting to see them engage and kind of throw some suggestions-
Jon Krohn: 06:43
That’s a great idea.
Ben Taylor: 06:44
… into the comments, because I have three kids, but I have not prioritized them learning. Like the STEM camp is really coming from my better half, from my spouse.
Jon Krohn: 06:56
Oh yeah.
Ben Taylor: 06:57
Where you think it’d be coming for me, but it’s not.
Jon Krohn: 06:59
Oh, that’s so funny. Yeah, I mean, that was my assumption. That’s why I picked this question for you.
Ben Taylor: 07:04
Yeah, I’m such a great father that I encouraged my daughter to go to STEM camp. Not really, but I’m here in New York, which I appreciate because I get to spend time with people like yourself, and there’s a lot of really interesting people in the city, so I don’t mind being out here. I think there are some really interesting influencers on YouTube. Backyard Scientist, Mark Rober, where they build some robotics that can have AI. I think there is definitely a need for people that mainly focus on AI education for kids. I don’t know an influencer right now that I could call out, but I think that is because there’s knowledge gaps, and that’s something I’d like to start doing, go and speak to high schools and kind of show them some of the really cool content that can inspire them. Because I speak to colleges, but I don’t speak to high schools.
Jon Krohn: 07:57
Right. Yeah, neither do I. I don’t even know what I would say.
Ben Taylor: 08:00
Well, I think you could show them some of the-
Jon Krohn: 08:02
DALL·E 2.
Ben Taylor: 08:03
Yeah, DALL·E 2, generators, there’s a lot of different things you could show them that I think would surprise them, but then you could also inspire them with applications and healthcare and other things that you and I are actively working on that they would find interesting. Or even these supercomputing nodes. You have these young kids in high school that are building GPU gaming computers. Well, what about the DGX-2 or the A100? Like a $400,000, 32, I don’t know how many GPUs are in there, but like if you could show that physically to these high school kids, they’ve never gone to a supercomputing convention, so for them to see some of the stuff would shock them, and if they thought they could play on it…
It’s all about earning their attention up front, so even as silly as this might sound, demonstrating that you can play some gaming game on the A100, that probably has not been done, because it’s too valuable to play games on it, it might be the right thing to do to capture their imagination and really inspire them, and then you route them into the rational close on, “And this is why you should go into this domain.” But you have a lot of fun at the beginning, you can be silly. You want to give them something they never forget, and so playing Halo, or whatever the cool game is, the using the full capacity of the A100 for like 12K resolution or something silly, could be interesting. Or having a NLP, ASR bot that feels like it’s sentient and the kids have to talk to it.
Jon Krohn: 09:40
Right, right, right.
Ben Taylor: 09:40
And decide if it’s a human. Like if you actually did some Turing test stuff-
Jon Krohn: 09:44
Right, right, right.
Ben Taylor: 09:45
… I think they would, yeah, they would start to ask a lot of questions. They would say, “I can’t believe this is AI.” Like the Google engineer’s transcript. Yes, it’s silly that he thought it was sentient, but you read the transcript and you say, “Well, this is pretty neat.”
Jon Krohn: 09:57
Pretty good, yeah.
Ben Taylor: 09:58
Yeah, it’s pretty good.
Jon Krohn: 10:00
Well, so you said you didn’t have any ideas, but then you just reeled off like a dozen cool ideas for how people could be getting kids excited about STEM. So that was awesome, and I have one last suggestion. You mentioned a couple of cool robotics YouTube channels. I recommend Dave’s Armoury YouTube channel. Dave was a guest on the podcast, episode 529. You can listen to that episode for lots of specifics on how he used deep learning and machine vision systems and lots of other kinds of techniques to create fun robots that do fun tasks in his backyard. And that also I think would be a great channel for kids to watch and get excited about what he’s doing. He’s pretty funny too.
Ben Taylor: 10:44
You just reminded me about a kid, Mike Wimmer. Who’s 12, he was on The Ellen Show. He’s doing deep learning.
Jon Krohn: 10:52
Oh.
Ben Taylor: 10:52
Yeah, and he’s 12. I think he might be 13 now, but kids don’t need to wait. There are already kids that are building really interesting deep learning systems now. Hopefully, there’s a lot more in the future.
Jon Krohn: 11:05
Nice. And yeah, so your kid could be a trillionaire thanks to listening to this episode. Thanks, Ben.
Ben Taylor: 11:11
Not that being a trillionaire makes you happy.
Jon Krohn: 11:13
No, they’ll be miserable. All right, that’s it for today’s episode. I hope you enjoyed this series of four consecutive Five-Minute Friday episodes featuring the legendary AI entrepreneur Ben Taylor. We’ve finally wrapped it up. All right, keep on rocking it out there, folks, and I’m looking forward to enjoying another round of the SuperDataScience podcast with you very soon.
