Skip to content
PodSubs
ShowsPeopleGuestsJobsNewsEvents
More
BlogPodCred
Menu
ShowsPeopleGuestsJobsNewsEventsBlogPodCredNetworksSign in
Join nowSign in
  • Shows
  • People
  • Jobs
  • News
  • More
    • Guests
    • Events
    • Blog
    • PodCred
    • Networks
    Join nowSign in

Teaching Python

Teaching Python artwork

Episode 161: Teaching Hard Things Simply

Aug 7, 20261 hr 4 minAudio

In this episode of Teaching Python, Kelly and Julian welcome IBM Distinguished Engineer Jeff Crume to talk about teaching hard things simply. The conversation begins with Jeff’s short, visually driven videos and the studio setup behind them, including the lightboard format, the editing process, and the amount of planning needed to turn a 15-minute explanation into something clear and usable. They then turn to the challenge of explaining complex ideas in a way that fits the audience. Jeff describes how he chooses topics from his work with clients, why he thinks teaching deepens his own understanding, and how he adapts for YouTube, classrooms, and conference talks. He emphasizes brevity, structure, and using visuals so viewers are not faced with a talking head and a blank background. A major part of the discussion focuses on AI, the humanities, and education. Jeff explains why he believes the humanities are essential for understanding meaning, purpose, truth, and context, and why those questions matter when using AI. He argues that AI should be treated as a tool to augment learning rather than something to exclude from classrooms. The conversation also covers cybersecurity and practical AI risks. Jeff discusses passwords versus passkeys, phishing, public chatbots, data privacy, cloud services, and the security concerns around agents and connected tools. He argues for private instances, stronger security practices, and doing security earlier in the process. Near the end, Jeff highlights communication, curiosity, and critical thinking as key skills for students. He also points listeners to IBM SkillsBuild and Coursera for training, and closes by encouraging lifelong learning in a fast-changing field. Special Guest: Jeff Crume.

Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.

Transcript

From the show. Tap a time to play from there.

Track 3Hi, and welcome to Teaching Python Podcast. This is episode 161, 00:00:23.931 --> 00:00:25.660 Teaching Hard Things Simply. 00:00:26.038 --> 00:00:28.876 My name's Kelly Schuster-Paredes, and I'm a teacher who codes. 00:00:29.120 --> 00:00:32.400 And I'm Julian Sequeira, and I'm a coder who coaches. 00:00:33.201 --> 00:00:38.130 I'm so excited. We got that right. It's like a thing that we always get scared about it when we come. 00:00:38.507 --> 00:00:41.501 So today, I'm really excited. I'm going to skip over all that fun stuff. 00:00:41.501 --> 00:00:47.441 Today, I'm really excited. We have Jeff Kroom here, and he's an IBM Distinguished 00:00:47.441 --> 00:00:52.881 Engineer, and he's pretty much like a person I stalk online because he has all 00:00:52.881 --> 00:00:55.281 these great videos that I can easily share with teachers. 00:00:55.281 --> 00:00:59.251 And like we were just talking before the show, they're all under like 15 minutes, 00:00:59.251 --> 00:01:02.116 so perfect for teacher bits. Welcome, Jeff. 00:01:03.156 --> 00:01:08.133 Thank you. Thank you. And I'll say right up front that a bit of disclaimer so 00:01:08.133 --> 00:01:11.983 that I stay out of IBM jail, that the views expressed are my own. 00:01:12.423 --> 00:01:16.523 So don't blame IBM or NC State University where I'm an adjunct professor. 00:01:16.523 --> 00:01:18.943 Neither one of them are responsible for it comes out of my mouth. 00:01:19.383 --> 00:01:20.620 Who is responsible, Jeff? 00:01:22.023 --> 00:01:26.363 I don't. I would say me, but people who know me know that I'm not a responsible 00:01:26.363 --> 00:01:28.123 individual. So I guess no one is. 00:01:28.563 --> 00:01:31.523 Perfect. You're going to fit right in with this conversation. 00:01:31.863 --> 00:01:34.837 I have a feeling we're going to have a great time. 00:01:35.863 --> 00:01:36.833 We kick the adults out. 00:01:36.833 --> 00:01:38.303 We kick the adults out, exactly. 00:01:38.303 --> 00:01:38.790 That's right. 00:01:39.383 --> 00:01:43.063 So before we start, because I always do this, so we get off tangent, 00:01:43.063 --> 00:01:46.553 I want to start with the wins of the week. And wins of the week are anything 00:01:46.553 --> 00:01:52.519 that happened at home, at school, at work, in the countryside. 00:01:52.676 --> 00:01:56.523 And we're always going to start with the guest. So, Jeff, go ahead. 00:01:56.903 --> 00:02:00.443 Okay. Well, I had a cool last seven days. 00:02:00.763 --> 00:02:06.893 I went to Austin, Texas, where I did a presentation at a customer seminar and 00:02:06.893 --> 00:02:09.543 then got to go to Cartagena, Colombia. 00:02:10.223 --> 00:02:14.623 I've been to Colombia a few times before, but never to Cartagena and a very beautiful city. 00:02:15.023 --> 00:02:20.892 Got to do a keynote on the topic of AI and jobs. Will AI take or make jobs? 00:02:21.328 --> 00:02:25.345 And the answer is definitely yes, it will. 00:02:27.403 --> 00:02:28.583 I bet you that went over well. 00:02:29.603 --> 00:02:31.249 Just yes. Yes. 00:02:31.428 --> 00:02:36.543 Yeah. And then I was done. And so that was easy work. 00:02:36.923 --> 00:02:41.273 That's crazy. I wish I could get to go to Cartagena. I actually went there one 00:02:41.273 --> 00:02:44.573 time. I loved the Wall City, but no one ever invites me. We have to go to things 00:02:44.573 --> 00:02:47.703 like Orlando to talk. It's not very fun. 00:02:49.836 --> 00:02:51.502 Uh, Julian, you want to go next? 00:02:51.717 --> 00:02:55.452 Yeah, sure. Um, all right. So I, I have a couple of wins, but the, 00:02:55.452 --> 00:02:57.542 the good one with the kids, uh, 00:02:57.542 --> 00:03:01.672 my kids got their mid-year results in for school and they all smashed it. 00:03:01.672 --> 00:03:06.562 So one of them, we've got a bunch of rewards, sorry, awards at a school ceremony, 00:03:06.562 --> 00:03:10.635 uh, yesterday, which was very proud, a very proud moment for us. So that was good. 00:03:11.093 --> 00:03:16.162 Um, and on a personal note for me with, with work, things are changing and going very quickly. 00:03:17.202 --> 00:03:21.122 On one hand, I start a brand new CTO gig on Monday. 00:03:21.502 --> 00:03:26.477 So I'll be the CTO of a financial company starting on Monday, which is very exciting. 00:03:26.819 --> 00:03:29.822 And on the other hand, people in my community have started reaching out for 00:03:29.822 --> 00:03:34.462 help. So yesterday, I was with an elderly lady, 82 years old, 00:03:35.442 --> 00:03:39.758 walking her through password management. She's hard of her sights going. 00:03:40.141 --> 00:03:44.031 And so I you know working through the accessibility stuff but also um, 00:03:45.002 --> 00:03:49.792 educating at the same time that's the relevance to this conversation um and 00:03:49.792 --> 00:03:56.592 answering her ai questions all at once so it was a it's been a good good period of time you're. 00:03:56.592 --> 00:03:57.442 Doing the lord's work. 00:03:57.442 --> 00:03:58.402 And if you watch. 00:03:59.042 --> 00:04:03.612 If you watch any of jeff's I think I feel like jeff is going to say something like pass key not. 00:04:03.612 --> 00:04:09.142 Not I was doing my best you could you could see me biting my tongue why would. 00:04:09.142 --> 00:04:11.442 You give an 80 year old woman a password. 00:04:11.862 --> 00:04:12.922 Yeah yeah. 00:04:12.922 --> 00:04:13.752 Passwords what are. 00:04:13.752 --> 00:04:18.602 Those my my first thing was just not in a notebook okay let's just get it off 00:04:18.602 --> 00:04:22.522 this one piece of paper in front of you that's that's the the bar small wins 00:04:22.522 --> 00:04:23.662 yeah exactly understand, 00:04:24.522 --> 00:04:28.592 you are you know what though jeff his thing she she did ask about pass keys 00:04:28.592 --> 00:04:32.422 so i want to hear about your um you're gonna have to talk about that in a minute then. 00:04:32.422 --> 00:04:33.232 That means we're making. 00:04:33.232 --> 00:04:37.203 Progress. Yeah, exactly. All right, Kel, go for it. 00:04:38.202 --> 00:04:43.222 So, I've had a really fun summer. I've done a couple of presentations, 00:04:43.222 --> 00:04:45.829 but last week, a girl from my gym, 00:04:46.485 --> 00:04:49.970 asked me on a friendly if, and I kind of, 00:04:50.885 --> 00:04:55.090 replicated the friendliness, I needed to work out some ideas because I'm doing 00:04:55.090 --> 00:05:02.530 a presentation for a group called Crew, which is Construction Real Estate Women in Palm Beach County. 00:05:02.808 --> 00:05:05.960 And I was trying to, I was a little bit nervous because it's a pretty big crowd. 00:05:05.960 --> 00:05:09.270 It was like 45 people and we're talking about AI and just how, 00:05:09.690 --> 00:05:13.020 you know, AI can be used as a colleague in the real estate market, 00:05:13.020 --> 00:05:15.397 just kind of helping out some of these companies and firms. 00:05:15.786 --> 00:05:19.830 So my friend at the gym, she's like, hey, come practice with us. 00:05:20.150 --> 00:05:24.806 So I went to her real estate office and it was a really fun time. 00:05:24.936 --> 00:05:29.450 And being able to present to people when I don't know really anything about 00:05:29.450 --> 00:05:33.650 real estate or finance, and I'm just talking about AI, it was just kind of, 00:05:33.650 --> 00:05:35.290 it was a nice feeling that. 00:05:37.445 --> 00:05:42.810 People in industry are experiencing stuff that we've already experienced in 00:05:42.810 --> 00:05:45.490 schools like three years ago and so, 00:05:46.270 --> 00:05:49.530 feeling like the expert in the room even though i didn't know the the subject 00:05:49.530 --> 00:05:54.890 area was just uh another one of those you know wins and here we go so it's kind 00:05:54.890 --> 00:05:58.908 of it was a good win yeah that's cool very nice yeah all right, 00:06:00.393 --> 00:06:04.720 all right let's get to the fun stuff so um i'm just gonna tell a little story 00:06:04.720 --> 00:06:08.870 so i've been you know silently stalking and learning from Joss' videos, 00:06:10.010 --> 00:06:11.280 sharing the little tidbits. 00:06:11.280 --> 00:06:13.490 The first one I shared with the teachers was... 00:06:15.622 --> 00:06:20.288 This video, I think it was where you defined deep learning, AI, 00:06:20.288 --> 00:06:23.708 machine learning. And Jeff, I don't know if you've ever watched any of his videos. 00:06:23.708 --> 00:06:29.668 He has this really cool see-through screen and he writes everything. I want one. 00:06:30.298 --> 00:06:31.538 Yeah, very cool. 00:06:32.032 --> 00:06:38.558 But it's very simple and well-explained and, like I said, short tidbits. 00:06:39.018 --> 00:06:45.638 And so I've been sharing out some of his knowledge pieces to teachers and just 00:06:45.638 --> 00:06:49.148 loving the stuff. And that's where I'm starting with. There's no question. 00:06:49.148 --> 00:06:53.970 I'm going to let you tell us a little bit about everything that you've done. 00:06:54.318 --> 00:06:57.025 And then Julian will go for there. 00:06:57.193 --> 00:06:57.358 Okay. 00:06:57.358 --> 00:07:00.488 You bet. Well, first of all, thank you. Thank you for watching. 00:07:00.488 --> 00:07:01.908 Thank you for the kind words. 00:07:03.168 --> 00:07:09.138 There's, to me, this is a lot of fun. And I know it probably people leave with 00:07:09.138 --> 00:07:14.506 the impression because I'm such an amateur at this that I'm probably doing this in my basement. 00:07:14.924 --> 00:07:18.358 But actually, there's a professional operation behind me supporting me. 00:07:18.358 --> 00:07:22.848 So I go out to the, I live in Raleigh, North Carolina, and I go out to the IBM 00:07:22.848 --> 00:07:26.604 office at Research Triangle Park just about every Friday morning. 00:07:27.051 --> 00:07:34.673 And I record a new video, 10 to 12, every Friday, unless there's some major reason not to. 00:07:34.975 --> 00:07:40.828 So I'm out there doing those. And we have a studio, and it's got professional 00:07:40.828 --> 00:07:46.078 sound, lighting, cameras, all kinds of stuff in there. 00:07:46.078 --> 00:07:48.878 I mean, it's probably $30,000 or $40,000 worth of equipment in there. 00:07:48.878 --> 00:07:51.391 And then we have professional editors that fix things. 00:07:51.967 --> 00:07:55.112 Things and try to make it look presentable because, you know, 00:07:55.112 --> 00:07:58.592 I mean, talk about Mission Impossible. They've got me as source material and 00:07:58.592 --> 00:08:02.770 they have to put enough lipstick on that pig to make it work. 00:08:03.687 --> 00:08:08.452 So that's what we do. Now, what I basically do is I come up with the subjects. 00:08:09.052 --> 00:08:11.352 Every once in a while, someone in IBM will come to me and say, 00:08:11.352 --> 00:08:14.373 hey, we really need you to do this topic or that topic. 00:08:15.152 --> 00:08:19.032 But for the most part, I come up with the subjects myself because Because in 00:08:19.032 --> 00:08:23.062 my day job, my primary responsibility is working with clients throughout the 00:08:23.062 --> 00:08:27.475 Americas, North and South. So that's my whole coverage area is that half of the world. 00:08:27.957 --> 00:08:34.232 And I will run across things that they are curious about, that they don't understand, 00:08:34.424 --> 00:08:37.703 that I can see they're struggling with. And then that often becomes. 00:08:39.156 --> 00:08:43.920 A food for thought for me, oh yeah, maybe I don't understand it either. Maybe I need to learn. 00:08:44.170 --> 00:08:46.616 Well, one of the best ways to learn something is to teach somebody, 00:08:47.096 --> 00:08:50.056 because in order to teach somebody, you have to learn it yourself. 00:08:51.436 --> 00:08:55.786 In learning it deeply enough to be able to explain it, then it causes you to 00:08:55.786 --> 00:09:01.156 dig deeper. So that's, I look at, for me, number one benefit I get from doing 00:09:01.156 --> 00:09:03.848 these videos is it's my own learning. 00:09:04.776 --> 00:09:09.636 It forces me to develop, you know, a structure around the ideas so that I can 00:09:09.636 --> 00:09:13.296 explain them in a way that hopefully a lot of people understand. 00:09:14.605 --> 00:09:17.436 And then there's the number one question that we always get, 00:09:17.436 --> 00:09:22.046 if you've ever seen any of these videos, as Julian told you, 00:09:22.046 --> 00:09:25.906 you know, there's a glass. It looks like I'm riding backwards in the air with 00:09:25.906 --> 00:09:27.828 my left hand in a dark room. 00:09:28.374 --> 00:09:30.423 And none of those things I just told you are true. 00:09:31.336 --> 00:09:35.556 The room's very bright, at least in terms of the lights shining in my eyes. 00:09:35.556 --> 00:09:36.916 I feel like I need sunglasses. 00:09:37.976 --> 00:09:40.876 The background is black, though, so it looks like the room is dark. 00:09:43.016 --> 00:09:48.476 If I write in the air, you can see nothing happens. So I'm obviously not writing 00:09:48.476 --> 00:09:51.616 in the air. There is a glass in front of me, and we use special markers. 00:09:52.136 --> 00:09:58.116 And there's a black light that's built into the glass that goes around the rim of the glass. 00:09:58.796 --> 00:10:03.276 And then these markers fluoresce in the black light. So that's what gives it 00:10:03.276 --> 00:10:05.536 that visual pop that you see. 00:10:06.696 --> 00:10:09.691 But the main thing, the number one question I always get is, 00:10:09.796 --> 00:10:11.536 how did you learn to write backwards? 00:10:14.660 --> 00:10:19.676 And so many people, what kills me is the ones who write in the YouTube comments. 00:10:20.156 --> 00:10:26.026 It's like, one of them I always remember, dude, I can't even concentrate on 00:10:26.026 --> 00:10:27.876 what he's saying. He's writing backwards. 00:10:29.276 --> 00:10:30.311 I'm like, okay, well... 00:10:30.897 --> 00:10:35.768 Sorry to let you down. I can barely write forwards. So I'm definitely not writing 00:10:35.768 --> 00:10:38.147 backwards. And I'm definitely not left-handed. 00:10:38.496 --> 00:10:40.324 So there should be your clue. 00:10:40.678 --> 00:10:44.308 And a lot of people will write in comments like, it looks like, 00:10:44.308 --> 00:10:47.878 because I'm not the only one recording these videos. They go on the IBM technology 00:10:47.878 --> 00:10:49.769 channel. So we have a lot of other IBMers on there. 00:10:50.100 --> 00:10:52.938 And somebody said, it looks like IBM only hires left-handed people. 00:10:53.818 --> 00:10:58.448 And I'm like, okay, that should be your clue. Yeah, that should be the clue 00:10:58.448 --> 00:11:02.728 that tells you, yeah, what we do is we record it. It looks backward if you're 00:11:02.728 --> 00:11:05.476 in the studio looking at this from the other side of the glass. 00:11:05.668 --> 00:11:08.458 And then the editor takes the video and then just flips it like this. 00:11:10.018 --> 00:11:15.148 And now magicians never supposed to tell, you know, how their magic is done. 00:11:15.148 --> 00:11:18.418 Now, you know, the secret and now you'll be able to, you know, 00:11:18.838 --> 00:11:21.198 no big deal. You know, that guy can't write backwards. 00:11:22.758 --> 00:11:26.388 But I did a video where I explained how we make the videos. 00:11:26.388 --> 00:11:26.578 Okay. 00:11:26.938 --> 00:11:31.078 And I love the comments on that. One after another is like, finally, I can sleep. 00:11:31.958 --> 00:11:36.418 This was such relief for people who had been working on this in their head. 00:11:36.758 --> 00:11:40.838 Smoke, mirrors, no, there's no smoke, no mirrors. Just a simple editing trick. 00:11:41.238 --> 00:11:45.118 And then that's how it comes out. So now you know how the sausage is made. 00:11:45.118 --> 00:11:49.058 I thought you had that board. I actually contacted the people where you write. 00:11:49.058 --> 00:11:52.768 And during the videos, it does like the reversed vision on there. 00:11:52.768 --> 00:11:56.018 Here, I thought you had like the $20,000 board. Now I have to tell everybody 00:11:56.018 --> 00:11:58.207 because I went to school and I said, I want one of these. 00:11:58.474 --> 00:12:01.488 If you give me one of these, my trainings are going to be better. I promise you. 00:12:01.518 --> 00:12:04.668 That's right. That's right. Well, there's something. It's really interesting 00:12:04.668 --> 00:12:10.388 there. We have the, I'll say the old fashioned board, the analog board in our 00:12:10.388 --> 00:12:15.720 studio, which is a piece of glass. And we're writing on the markers, aren't it with markers. 00:12:16.127 --> 00:12:22.148 There is a new technology that I have seen. In fact, we had an IBM Tech Exchange 00:12:22.148 --> 00:12:25.189 conference last year in Orlando, and I got to try this out. 00:12:25.410 --> 00:12:29.078 Digital boards, where you don't use markers, you use a stylus. 00:12:29.718 --> 00:12:31.656 And the stylus, as you're drawing... 00:12:32.283 --> 00:12:36.389 You know, you can just choose what color and the stylus will draw it in that color. 00:12:36.969 --> 00:12:43.329 You can also have animations that you put in and all that's done not during 00:12:43.329 --> 00:12:47.968 the post-production, which is what we have to do whenever you see animations in my videos. 00:12:48.473 --> 00:12:52.899 I'm pointing at stuff that's not there and having to pretend it's there and 00:12:52.899 --> 00:12:57.409 hoping the editor doesn't leave me looking foolish when I point at something 00:12:57.409 --> 00:13:00.157 that's not there. And every once in a while they do. 00:13:01.742 --> 00:13:06.829 So I love the digital boards and they're, you know, probably $30,000 or whatever, 00:13:06.829 --> 00:13:10.789 like you're talking about. So they're a lot. The analog boards are not as much. 00:13:11.109 --> 00:13:14.609 I mean, they're, I don't know, I don't know for sure. It depends on the size, 00:13:14.969 --> 00:13:17.329 but it's a fraction of that cost for sure. 00:13:18.349 --> 00:13:22.869 You know, less than $10,000 for most of them. But here's the downside of the 00:13:22.869 --> 00:13:27.889 analog boards that people don't think about is that there's one thing you never 00:13:27.889 --> 00:13:30.879 see us do in those videos. Erase. 00:13:30.879 --> 00:13:31.009 Mm. 00:13:32.142 --> 00:13:36.071 Okay, this is not like a dry erase board where, you know, you fill up the space 00:13:36.071 --> 00:13:40.623 and then you can just erase it and then you can reuse the space. No, this is a glass. 00:13:40.925 --> 00:13:44.781 So when you try to erase on it, it smears. 00:13:44.781 --> 00:13:45.231 Smudges, yeah. 00:13:45.631 --> 00:13:48.431 Then you get to erase it again. Then you get to erase it again. 00:13:48.751 --> 00:13:51.911 It's like, you know, five minutes of erasing to clear the board. 00:13:52.651 --> 00:13:55.771 And nobody wants to watch you do that on camera. 00:13:55.771 --> 00:13:56.011 Yeah. 00:13:56.351 --> 00:14:01.061 So that's why I don't recommend those for live presentations. 00:14:01.061 --> 00:14:01.391 Yeah. 00:14:02.077 --> 00:14:05.007 You know, you would have the problem, you need to flip it, but, 00:14:05.315 --> 00:14:07.585 you know, that can be done with the camera. 00:14:07.846 --> 00:14:14.101 But the digital board would allow you to do those kind of things and do them more in real time. 00:14:14.101 --> 00:14:17.371 And then the other thing that's different is, you know, if you think about the 00:14:17.371 --> 00:14:22.061 board space is this big, well, I'm taking up a third of it with me being in 00:14:22.061 --> 00:14:26.251 whatever section I'm in. And you don't want to write over your face. That looks really dull. 00:14:26.751 --> 00:14:31.221 And that's a rookie mistake. You'll see the people who are rookies writing on 00:14:31.221 --> 00:14:33.731 their face because to them, they're just looking at it. 00:14:34.231 --> 00:14:37.071 And you do that on a regular dry erase board. 00:14:37.574 --> 00:14:40.421 But on these, you want to use the margins, you know, the sides, 00:14:40.421 --> 00:14:45.241 or either you stand over to the side and put it all here. So there's a whole lot I've learned. 00:14:45.241 --> 00:14:45.751 Good Lord. 00:14:45.751 --> 00:14:47.951 And a lot of mistakes I've made along the way. 00:14:48.491 --> 00:14:50.071 Not to mention all the topics. 00:14:51.191 --> 00:14:51.451 Yes. 00:14:53.091 --> 00:14:58.301 When you're thinking about teaching these difficult topics, because you go from 00:14:58.301 --> 00:15:01.391 everything. I know you're an expert in cybersecurity, is that correct? 00:15:01.391 --> 00:15:02.731 Yeah, that's my main thing. 00:15:02.911 --> 00:15:08.411 But then you have to talk about everything else that encompasses tech. 00:15:08.951 --> 00:15:12.681 I've seen you've talked about quantum computing and everything. 00:15:12.681 --> 00:15:13.023 Yeah. 00:15:13.371 --> 00:15:17.447 When you go to build out these, and this is, I guess, like a really... 00:15:18.253 --> 00:15:23.979 Hard thing for teachers, is trying to tease out what's the most important, 00:15:25.459 --> 00:15:29.529 what is it that they really need to know, and how can you get it in a place 00:15:29.529 --> 00:15:32.039 where everyone's going to understand? 00:15:32.719 --> 00:15:39.339 What's your trick? Because I think you are quite the experienced educator when it comes to that. 00:15:40.659 --> 00:15:44.999 Well, thank you. I'll tell you, there's a lot of learning I had to do, 00:15:45.539 --> 00:15:49.159 and other people taught me, and then I learned from mistakes that I made as 00:15:49.159 --> 00:15:54.539 well, that it's different when you're in YouTube land than when you're in the 00:15:54.539 --> 00:15:58.883 classroom or when you're at a conference presenting. And I do all of those. 00:16:00.179 --> 00:16:03.519 We're used to, you know, if you've got a classroom, you've got a captive audience. 00:16:03.959 --> 00:16:05.553 So they kind of have to be there. 00:16:06.643 --> 00:16:10.479 If you're in YouTube land, you've got about 15 seconds. 00:16:10.479 --> 00:16:14.989 And if you don't grab them, They're out the door Onto the next video So one 00:16:14.989 --> 00:16:18.219 of the things That you can't do And I see this as a rookie mistake People will 00:16:18.219 --> 00:16:21.059 start these videos And they're standing there And they're introducing their 00:16:21.059 --> 00:16:25.108 topic And just talking So then it's just a talking head And a black background, 00:16:25.760 --> 00:16:30.335 Again, you're losing people. They are not walking. They are running for the 00:16:30.335 --> 00:16:34.676 exits at that point because YouTube land is short attention span. 00:16:35.465 --> 00:16:40.905 So that's the thing I had to learn to adjust on is that I need to get something 00:16:40.905 --> 00:16:43.085 happening on the board, something for people to look at. 00:16:43.525 --> 00:16:46.575 They don't want to just look at this. I mean, nobody wants to look at this. 00:16:46.802 --> 00:16:50.675 So I need to give them something else to distract them and, you know, 00:16:50.675 --> 00:16:56.735 watch this, you know, give them some bright, shiny object to engage them and 00:16:56.735 --> 00:16:58.510 then try to keep things moving. 00:16:58.852 --> 00:17:02.265 Because the videos, some of these are less than 10 minutes. 00:17:03.085 --> 00:17:06.495 I try not to ever get them beyond 15, but sometimes, you know, 00:17:06.495 --> 00:17:09.058 every once in a while they'll go longer. But, you know, if you're in a classroom, 00:17:09.325 --> 00:17:11.895 you've got them for an hour or two hours or something like that. 00:17:11.895 --> 00:17:16.325 So when I'm teaching at NC State University, I mean, I've got the hour. 00:17:17.185 --> 00:17:24.015 Or longer. And my approach is different, where you don't just take your PowerPoints 00:17:24.015 --> 00:17:27.725 or your slides and then put them onto that format. 00:17:28.125 --> 00:17:31.610 It doesn't work that way. So there's a lot of rethinking you have to do. 00:17:32.385 --> 00:17:35.365 One of the things that occurred to me as saying, I didn't come up with this, 00:17:35.365 --> 00:17:39.465 but I really love it, because we're trying to go for brevity, 00:17:39.465 --> 00:17:44.465 but still have something that as substantive is that if I'd had, 00:17:44.465 --> 00:17:46.905 it would have, I'd have made it shorter if I'd had more time. 00:17:48.225 --> 00:17:50.162 Which seems counterintuitive, of course. 00:17:50.676 --> 00:17:55.795 You know, I can put together a one-hour PowerPoint presentation with less time 00:17:55.795 --> 00:17:59.745 than it takes me to put together a 10-minute lightboard presentation. 00:17:59.745 --> 00:18:01.105 Lightboards are what we call those things. 00:18:01.696 --> 00:18:04.845 Because I got to figure all the things I'm taking out and what's going to be 00:18:04.845 --> 00:18:09.125 the flows and what's going to be the color of markers I use and how am I going to lay them out? 00:18:09.125 --> 00:18:13.785 Because I don't want to spend time, you know, even little things like capping 00:18:13.785 --> 00:18:17.834 and uncapping the markers that we've all been taught to do so they won't dry out. 00:18:18.293 --> 00:18:20.145 Nobody wants to listen to you go, hurry. 00:18:21.405 --> 00:18:24.905 As you're doing that with these markers. So I always tell the rookies, 00:18:24.905 --> 00:18:29.082 take the caps off and put them somewhere else so you're not even tempted to do that. 00:18:29.433 --> 00:18:33.155 And line your markers up in the order in which the colors you're going to use 00:18:33.155 --> 00:18:36.125 them and try to make the colors mean something. 00:18:36.865 --> 00:18:40.915 And sometimes I'll change in my script and say, okay, no, I need to switch to 00:18:40.915 --> 00:18:45.705 a different color because I've overused that one. Or here I want red to mean 00:18:45.705 --> 00:18:47.545 this and green to mean that, this sort of thing. 00:18:48.065 --> 00:18:51.975 So there's a lot that goes into it. But for me to do a 15-minute video, 00:18:51.975 --> 00:18:57.065 usually two to three hours, I would say on average, sometimes longer, 00:18:57.405 --> 00:19:02.182 of just trying to come up with a general script and outline for a 15-minute video. 00:19:03.005 --> 00:19:09.618 Yeah, that makes sense. Well, I think it was the most recent one on the humanities that you put out. 00:19:10.485 --> 00:19:12.965 I watched that last night because Kelly sent it to me. 00:19:12.965 --> 00:19:15.285 Oh, good. So I found somebody that did. 00:19:15.665 --> 00:19:16.265 I loved it. 00:19:16.785 --> 00:19:17.305 You're the one. 00:19:17.305 --> 00:19:20.105 Yeah, I was the one who said, how are you writing backwards? 00:19:21.125 --> 00:19:22.225 Yeah, right, right, right. 00:19:22.225 --> 00:19:24.715 Actually, that was the first video, because I watched a few of your videos now, 00:19:24.715 --> 00:19:27.215 and that was the first one where I kind of went, wait a minute, 00:19:27.215 --> 00:19:28.234 did he just ride backwards? 00:19:29.165 --> 00:19:32.865 Yeah. That was the first one that even triggered. But no, I think, 00:19:34.114 --> 00:19:38.166 like, that one was so, it hit home a lot with me because, you know, 00:19:38.485 --> 00:19:42.182 last week I gave an AI presentation down in Sydney, and... 00:19:43.099 --> 00:19:45.926 This is the kind of stuff I brought into it, where I said, look, 00:19:46.094 --> 00:19:51.219 the human aspect, the experience, the lived experience, our understanding of 00:19:51.219 --> 00:19:54.289 literature and just human history and knowledge, our, 00:19:55.208 --> 00:19:58.759 experience as humans is not replicated by AI. 00:19:58.759 --> 00:20:02.509 It's not there. It's just this database of information, right? 00:20:02.509 --> 00:20:07.729 So I won't repeat what you said, but it was just such a enlightening, 00:20:08.569 --> 00:20:13.389 well put together talk that part of me was like, I should have just shown them 00:20:13.389 --> 00:20:19.229 this last week because it was exactly what I try to articulate, but I couldn't. 00:20:20.827 --> 00:20:25.794 First of all, have you had any feedback on that video specifically? 00:20:25.794 --> 00:20:31.044 Because that was one that wasn't overly technical and more about our experience 00:20:31.044 --> 00:20:34.564 as human beings. And I loved it for that reason, because I'm the mindset guy. 00:20:34.564 --> 00:20:37.954 But yeah, so tell me, any feedback on that one? 00:20:37.954 --> 00:20:42.410 Yeah, yeah. And I'll even tell you the story behind why I decided to do that one in particular. 00:20:43.434 --> 00:20:46.894 So you're right. My background is cybersecurity. That's my PhD. 00:20:47.574 --> 00:20:52.114 And that's what I do in my day job. It's what I teach. I've also had an abiding 00:20:52.114 --> 00:20:56.414 interest in AI ever since I was an undergrad student back in my computer science 00:20:56.414 --> 00:20:59.714 days, a thousand years ago, writing my dinosaur to class. 00:21:00.174 --> 00:21:06.614 So as AI finally became real, I was like, wow, at long last, 00:21:07.074 --> 00:21:12.684 the thing that is so fascinating to me is AI and cybersecurity. 00:21:12.684 --> 00:21:15.814 So a lot of times I'll do videos that talk about the cross-section, 00:21:16.134 --> 00:21:17.832 the confluence of those two topics. 00:21:18.057 --> 00:21:24.174 But there's enough things in AI that, to me, I thought a lot of people are missing the point on. 00:21:25.394 --> 00:21:30.074 I love what AI can do, and I hate what AI can do because it can do some amazing 00:21:30.074 --> 00:21:31.942 things and it can do some awful things. 00:21:33.014 --> 00:21:37.097 Any dual-use technology is like that, can do good, can do bad. 00:21:37.409 --> 00:21:43.914 Well, so I was doing a talk for the University of North Florida University talking 00:21:43.914 --> 00:21:49.711 about AI and jobs and this kind of subject, and in particular... 00:21:50.193 --> 00:21:53.901 Talking about how we need to, as educators, in fact, there's a video I have 00:21:53.901 --> 00:21:55.601 on AI and the future of education, 00:21:56.321 --> 00:22:01.421 where I make a very unpopular stance that instead of us saying, 00:22:01.421 --> 00:22:06.171 keep AI out of the classroom, you know, because the fear is if we let students 00:22:06.171 --> 00:22:09.551 use AI, then they're not going to learn anything. They're just going to copy and paste. 00:22:09.935 --> 00:22:13.141 Valid fear. But we're going to have to get over that. 00:22:13.561 --> 00:22:19.791 We're going to have to adapt to that. Because as I asked my students last semester, 00:22:19.791 --> 00:22:24.101 I said, how many of you, and this is college level, how many of you have a professor 00:22:24.101 --> 00:22:27.321 that ever said you can't use a calculator in class? 00:22:28.633 --> 00:22:32.611 And they all laughed. I mean, nobody does that. Now, if you're in elementary 00:22:32.611 --> 00:22:35.101 school and you haven't shown proficiency 00:22:35.421 --> 00:22:38.343 with arithmetic, absolutely you shouldn't be using a calculator. 00:22:38.661 --> 00:22:42.681 But once you've got proficiency, you should use the tools that are available to you. 00:22:43.226 --> 00:22:48.591 If we require students not to use the tools, the best tools available to them, 00:22:48.591 --> 00:22:50.656 then we're preparing them for the jobs of yesterday. 00:22:51.421 --> 00:22:54.238 And they will not be suited for the jobs of today and tomorrow. 00:22:54.523 --> 00:22:58.197 Because their boss is never going to come to them and say, I want you to do 00:22:58.614 --> 00:23:03.241 X, Y, and Z and don't use AI. No boss is going to say that. 00:23:03.681 --> 00:23:05.656 So why would I as a teacher say that? 00:23:06.921 --> 00:23:11.161 Now, there's a whole lot about how we should use AI in the classroom. 00:23:11.161 --> 00:23:14.723 And a lot of people are doing a lot of thought on this. I'm not saying I have all the answers. 00:23:15.200 --> 00:23:18.881 What I'm convinced of, absolutely convinced, is that we can't say no to this. 00:23:19.461 --> 00:23:23.461 We have to figure out how to incorporate it as a tool, just like we did with 00:23:23.461 --> 00:23:25.463 calculators, just like we've done with a lot of things. 00:23:26.424 --> 00:23:31.330 So I was giving that talk to University of North Florida, and it came to the Q&A session. 00:23:31.810 --> 00:23:34.550 And of course, I've been talking about using AI throughout all of this. 00:23:35.010 --> 00:23:37.350 And someone raised their hand. Now, I'm doing this remotely, 00:23:37.350 --> 00:23:39.636 so I can't really see all that well who it is. 00:23:40.710 --> 00:23:44.652 And she said, I'm an English professor here, and how dare you? 00:23:46.030 --> 00:23:47.330 And I was like, uh-oh. 00:23:48.950 --> 00:23:53.520 She said, why would students ever read these great works of literature if they 00:23:53.520 --> 00:23:56.810 could just go to AI and ask it, you know, essentially for the summary. 00:23:56.810 --> 00:24:00.308 And I was like, okay, all right, all right. Yeah. You raise a good point. 00:24:01.410 --> 00:24:04.250 And I tried to answer it as best I could during, in the moment. 00:24:04.790 --> 00:24:10.617 And I think I gave a passable answer, but what she said really stuck with me 00:24:10.900 --> 00:24:14.630 because I thought, yeah, there is a lot more to this. 00:24:15.250 --> 00:24:18.980 And I can imagine if I was an English teacher, I'd be feeling very marginalized 00:24:18.980 --> 00:24:22.760 in all of this STEM, STEM, STEM for everything, you know, Science, 00:24:22.760 --> 00:24:24.131 technology, engineering, and math. 00:24:24.990 --> 00:24:28.800 That's everything. And then I got to thinking, yeah, you know what? 00:24:28.800 --> 00:24:33.668 AI doesn't know what to do. We tell it what to do. 00:24:34.190 --> 00:24:39.790 I mean, once we tell it, it will go do it. But we decide what and we decide why. 00:24:40.690 --> 00:24:44.046 And why is a question of meaning and purpose. 00:24:44.644 --> 00:24:48.790 And you know what you don't learn in calculus class is purpose. 00:24:49.235 --> 00:24:53.580 What you don't learn in computer science class is meaning. What you don't learn 00:24:53.580 --> 00:24:58.810 in, you know, all of these STEM classes in engineering are the whys. 00:24:59.690 --> 00:25:04.800 So if we're going to guide this AI, we need to be, and now that we have this 00:25:04.800 --> 00:25:08.950 tool that is so powerful, we need to be really careful about the why before 00:25:08.950 --> 00:25:10.818 we turn this monster loose. 00:25:11.390 --> 00:25:16.610 And so my contention, and that's what this video is about, is that the humanities 00:25:16.610 --> 00:25:21.796 have never been more important than they are today because the humanities are where we learn the why. 00:25:22.203 --> 00:25:24.942 You know, where do we learn, first of all, what's true and what isn't? 00:25:25.761 --> 00:25:31.247 Well, that's philosophy. You know, that's where we look at logical proofs and things like that. 00:25:32.107 --> 00:25:36.657 Where do we learn, you know, about man's humanity against man and these kinds 00:25:36.657 --> 00:25:38.550 of eternal struggles? Well, that's in literature. 00:25:39.003 --> 00:25:44.187 You know, where do we learn about the past so that we don't keep making the 00:25:44.187 --> 00:25:46.328 same mistakes in the future? Well, that's history. 00:25:46.828 --> 00:25:51.977 And I could go on. You know, those are all humanities, and those are all incredibly 00:25:51.977 --> 00:25:54.587 important in how we wield this tool. 00:25:55.067 --> 00:25:59.638 So that was what motivated me. So my hat is off to that English professor. 00:25:59.848 --> 00:26:03.227 And I even called her out when I posted to LinkedIn. I said special thanks to 00:26:03.227 --> 00:26:08.346 her for motivating me on this topic because I believe that. Now, 00:26:08.509 --> 00:26:11.603 some of the commenters really liked it, okay? 00:26:12.127 --> 00:26:16.917 Some of them said, oh, great, here we go. Another liberal. We have too many 00:26:16.917 --> 00:26:19.807 of those in our education. I was like, what in the world? 00:26:20.567 --> 00:26:25.357 This was not a political statement. The humanities are something. 00:26:25.357 --> 00:26:28.777 I mean, one guy even made a comment about, yeah, here we go. 00:26:28.787 --> 00:26:31.827 Another God in the gaps argument. That didn't hold up well or something. 00:26:32.207 --> 00:26:37.237 And I said to him back, the reason you even know about that reference is because of the humanities. 00:26:37.237 --> 00:26:39.847 Oh, nice. Excellent. 00:26:40.527 --> 00:26:42.567 So anyway, so many. 00:26:42.567 --> 00:26:47.337 So many thoughts. And I wish I had the whole time just to go into all the things 00:26:47.337 --> 00:26:49.461 in my head. Two things, though. 00:26:51.353 --> 00:26:57.468 AI, I just got sent yesterday or last night, an English teacher sent me a notebook 00:26:57.468 --> 00:27:02.328 LM. It's an idea he's going to try, I guess, and coming up. And he put the telltale 00:27:02.328 --> 00:27:04.698 heart in there. And, you know, and everybody raids this short story. 00:27:05.098 --> 00:27:10.578 And all the English teachers are, like, diving deep into the hidden analogies 00:27:10.578 --> 00:27:14.008 and whatever else is in there, metaphors in this thing. And as a student, 00:27:14.008 --> 00:27:15.498 you're just like, yeah, he murdered somebody. 00:27:15.898 --> 00:27:18.708 But as I was playing with this notebook LM. 00:27:18.708 --> 00:27:19.358 Spoiler alert. 00:27:19.358 --> 00:27:23.638 Spoiler alert. But it was pretty cool because you could ask, 00:27:24.038 --> 00:27:26.918 I don't know what he put in there as the sources because I didn't look, 00:27:27.258 --> 00:27:31.838 but it was really diving into what the meaning and what was the real point that 00:27:31.838 --> 00:27:34.718 English teachers always use this short story. 00:27:34.718 --> 00:27:38.668 So I think there's a lot to be said about things that can help and learn. 00:27:39.278 --> 00:27:41.588 I think that's great. In fact, one of the things when I was talking to that 00:27:41.588 --> 00:27:47.058 English professor, I told her, look, there's some works of literature that I 00:27:47.058 --> 00:27:49.908 think are just, you know, transformational. When I read that book, 00:27:49.908 --> 00:27:52.090 it changed me. I'm not the same person anymore. 00:27:52.500 --> 00:27:56.638 And then there's some others that English teachers made me read that I'll never forgive them for. 00:27:58.098 --> 00:27:59.848 And Moby Dick is one of those. 00:27:59.848 --> 00:28:01.038 But I'm sure if you had... 00:28:01.378 --> 00:28:02.538 I don't think anyone should have to read that. 00:28:02.538 --> 00:28:07.128 I'm sure if you had an AI where you can ask it questions about only Moby Dick, 00:28:07.128 --> 00:28:11.758 you would be able to understand the meaning of life now because of that. 00:28:12.658 --> 00:28:14.088 Exactly. And wouldn't it be 00:28:14.088 --> 00:28:18.084 great to have that tutor help you if you're trying to read through this. 00:28:18.269 --> 00:28:21.708 And it doesn't make any sense. And Shakespeare doesn't make sense to me until 00:28:21.708 --> 00:28:24.365 I get my head in that space. And sometimes I never can. 00:28:24.800 --> 00:28:27.778 If it could, okay, what in the world did he mean by that? Oh, 00:28:27.778 --> 00:28:30.808 okay. Well, this is what those words meant in those days. 00:28:31.691 --> 00:28:36.277 I can't just call up the professor every time I stumble over one of those things. 00:28:36.277 --> 00:28:36.457 Exactly. 00:28:36.917 --> 00:28:40.297 So it could be very helpful and augment that intelligence. 00:28:40.297 --> 00:28:43.943 Hey, I want, I'm sorry, I'm going to jump in, Shilling, because I have all these questions in my head. 00:28:44.251 --> 00:28:48.527 So part of my role now, because it's kind of like a funny to me, 00:28:48.527 --> 00:28:51.797 I don't think it's funny to everybody else, but because I taught Python, 00:28:51.797 --> 00:28:55.652 because I'm a person probably like everyone, like the two of you, 00:28:55.977 --> 00:28:57.237 where I just love learning, 00:28:57.697 --> 00:28:59.537 I got dumped into AI. 00:28:59.537 --> 00:29:02.496 So every day has been like a major learning moment. 00:29:03.042 --> 00:29:07.887 And since you're the cybersecurity guy, and since I talk to a cybersecurity 00:29:07.887 --> 00:29:13.277 team every Friday and we dissect the subprocessors and where's my AI model and 00:29:13.277 --> 00:29:15.817 where my OWASP risks are. 00:29:15.817 --> 00:29:16.877 I'm sorry you have to do that. 00:29:17.237 --> 00:29:20.282 Apologies. But it's actually fun. 00:29:21.097 --> 00:29:24.157 My people are difficult ones to deal with. I'm just going to tell you. 00:29:25.577 --> 00:29:30.557 What do you so and I think this is something that most schools do not have as 00:29:30.557 --> 00:29:36.737 a person that is actually loves talking about these things but what could you say to, 00:29:37.457 --> 00:29:42.287 some educators who don't have any idea about all the cyber security risks what 00:29:42.287 --> 00:29:48.017 are like some top three things that they should look for um, 00:29:49.503 --> 00:29:51.732 Yeah, without expertise. 00:29:52.487 --> 00:29:56.848 Yeah, well, so watch all of my videos. That would be the thing I would say, 00:29:56.848 --> 00:30:00.228 right? That's not self-serving, is it? No, they're good. 00:30:00.908 --> 00:30:02.008 I would second that. 00:30:02.768 --> 00:30:07.428 There you go. No, no, some of them would not be necessarily all that on target, 00:30:07.428 --> 00:30:09.257 but hopefully some of them would be. 00:30:09.375 --> 00:30:12.730 Um, I, I think I look at it this way. 00:30:13.072 --> 00:30:17.068 Um, cybersecurity is a really interesting topic to me, but it, 00:30:17.068 --> 00:30:21.500 I understand it's an acquired taste for others and most just never quite acquire it. 00:30:21.860 --> 00:30:25.518 Uh, but for me, it was, it was always interesting from the start because I like 00:30:25.518 --> 00:30:29.697 taking things apart and knowing how things work and how things cannot work. 00:30:30.039 --> 00:30:33.238 And of course hackers are taking things apart and making them not work. 00:30:33.603 --> 00:30:38.238 So I like the hard challenge of trying to figure out what they're going to do 00:30:38.238 --> 00:30:41.648 and then try to keep them from doing it, which is an even harder problem than 00:30:41.648 --> 00:30:42.810 actually taking it apart. 00:30:44.012 --> 00:30:46.758 But the kinds of things that I think people need to focus on, 00:30:47.106 --> 00:30:49.558 we still have a lot of problems and we mentioned this, you know, 00:30:49.558 --> 00:30:52.609 about pass keys, you know, passwords. 00:30:52.928 --> 00:30:58.568 Everybody's got passwords and I think everybody hates them and I don't blame, 00:30:58.888 --> 00:31:00.086 you know, They're terrible. 00:31:02.068 --> 00:31:05.448 They used to be the least worst alternative, but we have better alternatives now. 00:31:06.628 --> 00:31:09.108 And those things are called pass keys. And if you want to know about it, 00:31:09.108 --> 00:31:10.726 I've got a couple of videos on them. 00:31:10.893 --> 00:31:14.628 But in general, you'll find more and more sites offering these. 00:31:15.128 --> 00:31:18.388 And one of the big reasons, without getting into technical details, 00:31:18.525 --> 00:31:25.330 about why they're useful is that the number one cause of data breaches for the last few years, 00:31:25.868 --> 00:31:31.118 And IBM just came out with a report today on the cost of a data breach for 2026. 00:31:31.118 --> 00:31:35.618 And it was number one in terms of cost of data breach. It was number one in 00:31:35.618 --> 00:31:37.388 terms of frequency, phishing attacks. 00:31:38.165 --> 00:31:43.408 So phishing attacks are largely ways to try to steal your password, 00:31:43.408 --> 00:31:44.562 your login credentials. 00:31:45.629 --> 00:31:49.230 If the best way to make sure your password isn't stolen is to not have one. 00:31:50.049 --> 00:31:55.789 And a passkey is not easily stolen. So this thing is essentially phishing resistant. 00:31:56.199 --> 00:32:00.479 That is one of the biggest things that we could do if we were to get more and 00:32:00.479 --> 00:32:04.729 more people to adopt passkeys and use them in place of passwords. 00:32:04.729 --> 00:32:06.569 And the great thing is you don't have to remember them. 00:32:07.102 --> 00:32:10.950 You know, they get stored on your device and you unlock the device with your 00:32:11.234 --> 00:32:14.829 face print, you know. And I don't know about you, But so far, 00:32:14.829 --> 00:32:17.129 I've never forgotten to take my face with me everywhere I go. 00:32:17.769 --> 00:32:18.439 Happens once in a while. 00:32:18.439 --> 00:32:21.649 Right there on the front of my head. Yeah, yeah, right, right. 00:32:22.869 --> 00:32:29.349 So that's, I think, a really, it's, if you're offered the opportunity and more 00:32:29.349 --> 00:32:33.328 and more retail sites, for instance, are offering this, use it. I think that's a big one. 00:32:33.593 --> 00:32:36.939 So I learned, sorry, Julian, this is going to be all about me and Jeff today. 00:32:36.939 --> 00:32:42.009 Sorry. That's right. I just broke my face. I just broke my face out. Your face is great. 00:32:46.109 --> 00:32:46.550 Sorry. 00:32:46.748 --> 00:32:47.149 You go you go all right. 00:32:47.149 --> 00:32:51.679 All right so i just learned and i don't know if this is a real thing but like 00:32:51.679 --> 00:32:58.665 with it's called what blast radius and it was kind of linked into the idea of the canvas hack, 00:33:00.089 --> 00:33:04.589 and and so my my take is i like we want to put it behind uh, 00:33:05.341 --> 00:33:10.710 behind put our apps and all our stuff behind sort of like a walled vendor that 00:33:11.140 --> 00:33:16.549 doesn't ingest any of the data but just kind of talk you know allows that talking, 00:33:17.398 --> 00:33:24.719 from the app to the LTI or whatever whatever connection they have and um is 00:33:24.719 --> 00:33:29.287 that like a thing now because for example canvas got hacked right a couple yeah 00:33:29.558 --> 00:33:30.789 whatever a couple months ago, 00:33:31.445 --> 00:33:35.353 And with everything that's connected, then they had to worry about these blast 00:33:35.353 --> 00:33:41.153 radiuses. Is that something you think like is going to be more encompassing 00:33:41.153 --> 00:33:42.533 maybe for schools? Oh, yeah. 00:33:42.533 --> 00:33:46.163 We'll see more of that. We've had it ever since we started really moving into 00:33:46.163 --> 00:33:49.564 cloud apps. We've had that as a major exposure. 00:33:50.034 --> 00:33:55.914 And we have it even more now with the public chatbots that most people use. 00:33:56.110 --> 00:33:58.531 And I keep trying to advise everyone. 00:33:58.985 --> 00:34:03.633 Anything you put into a public chatbot, you should consider to be public information. 00:34:04.517 --> 00:34:08.403 And I don't think people have that message yet because they can use what you 00:34:08.403 --> 00:34:10.617 put in to train their models. 00:34:11.058 --> 00:34:13.893 And if you're putting insensitive information into those, you know, 00:34:13.893 --> 00:34:16.863 you say, well, hey, look, I've got a spreadsheet with all my student grades 00:34:16.863 --> 00:34:20.853 and I'm just going to upload that here and have it, you know, munch on it for me. 00:34:21.333 --> 00:34:26.256 Well, okay, if that was okay for you to send out to the world, you're good. 00:34:26.749 --> 00:34:31.953 But my guess is no. So what we need to be doing is having private instances 00:34:31.953 --> 00:34:34.073 of these, and you have to pay extra. 00:34:34.573 --> 00:34:37.233 I mean, those public chatbots are free, so I get that. 00:34:38.493 --> 00:34:44.599 And I use them, but everything I put in them, I'm okay if everybody reads about me. 00:34:44.988 --> 00:34:48.907 You know, like if I'm going to ask it a medical question, it's going to be about my broken shoulder. 00:34:49.249 --> 00:34:53.249 It's not going to be about something I consider to be personal as an example, okay? 00:34:53.615 --> 00:34:58.033 I don't care if anybody knows. I did break my shoulder in January. Okay, fine. 00:34:59.033 --> 00:35:04.393 I put that out there. But those are the kinds of things that we have to be more 00:35:04.393 --> 00:35:08.933 aware of. And therefore, like if you're talking about Canvas or at NC State 00:35:08.933 --> 00:35:12.133 where we use Moodle, we'll be switching over to Canvas, I think, coming up. 00:35:13.733 --> 00:35:17.503 All of those services, we put our data in the cloud and then we don't see what 00:35:17.503 --> 00:35:21.408 happens to it after that. Now it's all trust. And, 00:35:21.929 --> 00:35:26.876 A lot of times that trust is misplaced. A lot of companies will, 00:35:27.296 --> 00:35:30.556 they spend, and this is true of most organizations, they will spend most of 00:35:30.556 --> 00:35:34.631 their effort in putting functionality into an app, into a service. 00:35:35.316 --> 00:35:39.836 And it turns out they really need to be putting at least as much into securing it. 00:35:40.476 --> 00:35:44.056 And they usually don't. Because securing it doesn't sell more licenses. 00:35:44.776 --> 00:35:49.596 There's a financial disincentive to do security right until they get caught. 00:35:50.776 --> 00:35:54.376 And that's what happened. And now all of this happened. But until then, 00:35:54.856 --> 00:35:58.166 you know, it's everybody's riding fast. And, you know, as I said, 00:35:58.166 --> 00:35:59.297 go fast and break things. 00:35:59.586 --> 00:36:01.946 Well, no, please not with my data. 00:36:01.946 --> 00:36:02.696 Yeah. 00:36:02.696 --> 00:36:07.980 So that's a risk. And it's a risk that will continue as we use more and more AI. 00:36:08.305 --> 00:36:12.396 And if you're going to use AI with sensitive information, you need to have your 00:36:12.396 --> 00:36:19.126 own organization's private instance of that so that your data is your data, 00:36:19.126 --> 00:36:21.096 not your data is training their model. 00:36:22.217 --> 00:36:26.066 And you still have to trust that those companies are not using it, 00:36:26.066 --> 00:36:28.466 that their terms and conditions, they're actually sticking to that, 00:36:28.466 --> 00:36:29.856 right? That's exactly right. 00:36:30.336 --> 00:36:33.236 At some point, you have to trust someone. You do. 00:36:33.716 --> 00:36:39.306 To the extent possible that you can have the information encrypted and you have 00:36:39.306 --> 00:36:42.116 the keys, then you don't have to trust quite as much. 00:36:42.476 --> 00:36:46.476 Yeah, that's right. So because, and you mentioned at the start of this with 00:36:46.476 --> 00:36:49.309 your background and what you currently do day to day at IBM, 00:36:49.727 --> 00:36:52.647 you know, you're talking to customers, enterprise clients and everything. 00:36:53.298 --> 00:36:56.374 Based on what you were talking about with cybersecurity and, 00:36:56.374 --> 00:36:58.510 Kelly, what you brought up with the blast radius, 00:36:59.793 --> 00:37:06.864 where are you kind of seeing people stop their cybersecurity efforts when it 00:37:06.864 --> 00:37:08.609 comes to their AI stacks? 00:37:09.504 --> 00:37:12.664 Like it's easy to go okay we need to make sure the logins are secure and all 00:37:12.664 --> 00:37:17.144 that kind of stuff but as you get further down and you have llms having blanket 00:37:17.144 --> 00:37:19.366 access to things or agents having, 00:37:20.022 --> 00:37:25.885 external internet access while also having internal access to databases and everything do you see, 00:37:26.680 --> 00:37:29.914 is there a trend or is there something you see more often where people aren't 00:37:29.914 --> 00:37:31.324 thinking there's a trend okay you're. 00:37:31.324 --> 00:37:33.734 Saying what the trend is they're not doing any. 00:37:33.734 --> 00:37:34.224 Security. 00:37:38.664 --> 00:37:42.558 I hate to be, you know, a doomsdayer, but I'm just telling you the truth. 00:37:42.901 --> 00:37:47.154 I mean, look what happened. There was a story that broke just last week and 00:37:47.154 --> 00:37:51.451 we talked about it on today's podcast on the IBM security intelligence podcast 00:37:51.840 --> 00:37:54.435 about the hugging face attack. 00:37:54.929 --> 00:37:59.804 OpenAI was running one of their cyber-capable models. 00:38:00.124 --> 00:38:05.493 A model is designed to be able to break things and find where breakages are, 00:38:06.264 --> 00:38:09.957 to try to do an exercise. And this agent... 00:38:11.440 --> 00:38:15.390 Kind of went rogue. I mean, it was told what its objectives were and it tried 00:38:15.390 --> 00:38:17.960 to achieve those objectives, even if it meant cheating. 00:38:18.540 --> 00:38:24.350 And that meant cheating meant using a vulnerability in hugging face, 00:38:24.350 --> 00:38:27.551 another company entirely, and breaking into their system. 00:38:28.040 --> 00:38:34.070 So here we had an agent that essentially went rogue and attacked from one legitimate 00:38:34.070 --> 00:38:35.740 company to another legitimate company. 00:38:36.380 --> 00:38:40.850 Now, if that can happen, that's where, you know, okay, ooh, this thing just 00:38:40.850 --> 00:38:43.660 slipped out of the lab, you know, kind of thing. 00:38:44.200 --> 00:38:49.590 But the bad guys have access to this technology too. And agents are becoming 00:38:49.590 --> 00:38:50.661 more and more available. 00:38:51.120 --> 00:38:56.820 Now, I'm not against these things. That's the hard thing is to try to strike the right balance. 00:38:57.280 --> 00:38:59.700 I don't want to tell people don't use agents because they're dangerous. 00:38:59.700 --> 00:39:03.930 I want to tell people they're dangerous and here's how you need to use them if you're going to. 00:39:04.300 --> 00:39:08.517 But agents are built on a lot of assumptions. 00:39:09.120 --> 00:39:14.460 An agent is basically a model that uses tools in a loop autonomously. 00:39:14.773 --> 00:39:18.160 So what could possibly go wrong? Well, models hallucinate. 00:39:18.580 --> 00:39:22.994 Models can be prompt injected. The tools they use could be compromised. 00:39:23.860 --> 00:39:28.747 The tools they use could belong to someone else who is a hacker who's just collecting, 00:39:29.420 --> 00:39:31.060 credentials that you're feeding into them. 00:39:32.140 --> 00:39:35.780 It's doing this in a loop, which means it can do it at machine speed faster 00:39:35.780 --> 00:39:40.380 than you can intervene and it's doing it autonomously. So it's not asking your permission. 00:39:41.390 --> 00:39:42.562 All of that is a, 00:39:43.241 --> 00:39:46.656 If you don't do this right, this is what I call a risk amplifier. 00:39:46.656 --> 00:39:46.946 Yeah. 00:39:47.646 --> 00:39:51.686 And we have to do this right, which has been the subject of a lot of videos 00:39:51.686 --> 00:39:55.106 I've talked about. So I'm preaching on this one, you know, constantly trying 00:39:55.106 --> 00:39:58.020 to make sure we get it right. But I don't think most people are. 00:39:58.298 --> 00:40:03.196 Yeah. I joke and I used to tell our tech director that he was the department 00:40:03.196 --> 00:40:06.386 of no, but I've become the department of no. Because the more I learn, 00:40:06.786 --> 00:40:09.226 the more I'm like, no, we don't need that. No. 00:40:12.266 --> 00:40:16.306 And that's the problem. security people become that. I did a video where, 00:40:18.366 --> 00:40:25.526 I, and this was not my idea. I actually heard a former CIO of the state of North 00:40:25.526 --> 00:40:31.021 Carolina say this. I don't even know where she is now, but she said, don't say no, say how. 00:40:31.805 --> 00:40:35.606 And I like that better. If you say no, people will do it anyway, 00:40:35.606 --> 00:40:37.772 and they'll do it in the most insecure way possible. 00:40:38.125 --> 00:40:44.065 If I say how, then maybe I can get them to do it right. So maybe for instance, I say, 00:40:44.907 --> 00:40:49.487 No to using public chatbots with AI, with confidential information. 00:40:49.907 --> 00:40:54.397 But the how means we've set up a private instance where you can go knock yourself 00:40:54.397 --> 00:40:56.727 out. So everybody use this one instead. 00:40:57.387 --> 00:41:01.337 But if I just say no and don't give an alternative, oh, they're going to do it anyway. 00:41:01.337 --> 00:41:06.297 Yeah, I'm trying that with that whole new Claude release all for K-12 educators. 00:41:06.297 --> 00:41:08.008 They've released a free. 00:41:08.199 --> 00:41:14.827 They said, hey, all K-12 educators, if you give us your driver's license picture 00:41:14.827 --> 00:41:19.947 and prove that you're an educator with your paycheck stub, we're going to give you something. 00:41:19.947 --> 00:41:23.847 It's like these silly requirements. We're going to give you a free. 00:41:23.847 --> 00:41:27.727 Until somebody hacks in and gets the identity database. Thank you. 00:41:28.547 --> 00:41:31.407 That's the one. And then they're like, and it's completely safe. 00:41:31.407 --> 00:41:36.807 We've written it so that you as teachers can put your grade books, your teachers counts. 00:41:36.807 --> 00:41:40.837 But we're doing it for teachers and not for districts. So the teachers came 00:41:40.837 --> 00:41:44.957 out to me. I'm like, listen, no to Claude. And I love Claude. 00:41:44.957 --> 00:41:48.457 And I'm not saying anything about it. I know I'm going to assume positivity 00:41:48.457 --> 00:41:50.623 that they're doing this as a positive thing. 00:41:51.917 --> 00:41:54.836 Just trying to put that out there. in the space. 00:41:55.133 --> 00:41:55.987 I'll give the benefit of the doubt. 00:41:56.247 --> 00:41:56.867 I'm going to give the benefit of the doubt. 00:41:56.887 --> 00:41:58.307 But that doesn't mean they're going to get it right. 00:41:58.787 --> 00:42:02.173 And I said, you know, I'd rather pay 20 bucks a month for $20. 00:42:02.701 --> 00:42:06.939 For every one of you, because they don't even offer that enterprise yet for K-12. 00:42:07.107 --> 00:42:12.665 I'd rather them pay the 20 bucks a month for their own little one that they want. 00:42:13.105 --> 00:42:17.185 And then we worry about it. But it's just, it's become quite a thing. 00:42:17.185 --> 00:42:22.055 And hope that you're not just paying 20 bucks a month for just more tokens in 00:42:22.055 --> 00:42:24.735 the same service with the same level of security. 00:42:24.735 --> 00:42:25.195 Exactly. 00:42:25.195 --> 00:42:28.911 Because that's the other thing you kind of have to vet with these things. 00:42:29.865 --> 00:42:33.265 That, you know, so, I mean, even better still is if you can say, 00:42:33.265 --> 00:42:36.964 look, we're going to have a private instance in our own cloud space, 00:42:37.289 --> 00:42:39.205 and we're going to run those models over here. 00:42:40.245 --> 00:42:41.045 Yeah, you're going to come and 00:42:41.045 --> 00:42:45.373 teach me from IBM Watson. I went to IBM conference in Orlando last year. 00:42:45.606 --> 00:42:49.705 You come and teach me and sit me down with IBM Watson, and I want to build my own instance. 00:42:49.705 --> 00:42:50.905 There we go. 00:42:50.905 --> 00:42:51.885 He can fly me over as well. 00:42:52.345 --> 00:42:54.255 Julian wants to join too. We're going to have her. 00:42:54.305 --> 00:42:57.135 Come on, come on. This year, the conference is in Atlanta. 00:42:57.135 --> 00:43:00.767 Oh, well, might have to go there. Yeah. So... 00:43:01.672 --> 00:43:08.188 Quick question. You're coding language in cybersecurity, or is there like a 00:43:08.188 --> 00:43:11.218 coding back there, or is it consulting more so now? 00:43:12.158 --> 00:43:12.638 Yeah. So, 00:43:14.758 --> 00:43:19.338 it's interesting. After doing 200 videos, basically every question anyone ever 00:43:19.338 --> 00:43:21.368 asks, I can say, I did a video on that. 00:43:22.293 --> 00:43:26.678 And I did a video on that. The most important programming language of the AI era. 00:43:26.998 --> 00:43:27.278 Is English. 00:43:27.338 --> 00:43:27.938 Are you ready for this? 00:43:27.938 --> 00:43:28.388 English. 00:43:29.723 --> 00:43:29.958 It is. 00:43:30.038 --> 00:43:31.118 You said that on last night's video. 00:43:31.178 --> 00:43:33.438 If you think of, yeah, okay. 00:43:33.438 --> 00:43:34.998 We've been watching your videos. 00:43:35.458 --> 00:43:36.168 Okay, somebody. 00:43:36.168 --> 00:43:37.101 We know you, Chip. 00:43:38.038 --> 00:43:42.038 Yeah, great. I mean, think about it. The reason we invented programming languages 00:43:42.038 --> 00:43:47.618 in the first place is because we speak, you know, these high-level natural languages, English. 00:43:48.338 --> 00:43:50.758 Somebody may say, why are you talking English? It could be Spanish. 00:43:50.758 --> 00:43:55.687 Yes, it could. So, we'll use English as the stand-in for this. But natural language. 00:43:57.378 --> 00:44:02.638 Computers didn't understand that. So we invented programming languages as something in between. 00:44:03.098 --> 00:44:09.223 So you translate your thoughts, your intents into instructions, which is code. 00:44:09.520 --> 00:44:11.814 And then the computer knows how to run the code. 00:44:12.410 --> 00:44:17.038 Well, now with AI, generative AI, being able to understand natural language 00:44:17.038 --> 00:44:21.263 natively, well, we can skip that middle step in a lot of cases. 00:44:21.385 --> 00:44:27.458 We can go straight from intent to outcome, intent to results, and not have to code. 00:44:27.858 --> 00:44:31.178 And more and more, and this is the thing, I don't mean to sound like an IBM 00:44:31.178 --> 00:44:35.178 commercial here, but the big emphasis IBM's doing these days is on this product 00:44:35.178 --> 00:44:38.110 we have called Bob, which is a coding assistant. 00:44:38.370 --> 00:44:42.998 But it's more of a coding assistant for enterprise environments and does a whole 00:44:42.998 --> 00:44:46.997 lot more than just write code for you because any old chatbot can do that. 00:44:47.718 --> 00:44:52.337 But that's the future. Now, can it write perfect code? No. 00:44:53.040 --> 00:44:57.248 You know what? Neither can people. So that's not the standard. 00:44:57.701 --> 00:45:03.684 Can we write more code? Yes. Can we generate it? Can we do iterate on it faster? Yes. 00:45:04.084 --> 00:45:08.963 But then we've got to spend more time on making sure it did what we asked. 00:45:09.944 --> 00:45:13.464 And we need to move security. We've been saying we need to shift left. 00:45:13.464 --> 00:45:15.684 The video I'm going to do on Friday is on this topic. 00:45:16.684 --> 00:45:19.944 Shift left security. Get security earlier in the process. 00:45:20.364 --> 00:45:22.744 You know, it should be all the way in the requirements phase. 00:45:22.744 --> 00:45:27.464 But in the development process, don't wait until the end and then do the security testing. 00:45:28.024 --> 00:45:33.064 Start doing that early. And if AI is generating the code, we can use AI to test 00:45:33.064 --> 00:45:35.484 the code as well and look for vulnerabilities. 00:45:35.484 --> 00:45:39.784 Oh, my gosh. You just made me think of another question. So no code, like a no code apping. 00:45:40.104 --> 00:45:46.304 If you're coding with English and you're using, say, like Cloud Code or Lovable or something. 00:45:46.304 --> 00:45:49.864 Although the AI may create code and then the code runs. But in some cases, 00:45:49.864 --> 00:45:51.244 it doesn't have to. It can just do it. 00:45:51.860 --> 00:45:56.914 But if you wanted to build in security in the front, say you're working... 00:45:58.307 --> 00:46:03.304 With a code. That's the thing. Most people don't really know what they're building 00:46:03.304 --> 00:46:08.894 for security when they build with the English or the AI. So they kind of just 00:46:08.894 --> 00:46:09.987 say, hey, make it secure. 00:46:11.394 --> 00:46:13.713 That's like saying, here's my essay, make it better. 00:46:15.394 --> 00:46:17.494 Yeah. Sprinkle magic pixie dust. 00:46:17.494 --> 00:46:17.994 Exactly. 00:46:17.994 --> 00:46:21.929 Yeah. No, that's not going to be enough. I can assure you that. 00:46:22.634 --> 00:46:25.302 But my example is, Like I remember when I was in high school, 00:46:25.598 --> 00:46:31.244 one of the first programming assignments I had was to write a program using 00:46:31.244 --> 00:46:34.474 the basic programming language. There was a programming language called basic 00:46:34.474 --> 00:46:36.442 that nobody uses anymore. 00:46:37.714 --> 00:46:40.116 And find all the number of ways you can make change for a dollar. 00:46:40.840 --> 00:46:42.514 Okay? So that was the assignment. 00:46:42.854 --> 00:46:46.374 Well, I figured out how to do that and wrote that, and it's not a big deal. 00:46:46.734 --> 00:46:47.928 It's actually quite easy. 00:46:48.614 --> 00:46:52.514 Probably 12, 15 lines of code to write something like that. 00:46:53.339 --> 00:46:56.914 Now I can go to a chatbot and say, tell me all the ways to make change for a 00:46:56.914 --> 00:46:58.434 dollar. And it could just spit it out. 00:46:58.834 --> 00:47:02.624 It doesn't have to create code as an intermediate step. So that's what I'm saying. 00:47:02.624 --> 00:47:06.742 In some cases, it can give us the answer, the result without code. 00:47:06.986 --> 00:47:10.594 In other cases, we'll use it to actually write code because we want this to 00:47:10.594 --> 00:47:14.784 be, you know, something that is more predictable and more deterministic than 00:47:14.784 --> 00:47:18.424 a chatbot would be because, you know, how you can ask them the same question 00:47:18.424 --> 00:47:20.790 on different days and sometimes get slightly different answers. 00:47:21.377 --> 00:47:25.174 Sometimes that's charming. Sometimes that's annoying, right? 00:47:25.574 --> 00:47:30.454 Yeah. So, okay. So I've got a question then before we wrap it up. 00:47:30.454 --> 00:47:31.854 I think we're getting close to time. 00:47:33.534 --> 00:47:37.078 With all of this layer with AI and cybersecurity in there as well, 00:47:37.891 --> 00:47:42.067 And you want to move it to the front, which I actually, you can't see this for 00:47:42.067 --> 00:47:46.657 everyone listening, but I was drawing up a process diagram for an app. 00:47:46.657 --> 00:47:47.127 Look at that. 00:47:47.127 --> 00:47:51.357 An app I'm building. And there's the security abstracted, you know, 00:47:51.697 --> 00:47:55.087 out of this, let's just build an agent kind of thing, which is what they said. 00:47:55.087 --> 00:47:55.377 Good for you. 00:47:55.397 --> 00:47:55.937 Good for you. 00:47:56.197 --> 00:47:56.437 And I'm like, no. 00:47:56.437 --> 00:48:00.837 No, we need an MCP server here. I'm going to write very explicit scripts in 00:48:00.837 --> 00:48:01.556 here, you know. Excellent. 00:48:02.237 --> 00:48:06.887 Um, so that's, that's all there, but how would you suggest the people listening 00:48:06.887 --> 00:48:10.897 and learners out there who are thinking, starting to switch onto this and go, 00:48:10.897 --> 00:48:13.777 Oh crap, I didn't really need to think about this stuff as they're seeing more 00:48:13.777 --> 00:48:17.067 breaches and like hearing about the hugging face thing and all of that. 00:48:17.416 --> 00:48:21.037 Um, aside from your channel, which is amazing, by the way, I actually brought it up here. 00:48:21.037 --> 00:48:23.827 So I have the link to your channel. We can put in the show notes after, 00:48:23.827 --> 00:48:29.447 but, um, is there like a formal at this point in time, given this is all relatively 00:48:29.447 --> 00:48:33.607 newish, is there any formal learning people can do for this where they might 00:48:33.607 --> 00:48:35.144 get a cert? Is there anything like that? 00:48:35.505 --> 00:48:42.877 Yeah, yeah, there are some. In fact, there's a skillsbuild.org is actually a website that IBM runs. 00:48:43.277 --> 00:48:48.337 And there's a lot of training on there that we make available to the education market all for free. 00:48:49.197 --> 00:48:52.657 Training in AI, training in cybersecurity, training into a lot of different 00:48:52.657 --> 00:48:55.117 disciplines. So that'd be my first stop. 00:48:55.777 --> 00:49:01.127 Second would be Coursera.com, which they've actually taken some of my videos 00:49:01.127 --> 00:49:02.377 and made them into courses. 00:49:02.377 --> 00:49:02.897 Oh, cool. 00:49:03.237 --> 00:49:07.333 And, you know, people who have a Coursera, you know, subscription or whatever, 00:49:07.669 --> 00:49:11.977 they can, and they added some assignments to it that thankfully I don't have to grade. 00:49:13.237 --> 00:49:17.297 And then after you finish those courses, you can get a completion certificate. 00:49:17.737 --> 00:49:19.877 So that's kind of like the next level up from it. 00:49:20.617 --> 00:49:25.237 So those things are available. And I do think an advantage we have is right 00:49:25.237 --> 00:49:28.657 now with the internet like it is, it's never been easier to learn. 00:49:29.097 --> 00:49:31.348 We have so many resources available to us. 00:49:31.767 --> 00:49:36.495 I learned so much on YouTube myself from other people. 00:49:37.855 --> 00:49:41.975 I look at the IBM videos from my colleagues. I also look at lots of other sources, 00:49:42.555 --> 00:49:45.495 and learn so many things. It's such a great resource, I just think, 00:49:45.495 --> 00:49:47.811 in general. So there's all of that. 00:49:48.837 --> 00:49:55.885 I think wanting to learn and focusing on learning and having that kind of lifelong 00:49:55.885 --> 00:49:59.919 learner mindset is super critical to this. 00:50:00.615 --> 00:50:04.541 Because I remember when, again, back when I was an undergrad, 00:50:05.315 --> 00:50:07.072 I took a course in computer graphics. 00:50:07.576 --> 00:50:11.853 And now back then, again, we're talking dinosaur days. So computer graphics 00:50:11.853 --> 00:50:15.593 was a really new thing. I mean, we're talking Pong level graphics here, 00:50:15.593 --> 00:50:19.903 if you even know what that game is. But I mean, we're talking really crude stuff. 00:50:20.543 --> 00:50:25.113 And the professor I had that taught that was a really good professor. 00:50:25.113 --> 00:50:26.343 I'd had him for other courses. 00:50:27.903 --> 00:50:32.363 But I could tell he had just read one chapter ahead in the textbook. 00:50:32.727 --> 00:50:35.543 That's what he knew about this. That's what qualified him. 00:50:36.023 --> 00:50:39.333 I mean, his experience. But he didn't know any more about graphics other than 00:50:39.333 --> 00:50:41.923 he had read one chapter ahead in the textbook from us. 00:50:42.323 --> 00:50:46.283 Because I realized when I read that chapter, he was almost quoting it verbatim. 00:50:46.803 --> 00:50:53.364 So, okay. But the reason I bring that up is I feel like we're all a little bit like that now. 00:50:53.618 --> 00:50:58.213 What qualifies you as an expert these days, especially in AI, 00:50:58.213 --> 00:51:02.513 a fast-moving topic like this, is you read one chapter ahead in the textbook 00:51:02.513 --> 00:51:03.819 compared to everybody else. 00:51:04.423 --> 00:51:08.184 And nobody has finished the book. because it's still being written. 00:51:08.863 --> 00:51:14.443 So there's, you know, we have to be careful that we don't, and I love using 00:51:14.443 --> 00:51:17.263 analogies, that we don't outrun our headlights on this. 00:51:17.783 --> 00:51:22.443 And a lot of people are doing that, you know, but that's the thing. 00:51:23.083 --> 00:51:27.163 You keep pushing forward into the darkness, trying to fill it with more knowledge. 00:51:27.683 --> 00:51:31.833 And again, I think we're in a great place to be able to learn because there's 00:51:31.833 --> 00:51:36.082 never been more resources available to us than now, but it's moving fast. 00:51:36.395 --> 00:51:39.673 So if you don't like learning, this is not your era. 00:51:39.743 --> 00:51:41.423 This is not for you. I love that. 00:51:42.063 --> 00:51:46.583 I recommend you build a time machine and go back in time because it's not going to slow down. 00:51:47.003 --> 00:51:48.833 I actually really love that because 00:51:48.833 --> 00:51:54.063 you know what? I think that summarizes 26, 27 for a lot of teachers. 00:51:54.772 --> 00:51:59.403 I think we need to put that on there that if you don't love learning, 00:51:59.403 --> 00:52:03.683 you're in the wrong place kind of thing. Because if we don't get that through... 00:52:04.531 --> 00:52:07.548 Not just the teacher's heads, because that's kind of easy. We go, 00:52:07.548 --> 00:52:11.098 oh, yeah, of course. But I think reminding kids, 00:52:13.498 --> 00:52:17.318 that even though AI can do this stuff, if you're not learning with it or learning 00:52:17.318 --> 00:52:21.278 from it, you're really not doing too well for the future. 00:52:22.118 --> 00:52:25.518 Yeah. I think there's a push and pull on this. You know, the kids maybe don't 00:52:25.518 --> 00:52:29.228 necessarily want to learn. The teachers like learning. They've chosen this as 00:52:29.228 --> 00:52:31.090 their profession because they love learning. 00:52:31.678 --> 00:52:36.923 But then on the other side, what the kids are better at is flexibility and adaptability. 00:52:37.438 --> 00:52:38.858 And a lot of times the teachers are not. 00:52:39.698 --> 00:52:43.458 And I find they've become very resistant. They learned things in a certain method, 00:52:43.818 --> 00:52:46.936 and they assume that's the same method that we'll always have to carry forward. 00:52:47.156 --> 00:52:51.480 You know, I gave an example in the video on AI and the Future of Education, 00:52:51.724 --> 00:52:53.548 where I start off writing in cursive. 00:52:53.849 --> 00:52:59.158 And I say, you know, look, there was a time when that was a really important skill to have. 00:52:59.158 --> 00:53:03.178 It just became a log. again, and they want to play a law again in Florida. 00:53:03.578 --> 00:53:04.018 No, fine. 00:53:05.398 --> 00:53:10.448 Fine. Why don't they go ahead and teach all the kids how to put shoes on horses, too? 00:53:10.448 --> 00:53:13.078 I think so. Because that's going to be around as relevant. We might as well, 00:53:13.078 --> 00:53:17.768 you know. I mean, you know, we got to train some more OCR better with some really, 00:53:17.768 --> 00:53:19.438 really scribbly cursive heads. 00:53:20.158 --> 00:53:24.278 I mean, no, the bottom line is, you know, we free up, educate. 00:53:24.278 --> 00:53:27.298 We have precious little educational time. Instructional time. 00:53:27.758 --> 00:53:31.568 Is that really where we want to spend it? is that the hill you want to die on 00:53:31.568 --> 00:53:34.978 cursive i'm going to recommend that that not be the. 00:53:34.978 --> 00:53:38.818 One although it was very embarrassing when my son went for his driver's license 00:53:38.818 --> 00:53:41.518 and they told him to sign for his driver's license and he didn't know how to 00:53:41.518 --> 00:53:44.758 sign it and he starts printing like a two-year-old i'm like what is that, 00:53:47.458 --> 00:53:48.138 and you know. 00:53:48.478 --> 00:53:51.028 His kids are going to look back and say what was this. 00:53:51.028 --> 00:53:56.605 Scrawl that all these old people did why didn't they know how to write what is a pen and, 00:53:57.121 --> 00:53:58.649 What is a pen? Yeah. 00:53:59.419 --> 00:54:04.469 Well, most of what we write these days is not with a writing instrument. 00:54:04.869 --> 00:54:06.149 Most of it's done with a keyboard. 00:54:06.149 --> 00:54:07.879 Yeah, our thumbs. On our thumbs. 00:54:07.879 --> 00:54:13.219 Yeah. And my grandmother got awards for penmanship when she was in school. 00:54:13.219 --> 00:54:13.559 Yeah. 00:54:13.559 --> 00:54:18.679 Okay, nobody cares now. I'm sorry, but they just don't because when I type it, 00:54:18.679 --> 00:54:20.229 it looks the same as when you type it. 00:54:20.709 --> 00:54:25.229 You know, that skill that used to be really important in the past is not as important in the past. 00:54:25.569 --> 00:54:31.365 So, teachers have to be adaptable and flexible and looking forward. 00:54:31.469 --> 00:54:35.450 I mean, if we're the educated ones, we've got to lead people in that area, 00:54:35.786 --> 00:54:39.752 not just teach them this way because this is the way it's always been done. 00:54:40.158 --> 00:54:42.562 If we're smart, we know that's not an answer. 00:54:42.734 --> 00:54:46.839 Yeah. No, I love that. And that's actually why the only reason I drew on that 00:54:46.839 --> 00:54:50.279 piece of paper for that process flow I was putting together is because I didn't 00:54:50.279 --> 00:54:53.089 have a $30,000 electronic board. 00:54:53.589 --> 00:54:54.619 Oh, okay. 00:54:54.619 --> 00:54:55.878 So if you want to send one over. 00:54:56.033 --> 00:55:02.227 That's all. So, yeah, I hear the plea for a GoFundMe. Yeah, exactly. 00:55:02.471 --> 00:55:06.781 Can you imagine if we had a studio or a podcast? I would actually, 00:55:07.104 --> 00:55:09.002 you know, we'd go back to live streaming. 00:55:09.269 --> 00:55:11.389 If I'd be over there every week to record. 00:55:11.389 --> 00:55:11.869 It'd be great. 00:55:12.409 --> 00:55:13.029 Oh, sure. 00:55:13.809 --> 00:55:17.189 So just one last question from me, and it's kind of tongue-in-cheek, 00:55:17.189 --> 00:55:19.949 but you said you like to use analogies, right? 00:55:20.849 --> 00:55:24.729 And quotes and things like that. last week in my presentation, 00:55:24.729 --> 00:55:29.549 I was saying AI is not an oracle, right? It's not a know everything kind of 00:55:29.549 --> 00:55:33.729 thing. And I wanted to put a picture of the oracle from the matrix. 00:55:34.409 --> 00:55:37.997 And then I thought to myself, I don't know who's going to understand this in the room. 00:55:38.368 --> 00:55:42.339 So do you have, as you put your stuff together, do you ever have those moments 00:55:42.339 --> 00:55:46.929 where you think about your audience and think, who's going to get this? Is this going to flop? 00:55:46.929 --> 00:55:50.269 Constantly. Constantly. Because here's the thing. 00:55:50.789 --> 00:55:55.581 Um, so I, just to give you a general idea of, of how long I've been around, 00:55:55.889 --> 00:56:00.129 I got my, my bachelor's in computer science in 1984. 00:56:00.129 --> 00:56:00.419 Okay. 00:56:00.889 --> 00:56:01.276 So. 00:56:02.252 --> 00:56:04.592 That's when I'm talking about dinosaur era. 00:56:04.932 --> 00:56:10.122 Now, I'm up talking to my students every week, and I mean- 00:56:11.227 --> 00:56:16.607 When I was already 20 years into my IBM career, they were not yet zygotes. 00:56:17.407 --> 00:56:21.627 So they're not getting the reference that I'm making on these things. 00:56:22.447 --> 00:56:26.057 And I'll put these out there. And like what I just said to you, you all laughed. 00:56:26.057 --> 00:56:28.507 I'm a former biology teacher. That's why. 00:56:29.987 --> 00:56:35.647 I just get blank stares. And then I'm like, okay, I guess I have to explain this to you, don't I? 00:56:35.647 --> 00:56:38.483 No, they have it on recording. They're like, what's zygote? 00:56:39.187 --> 00:56:44.787 Yeah, right, right. So, yes, there's a lot of analogies I would like to make, 00:56:44.787 --> 00:56:48.997 and I have to try to figure out what my audience is, and are they old enough 00:56:48.997 --> 00:56:50.287 to understand these things? 00:56:51.227 --> 00:56:58.587 When I was in Cartagena last week, I was talking about the movie 2001 A Space Odyssey. 00:56:59.247 --> 00:57:02.587 And, okay, so that movie came out in, I think, 1969. 00:57:03.907 --> 00:57:09.287 And, you know, I realized most of the people in here probably weren't alive yet. 00:57:09.767 --> 00:57:13.257 So they had they haven't seen this movie they probably some of them probably 00:57:13.257 --> 00:57:17.427 most of them probably not even heard of it uh so i'm gonna have to explain to 00:57:17.427 --> 00:57:22.547 them what that reference is about you know so yeah so yes but um, 00:57:23.267 --> 00:57:27.587 but yeah there's there's a lot of those i do think analogies are really powerful 00:57:27.587 --> 00:57:32.547 though because we take something that we know already and we're comparing it 00:57:32.547 --> 00:57:34.607 to something that we don't know and 00:57:35.047 --> 00:57:39.134 basically saying oh yeah you kind of already know this because it's a lot like that. 00:57:39.295 --> 00:57:42.267 But then, of course, analogies, you stretch them too far and they break down. 00:57:42.687 --> 00:57:47.307 So they're not always a perfect replication, but they're good for a quick illustration. 00:57:47.307 --> 00:57:49.557 And I think most people can latch onto them. 00:57:49.557 --> 00:57:53.427 Okay, real quick, because we have like a couple minutes and we're usually trying 00:57:53.427 --> 00:57:56.927 to cut it an hour because Julian has- That's because. 00:57:56.927 --> 00:58:00.767 You don't have me as a guest. You're able to do that. 00:58:01.087 --> 00:58:07.127 Sometimes we just go over. For kids, and this is kids pre-college, 00:58:07.447 --> 00:58:10.305 right? Kids graduating out of, 00:58:11.234 --> 00:58:17.741 school next year, 12th grade. And what are three things, because you're in college, 00:58:17.741 --> 00:58:23.227 you're a professor, what are three things that you wish every senior leaving, 00:58:23.900 --> 00:58:28.211 would have in their kind of back pocket? Besides a brain. 00:58:28.211 --> 00:58:30.321 Only three? 00:58:30.321 --> 00:58:31.021 Only three. 00:58:31.381 --> 00:58:32.282 Only three. 00:58:32.444 --> 00:58:33.401 We have to keep it short. 00:58:34.301 --> 00:58:38.341 Yeah. Okay. I'm going to go with communication. 00:58:39.580 --> 00:58:42.570 Incredibly important, especially for people in the STEM fields. 00:58:43.661 --> 00:58:48.991 I can remember, I think a lot of my students even now would be happy to have 00:58:48.991 --> 00:58:52.361 a career where we put them in a closet, slide pizzas under the door, 00:58:52.361 --> 00:58:56.041 and they give us code out the other side. Okay. So what do I do? 00:58:57.521 --> 00:59:02.811 I don't ask them to write code. 20% of their grade is class participation. And most of them hate it. 00:59:04.041 --> 00:59:08.341 They're super introverts and they do not want to talk. And that's exactly why 00:59:08.341 --> 00:59:09.161 I'm going to make them do it. 00:59:09.661 --> 00:59:12.651 Because that's going to be the difference between who wins and who doesn't. 00:59:12.651 --> 00:59:16.611 We've got AI that can do a lot of that, slide the pizzas under the door. 00:59:16.611 --> 00:59:17.861 It doesn't even require pizza. 00:59:18.261 --> 00:59:23.271 So it can just give us code. So the ability to communicate is going to be incredibly 00:59:23.271 --> 00:59:29.030 important. The things that are human, unique, we need to lean into. I think, 00:59:30.009 --> 00:59:33.773 being curious is a big part of this as well. 00:59:34.789 --> 00:59:39.449 One of my favorite quotes here also that I saw recently is, curious people don't 00:59:39.449 --> 00:59:41.769 get replaced, they get liberated. 00:59:42.537 --> 00:59:46.659 And that's what AI will do for us if we're the curious type, 00:59:46.659 --> 00:59:50.029 because we'll move on. As it starts to do the grunt work for us, 00:59:50.029 --> 00:59:54.288 we'll move on to the next interesting thing and continue to win with that. 00:59:54.513 --> 01:00:00.290 And then the most important skill of the AI era is a third C. critical thinking. 01:00:01.269 --> 01:00:04.673 We need to not trust everything that comes out of AI because sometimes it's wrong. 01:00:04.992 --> 01:00:11.649 It will hallucinate, which is a convincing inaccuracy that it tries to tell us. 01:00:12.209 --> 01:00:16.591 Sometimes it's been manipulated to tell us things. Sometimes it just makes mistakes. 01:00:17.652 --> 01:00:21.959 What we wanted to do and when we wanted to do it and why we wanted to do it, 01:00:21.959 --> 01:00:24.689 again, all going back to humanities kinds of questions. 01:00:26.089 --> 01:00:31.189 I think that kind of critical thinking and deciding The main thing I try to 01:00:31.189 --> 01:00:35.049 get across to my students in one of the courses that I teach on secure thinking 01:00:35.049 --> 01:00:37.487 is just because you can do something doesn't mean you should. 01:00:37.852 --> 01:00:41.809 You could write enough code to do anything you want, but should you? 01:00:42.289 --> 01:00:45.239 What are the unintended consequences if you were to do that? 01:00:45.239 --> 01:00:48.769 How could it be used and abused? 01:00:49.409 --> 01:00:53.689 And let's try to put in the safeguards. So critical thinking, 01:00:54.049 --> 01:00:58.389 I think, is also more important than ever. And where do we learn critical thinking? 01:00:59.029 --> 01:01:03.339 Again, I'm going back to the humanities. So it's gonna be a balance. 01:01:03.339 --> 01:01:09.670 We don't want just a bunch of techie engineer types who don't understand the larger context. 01:01:09.803 --> 01:01:15.533 We've gotta be more versatile and spread our knowledge around and lean into the things, 01:01:15.998 --> 01:01:19.662 that ai can't do so that would be my advice. 01:01:19.662 --> 01:01:22.098 Very cool julian you have any one last. 01:01:22.377 --> 01:01:25.762 No no no i i love it this that was the perfect ending i. 01:01:25.762 --> 01:01:32.032 Think perfect ending so yeah so great um so if people wanted to reach out to 01:01:32.032 --> 01:01:34.706 you is there an easy way to get in contact with you. 01:01:35.031 --> 01:01:39.622 Linkedin's probably the easiest yeah if you can find me uh there's a there's 01:01:39.622 --> 01:01:45.742 another jeff crum if you if you google me and he is an evangelist and motivational 01:01:45.742 --> 01:01:47.562 speaker in California. That's not me. 01:01:48.102 --> 01:01:53.522 So, you know, just, but you can Google me or you can find me on LinkedIn. 01:01:53.522 --> 01:01:56.482 It'll be pretty obvious, which if you got the right Jeff Kroon or not. 01:01:56.482 --> 01:02:01.622 Excellent, excellent. I can attest to that. On a Saturday or Sunday morning, 01:02:01.622 --> 01:02:04.852 I was like, oh, and you did reach back. You're like me. I'm like, 01:02:04.852 --> 01:02:06.562 who's on LinkedIn this early like me? 01:02:07.982 --> 01:02:09.152 Anything to share, Julian? 01:02:09.571 --> 01:02:12.942 No, no, I'm good. This has just been inspiring for me. I'm definitely going 01:02:12.942 --> 01:02:15.582 to go through the rest of your videos. Good luck. 01:02:15.582 --> 01:02:17.042 There's like 5,000 of them. 01:02:18.462 --> 01:02:21.942 It's a cure for insomnia. That's not much I'm sure of. 01:02:23.122 --> 01:02:26.322 I went to bed super early last night after watching your video, so you might be right. 01:02:28.262 --> 01:02:29.162 You're making my point. 01:02:29.162 --> 01:02:32.742 Exactly. No, I'm just really appreciative of what you're doing, 01:02:32.742 --> 01:02:37.362 Jeff, because as I'm seeing over here with a lot of the companies I'm talking 01:02:37.362 --> 01:02:41.792 to is exactly the kind of pain point you're solving for people. 01:02:41.792 --> 01:02:46.065 So love it. Absolutely love it. Thank you for joining us. 01:02:46.547 --> 01:02:46.832 Cool. 01:02:46.832 --> 01:02:48.092 Thank you all. I enjoyed it. 01:02:48.092 --> 01:02:51.022 Well, hopefully, maybe I'll try to get to Atlanta and go to IBM. 01:02:51.022 --> 01:02:54.002 I only went for one day last year because I was like, it's just a train ride, 01:02:54.002 --> 01:02:57.692 but maybe we'll see each other because I learned a lot in just one day at IBM. 01:02:57.692 --> 01:02:58.782 So hopefully- If you get there. 01:02:58.782 --> 01:03:03.442 Definitely look me up and I'd say that to anybody, the IBM Tech Exchange Conference in Atlanta. 01:03:03.442 --> 01:03:03.762 Excellent. 01:03:03.762 --> 01:03:05.354 Do I have one in Australia yet? 01:03:06.487 --> 01:03:08.686 No, I'm afraid. Not that I'm aware of. 01:03:08.686 --> 01:03:09.216 You should found it. 01:03:09.316 --> 01:03:11.436 Maybe they do, but I'm not invited, I guess. 01:03:12.536 --> 01:03:13.516 Only Americas. 01:03:14.216 --> 01:03:19.006 Yeah, yeah. They have higher standards for that one than to let me have. 01:03:19.006 --> 01:03:20.066 Yeah, that's true. 01:03:20.066 --> 01:03:24.409 All right. We're going to wrap it up. Otherwise, we'll talk forever. 01:03:24.711 --> 01:03:27.836 All right. So for Teaching Python, this is Kelly. 01:03:28.136 --> 01:03:29.617 And this is Julian. Signing off. 01:03:30.267 --> 01:03:31.696 Oh, he remembered. Well done. 01:03:32.056 --> 01:03:33.046 Got it. Got it. 01:03:33.056 --> 01:03:33.456 Got it. 01:03:36.322 --> 01:03:36.693 Peace.

Open the transcript file

Hosts and Guests

Hosts

  • Julian Sequeira host Feed-stated
  • Kelly Paredes host Feed-stated

Guests

  • Jeff Crume guest Feed-stated

Credit the people on this episode

Be the first to rate this episode

More episodes

All episodes
  1. Episode 165: Everyone is Becoming a Technologist

    Nov 13, 202644 minAudio

    How is AI changing job boundaries? AI is making it easier for people to take on tasks that once went to a developer, analyst, or other specialist. In its July 2026 analysis of U.S. ChatGPT use, OpenAI found that 43.5%...

    Credit the people on this episode

  2. Episode 164: The Science of Remembering

    Oct 16, 202658 minAudio

    How much of what we learn do we actually remember? And what if one of the most effective ways to learn is to regularly challenge ourselves to recall what we already know? In episode 164 of Teaching Python, Kelly...

    Credit the people on this episode

  3. Episode 163: How Do You Find Time to Keep Learning?

    Oct 2, 202653 minAudio

    How do you keep learning when there never seems to be enough time? The answer may be less about finding more time and more about building learning into the way you already work, solve problems, use AI, and connect with...

    Credit the people on this episode

  4. Episode 162: What Does “Teaching Python” Mean in 2026?

    Sep 19, 202630 minAudio

    What does teaching Python mean in 2026, when AI can generate code and programming sits inside a much larger computing landscape? Teaching Python started as a podcast about teaching programming in the classroom, but both...

    Credit the people on this episode

  5. Episode 160: Data Science, Math and Python, Oh My!

    Jul 16, 20261 hrAudio

    In this episode, Kelly Schuster-Paredes speaks with Mahmoud Harding about his work in data science education and the way he thinks about teaching Python, R, and statistics. Mahmoud explains that he is the instructional...

    Credit the people on this episode