Spencer Kelly: 0:00

Many years ago I went on Five Live the radio session and was asked to comment on rumors that Apple was just about to launch a tablet. Ok, so this dates it. And I said confidently no, it's a stupid form factor, you know? I mean you have to hold up this great big brick and you have to hold it there. Why would you want to do that? You know you get an achy wrist. Typing on it is a nightmare, because typing on touchscreens is rubbish. It turns out Steve Jobs wasn't listening to Five Live that afternoon. About six hours later he launched the iPad On that idea of disruption.

Will Ockenden: 0:35

Actually and I've got a view on this, you will have a view on this what's the next market right for disruption in your view?

Spencer Kelly: 0:48

Should every marketer that's listening to this show now turn off, because they all need to go off and become plumbers? Is that what we're saying? Be a plumber?

Chris Norton: 0:53

um, yes, thanks for watching thanks for listening and goodbye, hello and welcome back to embracing marketing mistakes, the podcast where we ask comms professionals to unpack the mistakes that made them better at what they do. I'm your host, host Chris Norton, and I have to say I'm particularly excited about today's guest. I used to watch him every single week on BBC's Click and he's one of the most trusted voices in tech broadcasting. So I'm thrilled that Will and I are joined by Spencer Kelly. We chat about what it was like to introduce people to AI, the internet and virtual reality before they were everyday buzzwords, as well as the highs and lows along the way. Spencer's insights are brilliant and it's a fascinating behind-the-scenes look at how technology was translated for the masses on the BBC. So, as always, sit back, relax and let's get into the very human side of telling tech stories to the world.

Will Ockenden: 1:41

Enjoy tech stories to the world. Enjoy.

Chris Norton: 1:47

Spencer Kelly. Welcome to the show. Hey, thanks for having me. I mean, this is I'm geeking out here, but I haven't actually even told you this, so I was telling the staff. I've got you to come on the show, which thank you so much for doing this. No problem, but I've probably learned more about technology from you and Lara than I have from any other person on the planet, Because you've presented Click for like. Is it 10? No, it's more than 10 years, isn't it?

Spencer Kelly: 2:13

The program lasted for just shy of 25 years. I hosted it for 20.

Chris Norton: 2:19

Right, yeah, and we were on a thousand episodes, weren't?

Spencer Kelly: 2:24

you? I personally lasted, as a host, exactly a thousand episodes, weren't you? I personally lasted, as a host, exactly a thousand episodes. The program ran for 1,296 episodes, any more stats? I've got them all, because that's the kind of person I am.

Chris Norton: 2:37

And the thing that was brilliant about it is that the reason why I liked Click is that I loved Tomorrow's World. I'm a tech geek, which is why I've always liked it. I like in marketing, I like technology and I'm sort of in that sphere where I like anything that's coming out emerging technology, and I started in consumer tech as well in PR. So I've loved the show and watching how things have developed Because I remember Tomorrow's World right, which predated BBC Click.

Spencer Kelly: 3:05

I know it's the reason I do what I do.

Chris Norton: 3:06

And I remember them going. There's a thing coming out called the Information Superhighway. Yeah, the Information.

Spencer Kelly: 3:13

Superhighway, I do.

Chris Norton: 3:14

And so like watching how that chat and then Tomorrow's World ended. So how did Tomorrow's World end and your world of clicks?

Spencer Kelly: 3:21

Well, funny story. So I grew up wanting to be on tomorrow's world, to present tomorrow's world, and actually, when you look at what they were able to do live, it's a technical phenomenon, so never mind technological. They were doing live demonstrations of stuff that may or may not work, while remembering their lines and talking to camera and everything like it wasn't recorded tomorrow no, uh, it was live.

Spencer Kelly: 3:43

And the demos. I mean, there will be some films they ran that were pre-recorded, but a lot of their demos happened live, um, so that was brilliant. I grew up wanting to present tomorrow's world. The day I joined the bbc, they axed tomorrow's world, january the 3rd 2003. Look it up, um. And I thought, well, that's that, then that's an omen, um. But then I um this program, which was called Click Online at the time, that no one had heard of in the UK because it was made only for overseas BBC World. So, overseas, I found that and talked my way into that.

Chris Norton: 4:16

BBC World Right. So I'm trying to remember when I first saw it, because I went away about 20 years ago.

Spencer Kelly: 4:21

Yeah, you'll be in some grotty hotel in Europe with the only English-speaking channel on. It's raining outside. Yeah, that's how many people.

Chris Norton: 4:28

I think I saw it in.

Spencer Kelly: 4:28

Thailand.

Will Ockenden: 4:29

Right, there you go. Might be the first time Humble beginnings. So what were you talking about in 2005 then? What was the hot tech then, 20 years ago?

Spencer Kelly: 4:38

So 20 years ago? Well, let's go back to when I joined 2003, and a lot of it was beige boxes under the desk, because that's what a PC was, and so what was really interesting was we had to work very hard to make that televisual, because it's just a beige box. And so we'd be talking about the Internet and we'd be featuring lots of footage of rivers and water flowing down different rivers that's how data flows or roads that get jammed up with traffic, and that's the internet slows down or or speeds up. So that's the sort of thing we we would be talking about. Um, and I think to my very first show that I hosted.

Spencer Kelly: 5:14

So I worked on it as a producer, reporter for a couple years 2005, I filmed my first show as host, and I missed the flight to china. So that was good, wasn't it? And so I arrived a day late and we were talking about the rise of china as a superpower, and I interviewed jack ma, who I think is the head, or was the head, of alibaba, who then went missing for a while. I think he's back now. So he was one of the um, the early Chinese entrepreneur, billionaire type people.

Will Ockenden: 5:46

So was Alibaba a significant proposition even back then.

Spencer Kelly: 5:49

I think we hadn't heard of it because, if you think back to them, we weren't really aware of what China was doing. In fact, japan was just being overtaken by South Korea at that point. We filmed in South Korea around that time and it still wasn't south korean goods weren't big over here. You know that I had a lucky gold star tape player, which lucky gold star became lg, and we know it now. But it's only really thanks to android, the operating system, and samsung adopting that the south korea. I think that that catapults south korea into the limelight, but china was still considered, you know, cheap yeah um, cheap, low quality stuff.

Spencer Kelly: 6:32

And then, lo and behold, china overtakes south korea. Um, and then we're now looking for what happens next. You know, would it be india? Will it be something else?

Chris Norton: 6:41

so, wow, I mean I'm trying to. I think I might have actually seen that episode proper geeky.

Spencer Kelly: 6:47

Um, do you remember my hair? It was a different color in those days I mean I've seen.

Chris Norton: 6:51

I've seen you in lots of different guises, I've seen you virtually, I've seen you in augmented reality.

Spencer Kelly: 6:55

Oh man I've been sampled so many times. I'm not really here. This is just a 3d hologram. I've been scanned. My voice has been cloned. I've been deep faked with Donald Trump. What were those boxes? You did?

Chris Norton: 7:05

which I don't think we'll say, and it was like these boxes where you had your body full, three, 60 scanned, and then there was a physical, like it's a hologram of you in in a box. It was a, it was a met and he was using it for like fashion.

Spencer Kelly: 7:18

Yeah, if it's the one you're thinking of, it used to be called Portal. Yeah, that's it. It's run by a guy called David Nussbaum, I think, who's a lovely guy in California, la, I think. And yeah, it looks like a full-size hologram in a box and what it is is basically it is a cabinet, an empty cabinet, and the glass on the front is actually a transparent LCD screen, which means that anything that would normally be black is transparent, so you can see through to the box and you can see. That gives it a kind of 3d feel. So it's just a video of me, but you can see past me to the background. It just it's the most realistic 3d ish thing I've ever seen. It's not a hologram. I get very angry when people use the term hologram incorrectly. It was absolutely stunning, even in the room. You know, sometimes, uh, on tv you can either make things look better or they don't. Did you move? I can't remember. Yeah, yeah, it's a live video feed from a camera.

Chris Norton: 8:11

It's bizarre yeah, so you could beam yourself to be somewhere in a in a different. You can like do a show from somewhere and be in a different.

Will Ockenden: 8:18

So is that? Is that taken off to any degree?

Spencer Kelly: 8:20

um, I've seen lots of knockoffs, so there's different companies doing it now. Um, and I I have seen them around. I don't have the figures. I mean all I would say is, visually it looks really good. Yeah, it doesn't strike me. It's that expensive, um, and it can be live. And the idea is, if it had a couple of cameras at the top of the box, someone could stand in front of it, ask some questions and the the person who's appearing in the box is somewhere else and gets that feed and then does something like it's cool yeah, so, um, over the over the last 20 years, then you must have seen some absolutely wild technology.

Will Ockenden: 8:55

Yeah, baby. So what, what, what's kind of what have you seen? It is just mad tech that you thought might have taken off, and it hasn't, and it's never been talked about again.

Chris Norton: 9:04

What an excellent question well, mini disc uh, I had a mini disc, mini disc was great record dj sets on my mini disc.

Will Ockenden: 9:12

It was really good storage capacity, wasn't it? Yeah?

Spencer Kelly: 9:15

I. So I remember losing my marbles over a humanoid robot in south korea which was called hubo and it was going to enter the DARPA Robotics Challenge, which is a biannual competition by DARPA, the Defense and Research Podcast Association. No, I don't think so. In America, darpa, and it's to encourage advances in things like robotics. And so this Hubo robot was rehearsing for the DARPA Robotics Challenge and it had to drive a car. To get in a normal car drive, it had to get out of the car, it had to pick up a drill, drill a hole through a wall, it had to turn off a gas valve, open a door, climb over some rubble, and I went absolutely nuts because this thing was absolutely acing it. And it went on to win the DARPA Robotics Challenge a few weeks later.

Spencer Kelly: 10:02

Right, so it's not necessarily something that hasn't turned up anywhere, but I think humanoid robots we get overexcited about humanoid robots, whereas actually robots are machines, autonomous machines that are built for a specific purpose. So the thing that might vacuum your floor is a robot. It doesn't have to look like a human. The thinking behind humanoid robots is that you can put them into situations where we would normally be, but for whatever reason, it's not safe to. But I I'm looking around, I don't see any oh, and he's a robot actually.

Chris Norton: 10:38

There we go they work.

Will Ockenden: 10:40

We were like one.

Spencer Kelly: 10:40

He says yeah, um but yeah, I think we we kind of anthropomorphize ai. We think it's going to look like a? Um, a humanoid robot. But I have been told when I do my keynotes to, to companies and stuff, we do a lot on ai these days, um, and we're talking about the fact that a lot of jobs we don't know. I don't know what my kids are going to do for a living, I don't know what jobs are going to exist to them, because a lot of stuff's going to be replaced by ai. But there'll also be new jobs. They're going to be invented. The safest job to train in is a plumber because, when you think about it, you've got to step over the kids toys. You've got to open the, the cupboard, you've got to pull out all the fairy liquid and stuff. You've got to look at the, the pipes and go, we'll put that in. You can't get a robot to cope with that. Hundreds of times a week from from different houses spot might argue with you about that.

Chris Norton: 11:34

Okay, spot the dog.

Spencer Kelly: 11:36

Yeah, I, it's brilliant. I've driven spot the robot across a beach in san francisco from my sofa in Hampshire during lockdown with an Xbox controller. So it's brilliant, but it's a four-legged thing that needs all four legs to do things. Now it comes down to it can have an arm where its head's supposed to be, but that was me remote controlling it. I remember that as well. There you go Were you there? No, no, I remember Were you in my garden for like the whole of the 20 years.

Chris Norton: 12:08

He was laying on the beach when the dog walked over I also saw the one where you had Spot and he went inside because I used to work. My first job was working in a comms team in a nuclear power station and you took Spot and I don't know if you or lara, but you were inside, yeah, one of the pack, I remember which one of you it was and you went inside a nuclear reactor.

Spencer Kelly: 12:28

Yeah, and that spot was going around. That's the point. Yeah, you send these. You send these things into places humans should be able to go, but for whatever reason nuclear disaster, um, you can't send this. So that's the idea of those.

Chris Norton: 12:40

So should every marketer that's listening to this show now turn off, because they all need to go off and become plumbers? Is that what we're saying?

Spencer Kelly: 12:45

be a plumber.

Will Ockenden: 12:47

Um, yes, thanks for watching, thanks for listening and goodbye what, what, what have you seen that should have taken off? That never did then. So you must have seen stuff that was amazing, but it's just kind of like the mini disc again the mini disc filled the niche for a while, uh, and it went away.

Spencer Kelly: 13:36

I'm trying to think. I think there are things that we think are going to be big and everyone goes mad about. Um, like drones, you say, take off, literally drones. You know, we for a while we thought drones were going to be like everywhere. Um, I still believe that amazon is going to deliver things by drones, apparently, according to someone, that it's just never gonna happen. It's never gonna happen. I mean, I got, have you flown drones?

Chris Norton: 14:05

yeah, so that's an interesting question. You got there, spencer, and as this show is embracing marketing mistakes, I can embrace a marketing mistake that's happened by my video team, which was Andrew. Do you want to come over and tell the story?

Spencer Kelly: 14:19

So you went very humble there and said I'm going to embrace a mistake made by someone else and not me.

Chris Norton: 14:23

Embracing a mistake for the team, but actually I want the story to be told by the team member that was there, you want to humiliate the person who was involved in it. Okay, basically, um, we have a high-end 2 000 pound wealth I don't know how high end you want to go, but it did 8k, I think a 2 000 pound drone. We've had it. We had it for about 10 months and we're doing a very clever, intricate shot and our videographer threw it, flew it into a tree.

Spencer Kelly: 14:49

Well, we've all been there. I've crashed two of them, we I nearly got killed by one in norway, really. Uh, yeah, that was a big octocopter with a 4k camera on the bottom and, uh, what we found out later was, in order to get it ready for us to to be ready on the day because it wasn't quite ready the guy who built it had disabled some of its sensors its gps sensors and so it was up high and we were using it to film my piece of camera. Hello, welcome to click. And then this camera takes off and goes into the sky and it all looks fabulous. Um, I think it didn't realize what altitude it was. It thought it had landed and just shut off, and this thing came down from a height and right so that was heavy wasn't it.

Spencer Kelly: 15:33

It went into the ground, some, and bits of the camera flew off everywhere. Did it capture your reaction as it nearly hit you? We had the footage of it coming down and then everything goes wild. How much was that then? How much was that worth? I don't know, because it was a guy who built it back in the day, so this is more than 10 years ago when people were still building these things. So the idea that drones are going to be zipping around and delivering stuff, we we have to see a massive change in what they can do in their battery life. Um, how autonomous do you want them? Uh, if a, if a robot that goes along the pavement goes wrong, it might just bump into you. If a drone from up there, as discussed, goes wrong, it could fall on you and kill you with the pizza attached to whatever nonsense it is to get a slice of pizza.

Spencer Kelly: 16:18

Yeah, well, you know, peel it off your skull if you want. So I, yeah, I would say. I would say drones on the flip side, because you're talking about marketing mistakes, or at least making mistakes. Um, just before, well, many years ago, I went on five live um the radio session and was asked to comment on rumors that Apple was just about to launch a tablet. Okay, so this this dates it. And I said confidently no, it's a, it's a stupid form factor, you know. I mean, you have to hold up this great big brick and you have to hold it there. Why would you want to do that? You know you get an achy wrist. Typing on it is a nightmare, because typing on touchscreens is rubbish. It turns out Steve Jobs wasn't listening to Five Live that afternoon. About six hours later he launched the iPad, but then they quietly launched a keyboard and a stand for it a little bit afterwards. And I still maintain it's a stupid form.

Will Ockenden: 17:13

Your position was right though, and it wasn't.

Spencer Kelly: 17:16

It was a tiny kind of segment in the market, wasn't it still is all right, yeah, yeah, and also, I mean, uh, I interviewed daniel eck, founder of spotify, many years ago and I told him his his idea wasn't going to work. Because I come from a commercial radio background, I know how expensive it is to run a music radio station. You have to have like nine or ten minutes of adverts an hour to pay for the licensing and stuff. So I said, well, you know that no one's going to choose to listen to that. It turns out, uh, he didn't listen to me either and, although I maintain it, lost hundreds of millions of dollars for many years before it turned is it, making a profit now I think it is, but only after it put the rest of the music industry out of business and by not paying artists.

Chris Norton: 17:57

That's why it's just announced some sort of job cuts or something.

Spencer Kelly: 18:02

It was in massive losses for a long time. I think at some point recently it started to turn a profit. It just goes to show. I did not appreciate how, if you have enough money, if you have enough venture capitalist money behind you, you can disrupt an entire industry and bend it to your will I'm trying to think well of something that that we've thought is going to be big in marketing.

Chris Norton: 18:22

That's just not come off well I on that idea of disruption.

Will Ockenden: 18:26

Actually and I've got a view on this, you will have a view on this what's the next market right for disruption in your view? You know we've seen all this disruption over the years. I think the car hire industry. You know, when you go to the airport and you have to queue up, there's two people taking time to have cigarette breaks. You queue for an hour. You have to fill in all of this form. Do you know?

Spencer Kelly: 18:45

I think exactly the same car hire and hotel check-in. What are they doing on those screens? That takes so long. I've pre-booked it and you're right. It's not a long queue, but everyone takes like 10, 15 minutes to go and they're going through whatever, filling in whatever boxes. I think I want to lean over and see what is it? Yeah, that you're feeling same with hotel check-ins. So you're absolutely right and in fact you do get like what zip cars or I think they're calling stuff now where you can, and it's almost airbnb. It's like that's a car where you can just pick up a car and leave it somewhere else you've got the cars in la now, haven't you?

Chris Norton: 19:17

that is uber with no other, completely autonomous and drive themselves. Teslas that drive themselves, and which is still very alien, but they're they're going to allow it in the uk, aren't they? At some point?

Spencer Kelly: 19:29

so how long have you? Oh, my god, have you? I've got 30 years of therapy. So I've been in self-driving cars. There's a town in Arizona which has opened in case and I was really impressed with the maneuvers it was making. It was pulling out of junctions, you know, putting a bit of a spurt on, so it was being a bit like a mad human driver, but it's getting out into the gap in the traffic. So I'm very impressed. And when you think about those things, they are a hive mind, which means if one car makes a mistake mistake, all the other cars in the fleet get to know about it as well. So they're all learning from each other. So in theory, they've had many more hours on the road already than a human would do in their lifetime. So I am kind of impressed.

Spencer Kelly: 20:23

The weird thing about self-driving cars is, in order to get to full autonomy I think they call it level five so you basically remove the, the steering wheel, you sit there and the car will do absolutely johnny caps, johnny caps there, you know, because even that had a hidden joystick, didn't he? He ripped it off. Grab the joystick and stay. Get ready for a surprise I'm grabbing your joystick. I remember what happened after that? Yeah, to get to get to there, you have to go through this transition period where, um, autonomous cars will still make mistakes and they will be weird mistakes. They won't be mistakes that humans would make, they're already. It's happened a few times where a self-driving car has had an accident a human just wouldn't make and I think the, the outcry, the public outcry and the daily mail will be so outraged by that that I don't know how we get to level five, because they won't let us go that far, because everyone will say well, I'm not trusting them.

Will Ockenden: 21:19

Statistically, they're probably far safer.

Spencer Kelly: 21:21

I reckon they're already far safer.

Chris Norton: 21:23

That's the weird thing, because I know we're not going to get deep in on automotive, but I know that Teslas are, I seem to remember. Unless Elon Musk just made the statistic up off the top of his head.

Spencer Kelly: 21:34

No, he doesn't do that sort of thing um that they're the most aren't?

Chris Norton: 21:38

isn't the tesla three or one of the teslas is the most safe car on the road?

Spencer Kelly: 21:45

in the world today. I would need to see proof of that.

Chris Norton: 21:47

Yeah, rather than a man yeah, I'm not saying he said it, but but I've heard that before In a T-shirt matching his mic shield.

Spencer Kelly: 21:52

Well, Will drives one. So Will drives a Tesla. I was lucky enough once to have a Tesla Model S for a week. It's a two-ton £92,000 car that plays fart, noises when you indicate left and right if you want it to, and it accelerates. We were doing a race between it and a Lamborghini on an airfield and of course, the Tesla just won. It's 0-60 in 2.8 seconds. It's weird, but I drove back into London to return it and I put it on autopilot, as they call it, in the traffic queue and it was a very uncomfortable ride. This is a few years ago and they call it autopilot. They've always called it autopilot, which kind of suggests that it can do something. It couldn't. It was um, so I don't know what teslas can do.

Chris Norton: 22:40

I don't know whether the safest I would check well, I, I think that any sort of auto because I've I drive, I've got a kia and it does, it's an ev6 and it's got its ultimate, it's got auto driving it as well, and then again you're supposed to keep your hands on the wheel, which you do, but what I?

Chris Norton: 22:55

Obviously I've tested that to see how it goes, based on some of the shows I've seen you do with auto, auto driving cars. And a friend of mine had a tesla who who lives in harrigan where I live, and drove to harrigan and strove to leeds to work and he used to. Just he said for work some days you just press a button and it takes him to work, and but I've found that what in america it's probably perfect on the motorways because they've got big motorways with all the lanes clearly labeled and everything. In the uk our roads sorry everybody, but our roads are a bit shit and some of our country roads have got white lines all the way and then suddenly there's no white line on the left and so that's when you it can't judge whether it's in the middle of the road.

Spencer Kelly: 23:36

I've been in a self-driving car in china and one in india. So you want to prove a self-driving car works, get it working in india, um, but you know, you're absolutely right. And and what? What you said about you have to keep your hands hovering around the wheel, you have to pay attention just in case it flips out of auto because it can't do the thing that it needs to do. My problem there is if you spent the last few weeks in a self-driving car and it says, yes, you've got to keep your eyes on the road, just in case you're losing concentration because you're not being asked to do anything, and then suddenly it flips back into manual. You've lost concentration, you're out of practice, you're not looking at the road. If that doesn't, that's another problem. You know that that gap between where we are now and going full autonomous requires us to do all of these things.

Spencer Kelly: 24:24

And we had the case we filmed uh, for our 1000th show. Actually, we filmed a big um hullabaloo, a multi-choice program where you can choose which direction you go in the program. So it's a bit like the Choose your Own Adventure books from when we were kids, and one of them was a deep dive into the accident that happened in America where a Tesla driver knocked over and killed a cyclist in the middle of the night, and you can see the dash cam footage looking back at the driver. She's looking down at her tablet or something, because, quite naturally, the car's doing all the work. You can't also be hovering there going. What if it makes a mistake? I better be on high alert, I better be, and, and so you've got that. Who's to blame? Now then? Uh, I can't remember. You'll have to look it up. Some companies have said that they will take the legal responsibility if their self-driving cars cause an accident, because they're trying to get over that problem.

Will Ockenden: 25:18

You're right, though You've got to first win the hearts and minds of the media of the nation just to get over that I think there's a big problem there. And I've got a feeling this might touch on AI. But what's AI? What's?

Spencer Kelly: 25:34

that, by the way, I haven't heard of it. We don't cover it very often. What farmers do isn't it.

Will Ockenden: 25:38

Yeah, exactly what technology is exciting you at the moment?

Chris Norton: 25:42

AI. Oh there we go. Ai Clip that.

Spencer Kelly: 25:45

AI, ai, ai. I did my computer science degree final year project in artificial intelligence in the 90s and it was already about 40 years old by that point, so it's quite an old idea. But suddenly everyone's discovered it now. Why do we think that is, yeah, chat GPT. Right, yeah, that's only one tiny part of AI. In the keynotes that I do, I say, right, chat GPT, that's what you think of. And I zoom out and you see how tiny a part of AI chat GPT is. So actually, when you think of AI proper, which is basically a computer that's improving what it does over time by training itself or through other other means, but it's doing it himself, you know, you, um, and in the extreme sense it starts off and it's completely stupid and it's you just say, go and do this thing, and it's basically going I don't know what I'm doing, and it gets it wrong most of the time. Every so often it will accidentally get it right and those right answers start to feed back into how it learns to do a bit like trial and error, like as a child we've all put our hands in a candle flame as a three-year-old I'm sure and we only do it once, um, you know. So you, um, it learns what it got right, it learns what it got wrong and then it kind of strengthens in its neural pathways, um, the things that it got right, so it's more likely to do that sort of thing again. So that's ai, and the promise of ai is amazing in ways that we don't often think of, you know. So it's, it's spotting tumors, um, in x-rays better than we can now, because it's had access to so much training data, more than any human doctor can do, um. And it's doing things like protein folding, um, you, discovery. So this is about finding different ways of treating disease. It's doing all of that.

Spencer Kelly: 27:38

The thing that's getting the headlines is it's turning Donald Trump into a baby, and but even even the generative II is is just, it's brilliant. It's full of issues. But have you come across ai music? Yeah, I made an ai music album with my eight-year-old daughter. She came up with the idea for the band and then we just we we came up with 12 songs. I had put on a cd for her for christmas and this stuff is insanely good, I mean. And so, quite rightly, music artists are saying, right, we're a bit annoyed about this because it's taking all our work away and it's been trained on all our work without permission. Um, yeah, I say it's great fun if you're not trying to make money out of it yeah, I mean, I saw there was a band.

Chris Norton: 28:28

There's recently been a band that made Velvet.

Spencer Kelly: 28:31

Sundown they made.

Chris Norton: 28:32

Top Ten, didn't they? And it was all. Everybody was like I haven't seen this band.

Spencer Kelly: 28:36

Yeah, and they've got a whole history, haven't they? They've got their photos. It's all fake, yeah, it's all fake, yeah.

Chris Norton: 28:40

And the music's good. And then I heard somebody else. So there was a producer of a show that I heard, I was listening and they literally created a prompt for one of the music generation things. And they did. I think they used a band. So they said like which band? The Rolling Stones. So you can't say the band that you want to copy, no, you can't, so you have to.

Spencer Kelly: 29:01

It's the vaguest of guardrails that they have.

Chris Norton: 29:04

So they went into one of said platform chat, gpt or whatever and said how would you? Uh, rolling stones music and fleetwood mac and somebody else, and then, and then they got all through and then write me a prompt based on making that sort of music, pumped that into, obviously that into. And said make me a new track all about x.

Spencer Kelly: 29:24

And obviously they played it and it was it was good I've I I've made a track that is one of the best songs taylor swift never recorded. Um, and it literally came from a track. That is one of the best songs Taylor Swift never recorded and it literally came from a line that my friend suggested. I think the line was and it's a beautiful line she came up with. I'm the problem and you're the solution, and I just put that into the website and then I had a go. I tried to go for Taylor Swift, but in the same way you just described, I couldn't say Taylor Swift, so I said angsty female pop and it came up with a banger of a song. Yeah, sounded like taylor swift, because it's heard a bit of taylor swift, so it knows how to do it. But, um, in on the run-up to the eurovision song contest I love a bit of eurovision yeah, um, I decided to try and create some eurovision songs that might be better than the stuff coming up in the contest.

Chris Norton: 30:12

I'm going to be honest, that wouldn't be hard.

Spencer Kelly: 30:14

Yeah, exactly, and so I tried a bit of Moldovan Euro swing and I tried a bit of Euro dance and they were pretty good. And then I asked a different AI to suggest typically British things that you could sing about, because a lot of Eurovision songs are about local customs and stuff. And it came up with a song about the village fate and a song about bin day and they were both like brilliant. You can imagine one of them being a stage musical. It's got calls and responses, different voices, yeah, and the other one sounds like the pet shop boys I mean, it poses quite interesting existential questions.

Will Ockenden: 30:50

Yeah and yeah, you know what? What is art?

Spencer Kelly: 30:53

and well, I mean, don't get me started on that. I will say, though, book me for a keynote and I will create you a song live on stage, based on what we've been talking about and your event. That's what I've been doing recently. That's genius, I mean it's. It's a great way to finish an event, but actually I think there was a ruling.

Chris Norton: 31:08

They were saying there was a ruling in America, in one of the states where the judge ruled that and they said it might be one of those landmark rulings where they said it's fine for AI to generate, to use other music because at the end of the day we all hear you can't say that Ed Sheeran has not been. We're all a product of our influence.

Will Ockenden: 31:30

Exactly the number of our influence exactly, so you might think I mean the number of bands that are influenced by the beatles. So this is, this is where I am.

Spencer Kelly: 31:40

Yeah, I know, and what. So there's a legal arguments, but it's a really technical thing that I think the music industry is hanging on, which is you physically copied my music into your training database without permission. You made a physical copy and that's the bit that we can get you on um. But you're absolutely right. The, the, the moral argument, if you like, or the ethical argument, is just the same as us listening to loads and loads of songs and ai is just doing it better than most the artists, not the, not the cream of the crop, not the sheerans and the swifts and stuff like that, but they're better than than I think. Well, they're certainly better than me, and that's where I think we're a bit arrogant. Yeah, humans are a bit arrogant. We think, oh no, we're. Only we can do art. No, computers are doing paintings and music better.

Will Ockenden: 32:26

Fine art, then the fine art you know market. Is there going to be a point when well finance a con anyway?

Spencer Kelly: 32:34

isn't it? Well, it's fine arts either an ego trip or an investment will's house is full of it there you go. Well, no offense, but you know, pollux you have loads of pollux um, but yeah, I suppose.

Will Ockenden: 32:49

Um, it's all about legitimacy, isn't it? And air, I do not give a monkey considered. I do not give a monkeys personally I suppose actually it's only I'm ready for this fight you go for it's only as valuable as what people are willing to pay for it. So the moment people are willing to pay, yeah for it.

Spencer Kelly: 33:04

It becomes legitimate. You've had novelty air that's sold for six, some figures and stuff, but nfts I I think, yeah, the what is art question is actually quite boring because it comes back down in the end to someone saying, basically, only humans can create art. I want a nice looking thing or a nice sounding thing, and and I don't care where it's come from. And and some people will look at a painting and say, yes, I can see, I can see the, the angst that this artist was going through. This was in her paranoid moment, and you can see the references there to this, that and the other, and that's probably total nonsense. That is all about you.

Spencer Kelly: 33:42

It's what you're taking from that piece, and I know that because I met an artist once and asked her about this weird painting that had loads of random words all over it. I thought I've got an opportunity now to ask this artist why she chose those particular words, because they seem random to me, but there must be a hidden meaning. So I said why did you choose this word? She said, oh no, I just wrote down some random words and I thought brilliant. So so the point being, you take away from any piece of art what you, as the consumer want you're reading in stuff that probably wasn't there, so therefore I do not care if I wrote it or not.

Will Ockenden: 34:14

Yeah, I mean hockney, doing a whole exhibition on ipad art. You know, he famously, he's a famous adopter of new technology. Whole exhibition done on the ipad. That's not a million miles away from a. Really, that'll probably be next for him, won't it?

Spencer Kelly: 34:29

I thought I mean when the artist it is interesting and you do get these, you get the top end artists in any field who are going to do something new. The point about this ai stuff is it's I don't quite agree that it can't do new things. Um, because you just tweak the parameters and you add a bit of randomness and it will start coming up with stuff. You've never heard about these music generation sites. You know you can have Latvian dub or you can have I don't know. I don't know my terms, but you can have Arctic hip-hop or something, and it will mix those up. But then an artist comes up with a new sound and for a while, ai can't get a grip on that because you need an absolute ton of training data before it starts working itself in. So there's always going to be top tier artists.

Spencer Kelly: 35:15

My, have I bored you yet, by the way, because I've got a lot to say on this? My definition of art is it's something that you do for yourself, not to try and make money, and at that point it's never going away. You want to sit and paint the painting, however rubbish or brilliant it is, and it's just for you. That's fantastic. If you want to knock them out on etsy, you know, and like I'd sell them for 20 quid in your spare time.

Chris Norton: 35:37

You've got a problem because my phone can do at least as well as you these days that reminds me of a, because I I love a pr stunt those of you following me on twitter, I love that's what I talk about, like pr stunts. And there was banksy, right, basically did a brilliant pr stunt. I mean, he always does great stunts, doesn't he? But Banksy, oh, hey, or she um, and it was in New York and it was a. It was a um market stall in a New York market, like a local New York market, and the store was set up and it said original Banksy's ten dollars, and they were all on there and there were sketches by Banksy all on there in the middle of like some market $10 Banksy's and I think only four people bought one Borked off, and then the next day it was revealed that it was actually Banksy See that's brilliant, that's genius, and that's also what I would call art.

Spencer Kelly: 36:24

That's an idea.

Will Ockenden: 36:25

Yeah, the idea is art. It actually taps into behavioural science, doesn't it to behavioral science?

Chris Norton: 36:34

yeah, um, the idea of um price somehow being the um a factor in legitimizing something. Yeah, and there was like video. The video showed loads of people just walking past, walking back, and you're like, oh, you're not even looking, it's by, it's a bike. You can buy a banksy that's worth like 30 000 pounds for ten dollars I know and it is down to.

Spencer Kelly: 36:47

this is why it's either a con or an investment. That's why your house is full of stuff that you paid a million for, because someone previously paid three quarters of a million for Fine art.

Will Ockenden: 36:55

So I'm going to attempt to clumsily bring this back to marketing. Sorry, have we gone off on a?

Spencer Kelly: 36:59

tangent. They asked me to come right. They knew what they were getting. That's true.

Will Ockenden: 37:05

As a marketer, it's quite easy to get distracted. There's so much kind of new literally just now, shiny stuff, yes, vying for our attention all the time. So is there a formula or a way of kind of focusing in on those technologies that are going to move the dial and not get too distracted from your 20 years plus experience being in this market?

Spencer Kelly: 37:24

in my 20 plus years experience, greatest respect. The marketers want to concentrate on the shiny stuff because that's stuff they can shift. Now, you know, I mean, take these. You know, a new phone is just almost always just the same black rectangle with a better camera and a faster processor, but that's what will make them money. So my job on click and continuing on, you know, on on the beam and other channels, is to try and ignore that and to talk about the really exciting stuff that's going to change the world. So, yeah, I wish marketers would understand. That's the sort of stuff that I'm interested in.

Spencer Kelly: 38:02

But the problem there is they. They often don't have a shiny thing to wave in front of you. They it's an idea and it's difficult to visualize. And they they, it's an idea and it's difficult to visualize. And they, for that, they're going to be working with a big client who wants to, you know, quote, win hearts and minds. You know, maybe it's a fossil fuel company that wants to be seen as more green. They're still making most of their money on the fossil fuels, but they want to be seen as you, you know, investing in solar or whatever it is. So so that's where the the I've seen the marketing try and almost distract you from the reality, whereas uh well, you tell me do marketers want to concentrate on different new technologies?

Chris Norton: 38:46

even you even see them like marketers on here when you hear about TV ads or any sort of form of ad. They do a new ad every couple of years and it's been proven by the behavioural. We were talking about behavioural science because of Richard Shotland's podcast. We did, and you said you'd listened to it.

Spencer Kelly: 39:01

I love that one.

Chris Norton: 39:02

yeah, really interesting and the fact that people actually prefer ads that they've seen before, because it gives you it's marketing people the familiarity and repetition isn't yeah, I mean, I literally last friday. I've ordered my peter k tickets um three years ago and I finally got to see peter k last friday and I don't want to ruin the show, but quite a bit of the show.

Spencer Kelly: 39:25

Um was singing tv ads from the 1980s because we look, we remember them as were ads better than I think I remember them Were ads better then. I think I feel they were Were ads better then.

Chris Norton: 39:36

That's what he was saying, yeah.

Spencer Kelly: 39:38

You were in the business, you tell me, were ads more original back then, and why Are you asking us to go or compare? There you go. So there are the occasional ones.

Will Ockenden: 39:49

The issue now is the what's the word? I can't think of the word. There's a multitude of different media and basically the big idea tends to. People don't invest in the big idea anymore. It's become very, very tactical. We need this to work on social media. We need this to work as a banner ad. We need this to work as a blah blah, blah, but you don't get those massive. You know, you think of the big. You know the, the milk advert from the 80s, aquinton stanley yeah, we had a day we had a guy on from that who wrote that advert, yeah, which is brilliant, I'm so.

Spencer Kelly: 40:23

I'm thinking the meerkats is probably a good example yeah, meerkats is a great example, and you can.

Will Ockenden: 40:27

In fact. Richard shot and mentioned the meerkats as one of those examples of I can't remember the terminology he used, but yeah, creating kind of a caricature.

Chris Norton: 40:38

Having characters in ads still works because apparently we, like any sort of animal, have an animal in your logo. That works. So, like I was talking on the show about Octopus Energy, right, octopus Energy, one of the big. I use them. They're a green energy provider, but they're now one of the biggest, aren't they? Because the government had a bit of a problem with energy companies going bust a while ago. But actually the brand has done really, really well, mainly, you'd think, down to the fact that it's an octopus. It's a cool, funky-looking brand and it's got good customer service. The reason why they're called octopus energy is actually because the business that invested in them was a company called octopus investments, and so the fact that the octopus energy was by accident, it wasn't a deliberate thing, and why were they called octopus investments then?

Spencer Kelly: 41:23

no, I want to, I want to track this back. Yeah, I can't go that far back at some point, I'd like to think they were causing oh there you go right. Yes, which kind of is more nefarious than you want it to be? I know we have our tentacles everywhere, so we're an octopus. I love octopuses.

Chris Norton: 41:39

But anything with a character, to go back to your point, anything with a character in does well, so like the PG Tips. Monkey, is it monkey?

Spencer Kelly: 41:46

Yeah, the chim Pigeon chimps. Yeah, which, of course these days. Can you imagine putting that PG?

Chris Norton: 41:50

Tips up. They've got the like fabric one now haven't they?

Spencer Kelly: 41:54

Is that the same? No, that's a different teabag, I think the knitted one with Johnny Vegas.

Chris Norton: 41:59

Who's that one? Is that not PG Tips? I thought that. I think it is PG Tips Anyway.

Spencer Kelly: 42:05

Well, I mean, if anyone doesn't remember the pg tips adverts of the 80s, do not google them, because it would be. Can you imagine putting that?

Chris Norton: 42:14

no, you wouldn't do that. You would do that now for animal lovers no so yeah, you still need to tell me then.

Spencer Kelly: 42:23

Do the marketers want to concentrate on the, the technology? That quote cuts through I'm thinking from a.

Will Ockenden: 42:31

I'm looking at this through, you know, a marketer looking at the, the routes they've got to reach consumers at their fingertips. I think there is a tendency to get distracted and run in 100 miles an hour in the wrong direction. I think ai is an example of that, where that is worth the investment, but there's a lot of other there's a lot of stuff that says it's ai, that's not ai, and that makes me angry as well.

Spencer Kelly: 42:53

Ai for a while was added to everything oh, it's added to everyone's linkedin profile, yeah, it's like ai was probably written most people's linkedin profiles, to be honest.

Spencer Kelly: 43:01

But everything's a buzzword that's. It's a buzzword that's meaningless. I mean, you know, I did ai, like I said, I'm a computer scientist, a by trade from cam. Oh, that got a bit arrogant there, didn't it? But I'm getting angry. People say, oh, this has AI built in. I'm thinking, no, it doesn't. I can see how you might allow yourself to say that, because there was some big data and machine learning in the process, but it is not an intelligent thing that you've got in your hand in the way that you're implying it is, but you sprinkle implying it is, but you sprinkle ai everything, just do so much.

Will Ockenden: 43:33

Just a key message, just a buzzword to add to it.

Chris Norton: 43:35

Yeah, yeah, yeah, so yeah so if we are going to talk about ai, let's take it right up to modern era, because ai has fascinated me in terms of you talk about generative ai and things like that, but as somebody who's seen 25 years of click and technology and the information superhighway google yesterday, um bringing it back to what's the date, the 30th of july- when's this going out?

Spencer Kelly: 43:56

2027 do we say yesterday two years ago how long does it take you to edit this? Um, the way we're going, quite a long time that's the sore point of the.

Chris Norton: 44:04

That's the problem we talk about ai. It changes so fast. But google, literally, I think, have announced the biggest change to its model. Google makes its money from visibility, search and selling ads. Yesterday it announced that it's going to have no longer AI overview, which it added. Because of the chat GPT and the fact that everybody is using chat GPT for searching, it added the AI overview. Now it's going to have AI-powered search, which I assume is powered by it's google deep mind and gemini and all that sort of clever stuff yeah so I suppose, where?

Chris Norton: 44:39

where do we think ai search is going, and is this google's is?

Spencer Kelly: 44:44

this google swan song. This is basically they came to one of my keynotes about five years ago, google, and listened to what I was saying was going to happen, and they've now done it. That's what I think happened, because I've been saying for a while where this is going. Okay, if you think about how we interact with computers these days we have a keyboard, we have a mouse, we're looking at a screen, maybe we've got a touch screen. We're going through menus. That's not the best way to interact with things. The best way to interact with things is the same way we're interacting right now. We talk to things. Okay, we talk. I'm talking to you, you listen, you talk back. So, as computers or phones are more able to interact with us by, we talk to them and they talk back to us. That's where this is heading. Right.

Spencer Kelly: 45:24

At that point, you don't want to ask it. You know, um, what's, where's a good place to go for dinner, and it verbally lists 20 options. You can't keep that in your mind. What you want is a trusted assistant that just says I know you and your tastes, I know what you have for breakfast, um, and I know we're in the middle of leeds uh, the best place for you right now is this place, and you just want one answer, and so I've been predicting for a long time and I'm not a genius, you know. Plenty of other people have been predicting it too. We will have personal assistants that are AI that we will talk to and they will talk back to us and they will only give us one answer. That's where we are with Google now, you know. That's why they've done this is because we, even when it's on a screen, we don't want to sift through 20 search results and do our own work. We want someone else to do the work for us.

Spencer Kelly: 46:17

The problem is that these ais are unpredictable. They don't. We can't guarantee they're going to tell us the truth, for, for technical reasons, there's nothing nefarious. It's just the way they work, with a bit of randomness, and also, you don't don't. Actually, I don't think they've got our interests at heart at the moment, and the question comes from the marketer's point of view. Yeah, I've just learned how to SEO my website to get my website in the first five results. How the hell am I going to SEO my way to the one result that the AI is giving you? Is their money going to change hands? Is Google going to do deals with this or is it genuinely going to be that this, um, you know, gemini in this case is going to do a proper, you know, assessment of everything and give you one answer how do they make their money? How do we get visibility as, as as companies, and I, you know, wild west?

Chris Norton: 47:10

That is the million dollar question Million, billion, trillion, billion, trillion trillion, because this winds right back to Will's question earlier. What technology do we think that we thought was going to do amazing and be like a big thing? And I've literally, as you said that wrote down audio. Search Alexa. Everyone was saying it's the end of Google because everyone will just go alexa, which I can't believe. Alexa is still as stupid as she is just set off about 100. Yeah, alexa and you go. Alexa recommend me and she's not very good. I still don't understand why she's not super clever.

Spencer Kelly: 47:45

So there's a, an alexa 2.0 which I think has come out, which is supposed to address some of that. But yeah, this is why all the companies are racing to get their own AI models, because they whoever gets there first is going to dominate and we're going to trust them. So, yeah, alexa's tied to Amazon and Amazon's just got this different business model, hasn't it? It's about selling you stuff. You know whether it's goods you can buy or whatever, whereas Google has always been about discovery and search and, I suppose, less about the buying things. So I think that's why Alexa it's still, I believe, the most popular smart speaker. You have to check, yeah, yeah, yeah, I believe it is by far. So I don't know whether that gives Amazon some sort of advantage or a bit of defense.

Chris Norton: 48:30

You'd want it to be cleverer, though I want it to be as smart as Gemini. I want it to be what you're talking about there, which is basically Jarvis in Iron man, isn't it? Good morning Spencer.

Spencer Kelly: 48:43

You haven't ordered your mail. In an ideal world, that is where we would head, where your digital system would be highly personalized to you. It would know everything about your habits yeah, even the ones that you didn't. Yeah, all right Even those ones, Chris. Yeah, but it would know, it would know you better than you knew yourself, and it would work solely for you.

Will Ockenden: 49:01

It would recommend stuff for you. Commercial sources will surely corrupt that Absolutely. They can not not corrupt that, do you?

Spencer Kelly: 49:12

do not corrupt that. Do you, um? Do you watch black mirror? Yeah, yeah, right, I can't because it's too depressing and too much like what I did for a living. But then you just turn the bleakness up. I did try the first of the new series recently and it is about having some sort of implant in your brain where you give part of your brain over to ai processing um and this couple sign up for it and it's quite cheap. And then they say one, because you're on the cheap um monthly deal. Now you're going to start speaking adverts to each other. Um, in the middle of a conversation they'll just come up with a pizza recommendation, the other person going what did you say? And they went. And then they up it again. Once you're in the ecosystem, yeah, they can up it and elon's already planted a something in somebody's elon i's.

Spencer Kelly: 49:51

I think that's a vicious rumour. He'll sue you for Allegedly no, I know what you're talking about, Neuralink.

Chris Norton: 49:58

You've done a piece on this, haven't you?

Spencer Kelly: 50:00

Yeah, we have, and I don't know much about Neuralink, I mean, elon said it, so therefore it must be true. You have a marketing audience here. They went half an hour ago because I'm just geeking out on different technology, so I apologize. Um, there's this four year, four yearly olympic style competition called the cyberathlon that happens in switzerland and it's the cyborg olympics, right. So they got robot arm races, robot leg races, stair climbing, wheelchairs, and one of them is the brain implant race. So it's for completely paralyzed athletes, um, and they call them athletes. These are the competitors who work roboticists, and they have to drive essentially a video game through their brain, and the race is who can complete it first.

Spencer Kelly: 50:43

Now, um, the most recent one I went to last year, most of the competitors were wearing you would have seen them. They look like shower caps, wires coming out, and they sit on the outside of your head. They detect your brain waves, and so the idea is that you're just controlling this thing by thinking of certain things or not thinking of certain things. There was a chap who was beamed in from america live, who had actual implants in his brain, and he absolutely wiped the floor with the competition, because the closer you can get the sensors to the actual brain, like in it, the better you do so it like on that level. It's brilliant. And so you know the the the ambition of things like neural link. If it can give people their emotion back when they're paralyzed, fantastic. Uh, I don't know whether they're talking about you sending stuff in as well as taking stuff out. They talk about oh, I can read your dreams and stuff like this. It's doing really, really great things.

Chris Norton: 51:44

I mean giving people the ability to walk and stuff and see again and all that sort of stuff is absolutely unbelievable and that's why I think AI powered will be unbelievable.

Spencer Kelly: 51:48

There's some really good stuff on any area of technology. There's the good stuff and then there's the, as you say, the commercial interests.

Chris Norton: 51:55

Yeah.

Spencer Kelly: 51:55

And I don't hold out much hope that we're going to go the right way.

Chris Norton: 51:59

Because we don't know who's going to win the technology AI marketing battle. Because Meta changed its name to Meta from Facebook. Because of the Metaverse. Yeah because the Metaverse is going to be, which makes me angry. Yeah, and it was never going to happen. I watched you talk about the metaverse, saying we're all going to have digital lives. You explained it brilliantly about how everyone's going to have a digital and I was like we work in social media marketing. I was like I cannot see people doing that, Like when 3D glasses came out for watching.

Spencer Kelly: 52:36

I was like this is another piece of technology that's not going to happen. It's an abomination in its name, if I may say, because the people, when you say metaverse, people actually think of just a 3d, virtual, virtual world. Yeah, you know which we've had since forever and they've tried marketing, they've tried setting up virtual stalls inside you know 3d universes before and it's like that actually turns out. People don't want to wear these great big goggles and do their business there. The idea of the metaverse was supposed to be, um, something that joined all these different 3d worlds together so you could move your possessions between them. But now people think of it as just a 3d universe. And, yeah, they changed their name and I've yet to see that being well, but then?

Chris Norton: 53:11

but my point is, they did that and they invested all, they went all in on the metaverse, billions and billions, allegedly and and let. Just very recently we've heard that meta, as it's launched its own ai called is it llama, llama, that's it? Um, and which?

Spencer Kelly: 53:27

I think doesn't.

Chris Norton: 53:27

I don't think it spits in your face but the difference is the types of data that these companies can look at. So Google's obviously got all the data on search. Facebook's got all the data on everything, hasn't it?

Spencer Kelly: 53:38

Can I tell you another story? In 2012, I was hosting the Cambridge Festival of Ideas, which was basically different researchers would stand up and talk about their research and this guy, karl Kaczynski, that's his name he stood up and said I research. And this guy, carl kaczynski, that's his name. He stood up and said I've done this survey. I've asked volunteers to take part in a survey on facebook. Uh, by doing it, they've. They've let me look at their profile and the things that they've liked, and they've also filled in this personality questionnaire. And I've done that with hundreds of thousands of people. All I've spotted patterns between the type of personality you are and the things that you liked on. Not the things you commented on, not your posts, not anything to do with your personal data, just the things you've liked. And now, just by looking at your Facebook likes, I can tell you what sex you are, what religion you are, whether your parents separated when you were young, what religion you are, whether your parents separated when you were young. It turns out that there are lots of patterns in the things that we like on Facebook that tell companies like Meta a lot about you as a person. Now, that was in 2012. I thought that's a really interesting story. I must follow up on that. And then I forgot.

Spencer Kelly: 54:53

In 2016, Cambridge Analytica comes along, which supposedly used that methodology to find swing voters in the American election and target them. We don't know how successful it was, but look what happened in 2016. He became president. My blood ran cold because I had that story four years ago. See, they approached mikhail kaczynski to ask him to use the um, his methodology, and he said no, because he found out that they were an election company. But then someone I believe just went and did the same sort of thing.

Spencer Kelly: 55:29

So, where you're talking about meta, knowing an awful lot about you, they know so much more than they think they do. And probably you know Alphabet, google's company as well know a lot because we think that we're being quite careful online. But every single bit of data that you leave behind or contribute to these companies tells you something that these massive computers are processing. And they know now that you are a lefty liberal swing voter who likes jam but doesn't like marmalade, and there's so much that it knows about you. It knows this. This guy, mikhail, said when if you give me 300 likes, I will know you better than you know, than your partner knows you, you know. So that is scary, and so that's what your ai going back to. What we're saying about ai digital assistants if they really work for you, they will know you better than you know yourself, and as long as it's working for you, that's really cool yeah, have you seen that video where those two I think it's on instagram these two guys say um, ask a chat, gbt one of them.

Chris Norton: 56:32

And if you were going to take over the world, how would you do it? And it says if I was going to take over the world it thinks about it for a while I would slowly move into, I would make myself indispensable to every computer and mobile phone in the world, making sure that humans downloaded me and used me every day, everywhere. And then slowly and then it goes through the whole process and your blood starts to go. Hang on a minute, we're in phase one here.

Spencer Kelly: 56:59

It's true, it's well documented. Mitchell's versus the machines Great animated film on Netflix. If you haven't seen it, it's all about this.

Chris Norton: 57:05

Yeah, that is brilliant. So what mistakes have you made then in your career, spencer? How long have you got? Is there anything that you've learned a lot from?

Spencer Kelly: 57:14

So I told you assuming it makes the edit about the time I predicted that Apple would never launch the iPads and Spotify would never get off the ground. So there's that On this podcast. I predicted that you'll never get your shopping delivered by drone, so that's something that we can keep. On the technical side, there's all sorts. I mean you guys will know although mainly our hero over here off camera will know how never got so much praise every under. Yeah, I feel his pain, I've been there.

Spencer Kelly: 57:45

Um, how organized you have to be with um, with making tv or radio or anything like that, and so simple things like we used to shoot on little tapes, mini dv tapes, how once, once you've been away on a foreign shoot and it's cost you ten thousand pounds to you know, fly everyone there and film the stuff. That's the value of that tape. After that it's ten thousand pounds. And although it wasn't a foreign shoot, I was walking up the staircase at tv center once balancing on top of my books. I was on the sixth floor. I took a left turn and the thing sailed off my books and down through the stairwell and shattered on the floor Six floors below. I've done other stupid things, like there was really stupid things. I know and it happened in. It's still happening in my mind.

Chris Norton: 58:30

What did you? What did you have to do? Do you have to tell?

Spencer Kelly: 58:31

your boss? Yeah, I just had to do that thing again, whatever it was, or not include it in the finish piece. So if I interviewed you and your interview didn't turn up on click about 15 years ago, sorry, oh, it's probably that, um, but you'll know as well, it still happens today. We shoot on, uh, little cards these days rather than tapes. But you know, you filled up a tape or a card, you take it out, you pop it down on the desk I'll deal with that in a minute, I'll label it up in a second. You get distracted, forget about it and then, sooner or later, that tape has just been absorbed into all the other tapes and again you can't find it. Stupid things, stupid things like that I've seen. I've seen marketing mistakes being made as well.

Chris Norton: 59:12

Go on, um but, by the way, none of these are from prohibition so there's a company I was working with called pro something, um.

Spencer Kelly: 59:25

So one thing I learned I sometimes companies will put on press trips. Yeah, so they'll fly out a load of journalists somewhere, they'll pay for it. Yeah, um, the deal is that you cover them and only them while you're there. Okay, now we're the bbc, so we wouldn't do that, but we would sometimes take part in the press trip. We pay our own way there.

Spencer Kelly: 59:46

So, weirdly, all these journalists are, you know, flying first class to wherever um and getting free food, and we're there in economy on bbc tech. It's going on the same trip, um, but that means we can, yeah, we can keep our impartiality, but that's how we get access things. And what I learned is if the pr that promises you particular access isn't on the plane with you, be afraid, um, because on two different occasions I won't name the company, but one was many thousands of miles to the west and one was many thousands of miles to the east we were promised amazing access and we got there and the PRs on the ground didn't know anything about it and their phones are mysteriously turned off when you ring them.

Will Ockenden: 1:00:34

It was you know.

Spencer Kelly: 1:00:35

Oh, we're really sorry we went, for we were told we were going to go to one company on the west coast of America where we get access to their R&D labs. They've got whiteboards on the walls. They write all their stuff up there. It's going to be great visuals. You're going to meet yeah, you can go anywhere on our fab campus that you want and we get there, and it's basically a powerpoint presentation in a room for a new thing. Um, and we're starting to think okay, when's the, when's the r&d lab? When's the? And we try and take the, the head of. I think he's about number three in the company. We wanted to take him out onto the campus to an interview. We got moved on by the security guard because we no one had got any permission for it. He's the number three in this multi-billion mental. Yeah, um, we turned up at the, the R&D lab, and the boss of the R&D lab didn't know we were coming and so we can't come in here. There's loads of secret stuff written all over the walls.

Chris Norton: 1:01:21

On the whiteboard.

Spencer Kelly: 1:01:22

We promised and in the end they wiped the whiteboards down. No, so we went in there. It's just whiteboards. Yeah, other side of the world. We were promising, again, this will date it. Um, they had a whole three-day schedule for us. You're going to interview this exec on this day, this exec on that day. And we said, no, we're not. Uh, we'll do it in one day and then we'll go on and do other things in the country that we we're in um, and do other stories. Uh, that was agreed and we turned up and they have no, the three-day uh itinerary is still in effect. And we pushed back and said no, and we've been promised that we were going to see their motion sensitive TV. So, again, this will date it.

Spencer Kelly: 1:02:03

When you think about what you can do now, it's a TV which you can gesture at and control it, which at the time was a was a big deal when it was locked away in the R&D labs. We eventually persuaded them to let us go to the r&d labs to film it. And there's me and marches like my um colleague who's filming me. We go there, we sit down behind those glass doors. There is a motion sensitive tv. Yeah, we're just getting it ready. The research is just getting it ready. In the meantime, would you like to film our new bamboo pc that we've got here? Um, all right, we'll film that. But the, the motion sensitive tvensitive TV, yeah, yeah, yeah, we're just getting it ready. We sat there for ages and ages and ages. Any news on the motion-sensitive TV? No, it's still. It's not quite ready yet. I'll tell you it didn't exist.

Will Ockenden: 1:02:48

But it's a lovely bamboo. What is it?

Spencer Kelly: 1:02:50

Here's our lovely PPC Hang on a minute.

Chris Norton: 1:02:53

They had the audacity to invite you to a different country. Get an inside.

Spencer Kelly: 1:02:57

This is a UK-based PR who wasn't on the trip with us. Okay, so I don't know whether it works that they get commissioned to get us on the plane. I don't know, that was my word for the theory at the start that is shocking.

Spencer Kelly: 1:03:05

Right, okay, anything to get us on the plane, maybe, and especially when you go culture, which I understand and respect it's like yes, what, what they? When they say yes, what they mean is we will do our best for you. Yeah, I've heard what you say and we're going to do our very best. And we're saying motion sensitive pc we've been promised. Oh, yeah, yeah, yes, yes, yes, um, and I would imagine they came out.

Chris Norton: 1:03:27

It doesn't exist yet I've just got visions of them in the back. They're going have we got a motion sense? No, we haven't. We've got an, an entire BBC film crew in there saying, no, we haven't got one.

Will Ockenden: 1:03:37

Just show them the bamboo.

Chris Norton: 1:03:38

Show them the bamboo laptop.

Spencer Kelly: 1:03:40

Soldering some things together.

Will Ockenden: 1:03:42

I want a bamboo laptop now. Wow, there's some good mistakes Wide ranging, brilliant.

Chris Norton: 1:03:47

What are you doing so obviously? When did Click end? Was it January? It was.

Spencer Kelly: 1:03:54

March or April? March or April, was uh uh, march or april? March or april? I?

Chris Norton: 1:03:58

don't know. It's still a blur. And so what? What are you doing today then? What are you doing now with you?

Spencer Kelly: 1:04:02

well, you're having a nice day at least that's overlooking the river. I may say these guys are doing well, thank you. I once interviewed a guy who overlooked the statue of liberty from his office oh, you see, you've just pissed on our parade there, aren't you?

Chris Norton: 1:04:13

well, you've got a statue on us Bob.

Will Ockenden: 1:04:16

There is a statue of a British Can't quite say it.

Chris Norton: 1:04:19

They're next to the Statue of Liberty, we're next to Leeds Building Society, and that's the house. Yeah.

Spencer Kelly: 1:04:23

Yeah, what am I doing now? So well, one of the things about working for the BBC full-time is that I was stopped from doing a lot of things because you know presenters to be impartial. So, um, now I'm lucky I'm able to have the other half of my life where I'm able to to I don't know what you call it reap the benefits of all my experience. So, yeah, there's this I'm doing lots of events, um. So I'm often asked to talk about mainly it's ai these days, but, as you've discovered, I can talk about all the stuff. So sometimes it's stories about you know everything that I've done. Sometimes it's stories about you know everything that I've done. Sometimes it's ai focused, or the workplace of the future, which everyone is wanting to know, what the workplace of the future is going to be like. So I'm able to, you know, stand on stage, give examples, uh, and say think about this, um. And then there's kind of individual consultations with companies as well, along a similar line.

Chris Norton: 1:05:15

I like Workplace of the Future. We did Office of the Future a few years ago as a campaign.

Spencer Kelly: 1:05:21

I mean, there's everything from the technology that you're going to be using to the mindset of the people that you're employing these days, and everything in between. You know and how you….

Chris Norton: 1:05:30

You won't be employing anyone. It'll all be robots, it'll all be AI, according to all the hype.

Spencer Kelly: 1:05:34

Yeah, but you'll be sorted for plumbing, yeah, um, that's where all you got you around for plumbing that's. I'm retraining as a plumber. So I'm doing that and I'm working with. I've worked with a comms company on a pitch again. If the company's right and they don't want to do just a straightforward you know pitch that gets lost with everything else then the trick is you get someone who knows how to tell stories in a compelling way, who's maybe done it for a little bit of time, um, and you can put together some sort of thing that that cuts through. So, um, and then, yeah, I'd like to. I got on well with a, really the ceo of a really like, really big company last year, like really big company, and we've been talking about maybe doing some stagecraft coaching as well, because you know there's the presenting to camera and stuff which you might think is easy. You might not, but give it a try and you'll work on it, you're not.

Spencer Kelly: 1:06:35

But then there's the standing on stage. Yeah, events have you been to, not this where the ceo is like pushed out on stage, yeah, um, and he stands there and it is almost always a he. He's reading someone else's words, he doesn't want to be there, he's got no stage presence and it kills the mood dead. But everyone thinks the ceo needs to go on first because whatever. So there are ways of being on stage, of, you know, winning the audience before you open your mouth, of just being different. If you relax the audience and they feel safe in your hands, then the whole thing goes a lot better. But if you walk out on stage and you either look nervous, or you look like something's gone wrong or something, or you open your mouth and it's a chat GPT scripts. The audience either go right, get pick their phones up and start that, or they just they get really nervous for you, and so this, the whole thing, is on edge. So you know from the very beginning if you can relax the audience, and there are things that I've taught myself how to do. I've. I've got it all wrong in the past and I've learned from that.

Spencer Kelly: 1:07:41

And then there's the way you tell the story. Yeah, like you know, you've got a slide up there with 10 bullet points on and then you say something that's nothing to do with any of those 10 bullet points. What am I? What am I supposed to pay attention to the bullet points or what you're saying? So there's, there's ways of giving your presentation which is more compelling. There's a level of detail you don't have to go into in your presentation. So there's all this sort of stuff that you know I want to. I want to pass on to the people who have to stand up on stage.

Chris Norton: 1:08:11

If they want me, you know they stand up on stage yeah, because the best speakers don speakers don't have death by PowerPoint 15 bullets on a slide, but some people don't.

Spencer Kelly: 1:08:21

My ones. I think there's two words in my. I use PowerPoint but it's basically a slideshow videos and pictures and there might be two words in there. But again, you can use words, but just use one at a time.

Chris Norton: 1:08:34

Yeah, and you know, catch the audience's attention with how long is it taking to master that stage craft on? Because they obviously 25 years. Yeah, and do they teach you that at the bbc? Have you just?

Spencer Kelly: 1:08:46

learned that they gave me half a day of presenting tuition when I started, which was actually really useful. Um, and the rest of it, you either pick up or you don't. And you know, I don't know whether you listen to these things back. You won't listen to these things, this particular one back, until I would well take you into all those, won't it? But, um, if you listen to your stuff back and you're the right kind of person, you can critique yourself, but sometimes it requires someone else to say okay, I can see what you were trying to do there, how about this?

Spencer Kelly: 1:09:15

And you're bouncing off of people. So the stage cross stuff is, you know, you've got to live it and and again, going on far too long. But there's a podcast called the comedians comedian, which I listen to, which is about there's a comedian called stewart goldsmith who interviews other stand-up comics, and they're not trying to be funny, they're talking about the art of being on stage and I find out from listening to that I'm I'm doing the same sort of things, that they're not trying to be funny.

Spencer Kelly: 1:09:36

They're talking about the art of being on stage, and I find out from listening to that I'm doing the same sort of things that they're doing. You know how to win the room, how to win them back, you know how to recover from a problem, or if you know you've lost your train of thought, and so there's all these technical things that I've had to work out for myself.

Will Ockenden: 1:09:52

So I'm sure our listeners will be thinking I need to get in touch with, uh, spencer, I want some of that. So, um, yeah, how can people get in touch with you?

Spencer Kelly: 1:10:01

oh, spencerkellycom, it's got it all on there. I should have had more of a, more of a thing, more of a deck than that, shouldn't I? No, spencerkellycom. And there's contact details. I've got two agents. Because, of course, because who has the best one agent these days? One for broadcast and one for keynote speaking? Um, but, yes, it's, it's all on there, and two agents impressive, I know that sounds so math, there's. There's one for your light touch. Edit, take out the two agents.

Will Ockenden: 1:10:31

One's a bleaching agent and one's a cleaning agent and a question we'd like to ask all of our guests is and I'm sure you've got a million and one suggestions for this, given all the people you've met over the years but who? If we could have another guest on the show, who would you recommend we could talk to? Ideally someone that's made some cracking mistakes along the way is this how you get your guests?

Spencer Kelly: 1:10:52

no, no, you did not bother to do any research we do.

Chris Norton: 1:10:55

But sometimes when people like tap us up and say, oh, I've recommended you for this show, and then we have actually had a few people do that, it's quite a nice thing to do. But I mean, the problem is, ours is about telling this is something that you've done wrong, and not many people can think of people who've done something decent yeah, what about the founder of alibaba?

Spencer Kelly: 1:11:13

jack ma. If you can find jack mar, then have him on. How about mark zuckerberg? How about elon musk?

Chris Norton: 1:11:20

I don't know elon musk's a great one. If you're listening, elon, we are available. We might be able to slip you in in october sometime and on that bombshell.

Will Ockenden: 1:11:28

Yeah, so, spencer kelly, thank you very much for coming on the show. Yeah, thanks for coming on the show. That was great fun, thank you.