Hannah Clayton-Langton: 0:04

Hello world and welcome back to the Tech Podcast. I'm Hannah Clayton Langton.

Hugh Williams: 0:09

And I'm Hugh Williams, and we are the podcast that explains technology to curious minds.

Hannah Clayton-Langton: 0:14

Curious minds. Well, it's great to see you, Hugh.

Hugh Williams: 0:17

It's nice to see you. It's been a long time. I really think we haven't caught up since the end of season two, probably because we've both been busy.

Hannah Clayton-Langton: 0:23

Yeah, it's crazy where we've been and all the different we've been up to. More on that later, I guess. So, guys, we'll be back soon for season three, and we're busy working on some really awesome stuff for you guys. Before we get into the episode, it's been a little while we caught up, and I have one important question for you, and one tech reveal. And I didn't mean I haven't told you my reveal yet. reveal. So Hugh's for the listeners, a little reminder, I've a lot of time over in LA, and he's a big Waymo fan, has been from the beginning. One of our best performing episodes, actually, from is our episode on how Waymo works with uh Nick Pelley. But Hugh is on, he locked up something like 27 Waymo rides. He hadn't been in LA that long, but so it was pretty given the tenure. Um, I want to know how many Waymo's you've now taken.

Hugh Williams: 1:14

Now I knew you'd asked me this question, Hannah. So I opened up the app. Uh I didn't know the answer off the top of my head. And the answer is 138 rides, 613 kilometers, 1,900

Hannah Clayton-Langton: 1:25

So 138?

Hugh Williams: 1:26

138.

Hannah Clayton-Langton: 1:27

I actually thought it would be more.

Hugh Williams: 1:29

Well, as you know, I tore my hamstring and I had to have And this is this is a different podcast. But yeah, yeah, I wasn't very mobile. So I've got to say I've had probably the last six or seven of uh sitting on the sofa and not taking too many Waymo's. So that's uh otherwise I reckon they'd be closer to 200 by I love Waymo.

Hannah Clayton-Langton: 1:46

Well, I'm waiting for the London launch, they're teasing it. We've had a notification from Uber. There was like an opt-in for Waves because Uber are working with Wave. So I've opted in, but I have not seen any out on the let alone ridden in ones. So I I don't know when they're coming. I know that's imminent for London, but it's felt like it's that way for a while.

Hugh Williams: 2:04

Yeah. And when it does come, um, it'll be life-changing.

Hannah Clayton-Langton: 2:07

And then my big tech reveal for you is that about four ago now, I actually deactivated my Instagram. I don't know if you noticed, I disappeared from

Hugh Williams: 2:18

I did not notice. Wow. What are you doing with a little spare time?

Hannah Clayton-Langton: 2:22

I know. Well, I'm actually reading, I'm reading so many more And I also um have gotten really into the New York Times. You remember Wordle, which some guy made during COVID and the New York Times bought it. So I'm a paying subscriber to the New York Times I'm big into the mini crossword. There's this one called Spelling Bee, where you have to like make up different words with this set of letters. Anyway, I've got like leaderboards with my friends.

Hugh Williams: 2:46

It's so awesome. So awesome. The power social media has over us. I guess that's uh something we talked about during But uh, we're here today to just recap season two, build a to season three. Season three is coming soon. And you know, we're already busy working on some really episodes for everybody out there. Uh, we've got some great interviews coming up. We've got a new segment, which we'll tell you a little bit in our promo. But in the meantime, we had such a great time recording two that Hannah and I have put together this best of with some of our favorite insights and a few little anecdotes to tide you over.

Hannah Clayton-Langton: 3:20

And now you're spending so much time back in Silicon Valley, Hugh. I'm buzzing to get some more intel on what it's like on the inside. So we should definitely revisit that. Actually, back in episode five of season two, Hugh has already spilled some secrets on what it was like to work for Google, which will give you guys a flavor of life in big tech. And I will certainly be prying Hugh for an update in the season as well. But uh, have a listen to how he described life at Google back in episode five.

Hugh Williams: 3:46

Yeah, and look at all of it's true. Certainly when I worked at Google, I worked at the Mountain View campus, so the main campus of Google. I think it probably had around 50 different restaurants. Wow. Um, not all of them open for all meals, but you know, they breakfast, lunch, and dinner. Pretty much select any cuisine you want any time of the So I used to hit the gym. They they probably had, I don't know, more than 25 gyms. They'd sort of done a faux athletics track with the grass and things. They had big tractor tires there that you could flip that you could hit things with, put out sort of the cones and have running drills, you know, a bit sort of boot camp kind of style. I used to enjoy that. And when I finished the class, go and have a shower and uh the free towels, of course, uh that come with the uh the and then I'd head over to one of the smoothie stations and uh get them to make me up a really, really good protein smoothie with all the additives that that I wanted. Usually for lunch, you know, I'd pick one of the healthier quite like one of the Japanese restaurants they had, used go there. And then I didn't usually stick around for dinner, I'd jump on my bike and pedal home. But certainly, you know, the culture is designed to get you in there early so you get a free breakfast, feed you at and then get you to stay late, have dinner, stick around, do a little bit more work. So, you know, the culture is designed to get you to work hard by providing you with the uh the perks to help you do that. two wouldn't have been season two without AI, and this is still a super relevant topic. In fact, Hannah, I've been thinking in the back of my mind we should rename Tech Overflow to be AI Overflow. What do you think?

Hannah Clayton-Langton: 5:13

I love it. Do you know what? If we went back in time to when we came up with this things were so different. Like maybe it AI is a new tech in terms of educating the

Hugh Williams: 5:23

Yeah, I'm not sure we would have done coding as our first I mean, that just seems seems almost quaint that that one that we've talked about. It's like churning butter was relevant. Exactly. But uh but anyway, early on in this season we talked about people are actually using LLMs. I really enjoyed that episode. And here in episode four from this season, we talk about some insights from the white paper that OpenAI released about how people were actually using ChatGPT. There was a paper that came out of OpenAI, I think around last year, sort of a white paper. So not a peer-reviewed piece of research, but a white paper that came out. And uh, first thing is that almost 80% of conversations into three buckets practical guidance, seeking and writing. And so I think you and I, lots of practical guidance, lots of seeking information, but we probably didn't have in our top three writing, you know, so whether it's refining things, emails, you know, improving documents. And they also interestingly proposed a new about how these tools are used. And it's the asking, doing, expressing framework that up with. And what they figured out was that 49% were asking a question and getting back an answer. 40% of the prompts were getting it to do something, and then 11%, and this is super interesting, I know you love this. Uh, 11% is people just expressing something. So that's neither asking a question nor asking it to do it's just sharing a feeling with ChatGPT.

Hannah Clayton-Langton: 6:48

Okay, so to put that into real terms, let's take a cooking use case. Asking is like a simple question. So it might be like, is there cheese in pesto?

Hugh Williams: 6:58

Yep.

Hannah Clayton-Langton: 6:59

Right.

Hugh Williams: 6:59

100%.

Hannah Clayton-Langton: 7:00

And then doing would be like create me a dairy-free pesto

Hugh Williams: 7:06

Yeah.

Hannah Clayton-Langton: 7:06

And then expressing would be I love the pesto. Yeah, and that's it. I mean, I don't I think the expressing is the thing I'm the least. It's kind of hard for me to get my head around. But you said 11% of interactions are that expressing

Hugh Williams: 7:19

Yeah, exactly. People just saying something to their LLM that's a question or asking you to do something. So, you know, hey, I love the pesto. I love my running shoes.

Hannah Clayton-Langton: 7:28

Yeah. I guess maybe maybe pesto and running shoes aren't the best for expression, whereas expression might be like, I'm really stressed and overwhelmed. And sometimes people just say that to their LLM in lieu of it to anyone else.

Hugh Williams: 7:41

And I think this is pretty unique. Like I think this is not something that we're all doing with Google. It's not something we're used to doing with our computer. So I think this is kind of new.

Hannah Clayton-Langton: 7:52

Yeah, I think it'll be super interesting here, actually, to revisit how we are using AI between us as well, because my has massively changed since we started the podcast and since we finished season two at work and in my personal life. Like I'm massively rammed up how I'm using it. So we should definitely uh have a chat about that and think about where it's all going.

Hugh Williams: 8:10

Yeah, and we could also talk about how it's changed software engineering. I mean, I think software engineering is just changing every week. And, you know, the way that I see software engineering done today is just so different to a year ago or two years So, you know, definitely a topic we should come back and Hannah.

Hannah Clayton-Langton: 8:24

Well, AI overflow is coming soon. Speaking of AI, one of our highlights this season was guest Ramez Nam on the podcast. Mez is an AI futurist and science fiction author and just duper smart guy. And he came on the show and was exploring with us what AI and can't do today. I'm sure that's changed actually since the episode came out. Uh, its impact on jobs and coding, which we've already and the AGI hype, which I learned in that episode, stands for artificial generalized intelligence. There's so many great moments from this episode. We'd honestly just recommend you go back and give it listen. That's uh season two, episode five, because every time I to it, I learn something new. Um, but one of my biggest takeaways from the episode with Mez was that LLMs are basically great at memorizing like of data more than any human could in a lifetime. And there's a big difference between that and a GI, theoretically take it to the next level and solve problems like a human can.

Hugh Williams: 9:18

Yeah, and we actually asked Mez on the episode, Hannah, there's any real breakthroughs coming that might take that level. Mez believes that some of the biggest breakthroughs really are yet to come. Let's join him from episode five as he explores that

Ramez Naam: 9:33

I think some of what's happening is a mismatch between what measuring in the hope of achieving superintelligence or you know, really high quality in helping us on mathematics discovering new physics or curing cancer versus what everyday people do. So I think for most of us in most chat sessions, it appears to just be you know almost saturated. It does get better, but for simple questions, it kind of already hit the level that is mostly satisfying, except for like you know lowering that error percentage like bit by bit by But it's just in the last six months, I would say that the top mathematicians have started to say, this can actually me do my work and it's helped me like prove something that working on for a long time. And so that's sort of a new frontier. And how much is that worth? You know, it's hard to say, like in in strict math, it's worth a lot to society. But if that's like, oh, I came up with a new idea for a you a pathway for a cancer drug that's worth a tremendous amount. A lot of the work that the labs do in the hardcore training and in trying to devise better algorithms and so on to address some of those issues is really not aimed at like satisfying your, what's the best Moroccan stew recipe I could make with the in the picture I just took, which is amazing that it can do that at all. It's not aimed at that, right? It's aimed at we're gonna cure cancer and we're gonna like this physics problem, and we're gonna see if we can predict qualities of new materials that we might make before we even them in the lab and that sort of thing.

Hannah Clayton-Langton: 11:12

And taking us away from AI for a moment, more towards our with our phones, which is one of my favorite topics to I still can barely go a day without someone making a joke how their phone is listening to their conversations. And we talked about this in season two, episode one, where Hugh sort of blew my mind and explained why Alexa or phones or Google Home devices actually don't even need to listening to us because of all the sort of data and intel we're feeding them in other ways. So have a listen to this clip because it really changed my on how I'm using tech.

Hugh Williams: 11:44

You're providing lots and lots of context whether through your use of Instagram, through your use of Facebook, through your use of WhatsApp. So that's a that's a big signal, and you can figure out a lot about people's intent by their location, right? So I don't have to say out loud, or maybe I've said out in my house, oh, I'm I really want to buy a new surfboard, then suddenly I'm getting surfboard ads. It could also be because I searched for surfboards or I went to a surf shop. It's not that Siri or Alexa or Google Home is listening.

Hannah Clayton-Langton: 12:11

Yeah, we're we're putting out signals, right? That's the big thing. And we're putting out signals in what I'm texting you or you, what I'm searching for. And then they're like, it's like how I'm scrolling through right?

Hugh Williams: 12:25

Yeah, just if you think about any feed-based tool, right? There's lots of clues you're giving it by interacting with feed, right? The most obvious one is if you tap on something, right? So if you tap on something and expand the picture, know, showing pretty strong intent that that's something interest to you. If you scroll through the images, even more intent, right? Like, hey, I'm super engaged with this. So you're giving that signal back. If you skip something really fast, it's a great signal, I'm not interested in that. If I pause on something, watch, you know, more than a amount of a video, tells it something.

Hannah Clayton-Langton: 12:56

If you like it.

Hugh Williams: 12:56

If you like it, obviously massive signal. Yep, yeah. If you share it with somebody else, huge thing. Oh, I'm always sharing.

Hannah Clayton-Langton: 13:03

Add you sharing videos.

Hugh Williams: 13:04

Message it to somebody else, you know, all of these kinds things. So you can imagine working on that team and building a that understands people's intent based on positive things they do and also negative. While we're on our relationship with phones, Hannah, in three this season, we really talked about why it's to put down your phone and how big tech has smart people are really driving up metrics. And you know, in the case of social media, that's about you to engage more and use more minutes. And the essence of that really is experimentation. I'll tell you a quick story. The the most important, significant A-B test I've ever been involved in was slightly increasing the size of the images in the search results page at eBay. Instead of just showing users, you know, small thumbnails, showed them medium-sized thumbnails. That experiment alone put more than 400 million US dollars of revenue in the market.

Hannah Clayton-Langton: 13:59

That's crazy. Right.

Hugh Williams: 14:01

And it's such a small adjustment. It wasn't even technically very hard. Yeah. Right? We're just showing slightly larger images. And I think when you've seen that up close, when you've just how powerful it is to be able to measure changes on the user base and understand that change in engagement, then you want to do more of it. And so I'd say a lot of companies are resistant to because it involves humans letting go of control and machines have control and also being more about testing any idea not being right. Companies get quite resistant to that. But once you've seen it up close, uh, you just want to do of it.

Hannah Clayton-Langton: 14:35

Okay. And I imagine it's very rare that you actually like lose 10% of your users overnight because you've changed the size of a font and people have just deleted the app, right? Because if you're a company that's worried about I presume that's the worry, right? That you're gonna lapse a bunch of users because you made the wrong decision.

Hugh Williams: 14:52

Yeah, absolutely. And I think you have to have a budget. So you have to have a think about kind of if we're gonna our users in different treatments, how much are we willing risk here to do that? And you also want to be monitoring these metrics all of the time, right? And making sure that if treatment D is not a great and we're statistically sure that that is a poor treatment, then we just turn off D as soon as we possibly can, right?

Hannah Clayton-Langton: 15:12

Okay, and when you say budget, that's not budget for like to make changes to the platform. That's budget because if your conversion drops by not point two percent because of the way you formatted the button, you want to have planned for that eventuality rather than lose money.

Hugh Williams: 15:30

Yeah, absolutely.

Hannah Clayton-Langton: 15:31

And I assume this is kind of like we talked about in previous episodes about how reported behavior is very different be very different to actual behavior. Like we think or maybe want to behave in a certain way, and then our habits shine through with the reality. And I guess that's why emotionally led decisions around UX be pretty flawed because you think you want a button that's this color in this corner of the page, but you've not actually tested that that drives the results that you're after.

Hugh Williams: 15:58

Yeah, and I've had lots of arguments with UX people where, know, they've said, hey, it doesn't fit our sort of, you corporate ethos, the style guide, whatever else it is, don't want to do that. And I always say, like, let's just run the test. And then when we've got some data, we can have the about who's right, who's wrong, whether the style guide needs changing, whether you're right, that we shouldn't be afraid of gathering data and and learning. What we should be afraid of is thinking we're right.

Hannah Clayton-Langton: 16:24

Do you know what? That episode singularly has massively changed my behavior my phone because I used to think it was all about willpower, and now I've realized that like it's me against the product minds that work in Silicon Valley and Formta. So I tend to leave my phone. I think you recommended this in the episode. I tend to just leave my phone out of sight a lot more, and I just don't reach for it. Sort of one of the reasons that I've deactivated my account as well. I just realized that like I needed a circuit breaker, I need a better way of managing it because um there's some super people who are developing this stuff, and I don't think I can them. And since we're on the topic of experimenting, another guest we had this season was Derek Connell. So if you remember, Derek spent over 25 years working at and talked us through his most spectacular failure in his That is to say, an experiment that didn't quite pan So Derek led the team that launched the infamous chatbot which actually is one of the early cousins of today's LLMs, which makes the whole story even cooler. And what happened ended up laying the foundation for a lot learnings that the industry took as a whole when it ethics and AI. Um, but the story of Tay is ultimately about risk and It wasn't all bad. And here in this clip, Derek talks us through what it was to face his CEO on the day that everything went wrong.

Derrick Connell: 17:43

You've taken a risk. Sometimes they hit, most of the time they flop, sometimes fail publicly in a big way. So, you know, then you have to deal with that. Explain to the rest of the company what went wrong and go the CEO and explain what happened because you know, big PR I guess you could call it.

Hannah Clayton-Langton: 18:00

Were you the sort of the point person?

Derrick Connell: 18:01

Yeah, we came, there's a few of us. 7 a.m. go see the CEO.

Hannah Clayton-Langton: 18:04

7 a.m.

Derrick Connell: 18:05

Yeah, you know, first thing. You know what's a problem when you're the first thing in to see the headmaster.

Hannah Clayton-Langton: 18:09

Yeah, yeah.

Derrick Connell: 18:09

But you know, uh one of the things about innovation is, and this is a leadership opportunity for anyone out there who running a team doing incubations, things can go wrong. And the question as the leader of that organization or CEO a big company like Microsoft is what do you do? Because you can react in a number of ways. And I was so happy with the reaction of our CEO because we in, it's obviously a big problem. And the first question he asked was, What did we learn?

Hannah Clayton-Langton: 18:40

Not even no swearing.

Derrick Connell: 18:43

No swearing. What did we learn? And that is someone who knows how to manage innovation, how to motivate a team. You know, obviously hold us responsible because something's gone wrong. But first question, what did we learn?

Hannah Clayton-Langton: 18:56

And he's he but he's also, I guess, thinking back to he's expecting you to have that answer.

Derrick Connell: 19:02

Of course.

Hannah Clayton-Langton: 19:02

It would almost be worse if it had happened and you didn't the answer than if it hadn't happened at all.

Derrick Connell: 19:07

Yes. Now, was he happy with what we learned? He was happy with some parts of it and he was unhappy with parts of it, you know, because we made a mistake, rookie Repeat after me. Come on. Rookie mistake. But he was very happy that we tried. He was very happy with the learnings about conversational influenced a lot of the decisions that Microsoft have made the subsequent years.

Hugh Williams: 19:28

And it's not just our guests who make mistakes, Hannah. I've made plenty in my career, and I had a huge screw up I first worked at eBay. And look, it didn't really make the headlines quite as as Tay did, but it's similar to Derek's experience in tough day at work resulted in a super useful learning search team. So here I am talking about that in episode seven of this When I was working at eBay, different search problem, released the iPad. Okay. Super cool, right? So one day there's no iPad, next day, suddenly this new exists. And boy did we screw that up. If you went to eBay and you typed in iPad, we would say, some results for the query iPod.

Hannah Clayton-Langton: 20:07

Oh no.

Hugh Williams: 20:08

Because we were we were really, really sure that what you was iPod, right? So there's probably enough users out there in history accidentally put an A in instead of an O. And so there was no way we were going to show you iPad because that wasn't a thing. Wasn't a thing. That was a typo. Yep, we were really sure that you meant iPod. And so, you know, I got a call from uh the boss's boss and of said, Hey, there's all these customers who want to buy an iPad, and we're showing them iPods. Like, what are you guys doing? Please fix. Yeah. And so we learned, you know, the really hard way that we to have some overrides in the system. You couldn't just rely on history to do the query and the correction. You have to have some overrides where you kind of you can deploy the idea that this word should not be corrected. And then all of a sudden, we started to show iPads to who wanted to buy an iPad.

Hannah Clayton-Langton: 20:55

Who's buying an iPad from eBay the day it comes out? Someone who missed it at the store.

Hugh Williams: 20:59

And then so, you know, people are, you know, it's like reselling, I suppose. People would go and buy a bunch of these things and then list them on eBay at a premium, and people were happy to because they were such a big deal when they came out.

Hannah Clayton-Langton: 21:09

So, how quickly were you able to fix the did you mean iPod?

Hugh Williams: 21:13

I think it was a long night for the team because we'd needed this feature before. Uh, this had never happened to us. And so it was a long night, probably in the order of six eight hours before we had a fix out. And then uh and then we had to build a feature that allowed us to do this anytime. Well, we could do it in seconds in the in the future when a new word appeared. But again, you know, this is part of the query understanding team's job is to make sure that the right queries are the queries are annotated in the right way, these kinds of

Hannah Clayton-Langton: 21:39

And our last guest of the season was John Lilly, who joined us for episode 10. And John talked about how investing, like most things in life, is all about fit and relationships. I so enjoyed talking to him. John's got some really great stories about investing in Instagram. I think he invested in Duolingo as well. He's a super interesting guy. But my favorite quote of that episode was when John said, are the sun and investors are the moon. Investors are only reflected light relevant to which is pretty cool metaphor. So in that episode, we of course got John's take on AI and how much of a disruptive force he thinks it could be for the industry.

John Lilly: 22:17

Yeah, AI is changing everything. I um I would say that like more has changed over the last like 200 days than I think by any other time in my I'm an old guy now, I'm 55, and so like I've seen a fair of stuff. Like I just can't remember any time it's changed like I would say that two or three years from now ago, it was that LLMs in particular, uh large language models in that OpenAI and Anthropographer built on were useful. It was not clear to me that they were going to be uh, certainly not like they have been. I mean, it was even maybe just a year or two ago, people were debating whether these things could take over search like Google. And it's uh 100% obvious now that they can.

Hannah Clayton-Langton: 22:59

Yeah, well, we talked about this on a recent episode. Like, if you'd even said that, as you say, even six months for me, I would have been like, nah. And then now I don't use Google, I go to Chat GPT for

John Lilly: 23:09

Yeah, it's really shocking to me. And and then really, it's the last 200 days since Cloud launched Opus um 406. That was the first thing that the one that could code. And now, you know, I've coded more in the last two months I have in 30 years. Uh I'm building products, and so it's like for me as a product person, it's given so empowering to be able to say, Oh, I this thing should exist. Let me go play with it for a little while. And it's a shot, it's a complete reinvention of just about in software. And so I teach a class at the business school of Stanford, and we had a class yesterday, and we're like, this is the best ever to be an MBA student because like before it would be, I have this idea, and I got to go across the street to and find some computer scientists to try to prototype it me. And now, if you can't prototype, like you shouldn't be a company because like it's it's just right there for you. If you can describe it using Figma make or Claude, you can see it, you can build it in real life and see what gonna feel like. And then you can go talk to customers and you can get real in the loops, and so that changes just about everything, and it's just shocking to me how fast it's turned. So we're seeing lots and lots of startups. I'm in a new place now called Gigascale Capital, and we're doing sort of energy and hard tech type things. It's a little harder to vibe code like uh a nuclear power or a like electrical transformer.

Hannah Clayton-Langton: 24:35

For now, anyway.

John Lilly: 24:36

For now, yeah. But even those industries, I think, will be are by what you can do, especially with coding.

Hugh Williams: 24:45

And last but certainly not least, Hannah, was my favorite from season two, which was episode two. I remember us creating that episode. I think it was the middle of the week. We had the whole plan for the season, and then somehow we up in this discussion that it'd be fun to do some vibe and maybe use Claude Cowork, which was relatively new at time, a sort of an agentic framework to try and solve a And we had a great time doing that. We actually built a little calculator, and uh I think we a pretty good holiday for you for Florence.

Hannah Clayton-Langton: 25:13

I remember I was like actively having my mind blown as we doing that vibe coding, which is kind of crazy now because so much more commonplace. Like, even as you say, cowork was relatively new at the and now it's something that like using every day. Um, so I would definitely recommend going back and taking a listen to that episode because it was super fun. And then, of course, at the end of the episode, we got a bit into the more existential stuff around AI and the of work, which is such a huge and fast-moving topic. We'll definitely be getting into it more in season But have a listen here to what we were thinking about it back in episode two.

Hugh Williams: 25:47

Have you heard of the um the Stanford HIA AI Index Report comes out of Stanford University every year?

Hannah Clayton-Langton: 25:55

I have heard of it from you. We were chatting about this when I saw you, so yes, I have, but I can't lay claim to have found that one myself.

Hugh Williams: 26:03

So they write an annual report and it really looks at the of AI and how it's affecting the world and how people in society, investment, all these kinds of things. It really makes clear a whole bunch of points. And one that it really makes clear in the latest report is jobs aren't really being displaced yet by AI. It's just that tasks are being made faster. And I think we've seen that in today's show, right? Like you've seen vibe coding speed up a task and you've the agent solve your Florence holiday much more quickly. And so, you know, a marketing manager who might have of their time drafting the first copy that was for some piece of work is probably now spending 5% of their time on So they didn't lose their job, their job just really, you changed in shape.

Hannah Clayton-Langton: 26:45

The biggest game changer here that feels tangible is what we mentioned earlier on that you like, you no longer need to go and find the the tech person if you have a business idea. You can just sort of like spin something up. But other than that, it doesn't feel like there's a massive material difference in output that anyone could feel Like it's certainly fun to play with, but I haven't seen step-changing. As you say, there's a lot of smoke and mirrors at the minute.

Hugh Williams: 27:11

Like, but I but I do think this is coming, right? So I do think that routine cognitive basic work is at right? So people who have basic jobs around data entry, basic report generation, you know, first-pass document reviews, you really simple translation, dealing with standard queries. These jobs aren't gonna disappear overnight, but the number of humans that are gonna be needed to do these jobs is going to reduce measurably. So, you know, AI will march into these routine kinds of

Hannah Clayton-Langton: 27:42

I do feel strongly that everyone's like playing with this because they know it's huge, but I'm not sure we've fully how and when we're gonna have that Eureka moment in of it just yet. We're obviously getting closer with agents versus LLMs, but we're still sort of circling around it, it feels.

Hugh Williams: 28:00

Yeah, 100% agree. But look, you know, my honest take, Hannah, is that, you eventually AI is going to eliminate some jobs. It's definitely going to transform many more jobs. But the exciting news, I think, is it's gonna create many new jobs and they're gonna have titles that we haven't dreamed yet, right? So I think there's going to be a shift in what people do, but it just hasn't arrived yet. So that's a wrap of the best of season two, Hannah. It's uh being really, really hard to pick out some It was such a fun season to record with you. And if that's got your cogs whirring, definitely go back listen to your favourite episodes again. I'd recommend episode two. It's so much fun. And every time I re-listen to an episode, I actually awesome. Um, and I remember how much fun it is to be podcasting with you, Hannah.

Hannah Clayton-Langton: 28:43

A hundred percent. And likewise, if I were to go back and listen to another which I actually might do, it would be episode five because Maz just brought so many amazing insights to that episode. Um, anyone who's sort of AI curious, that's a really good to start. Tech Overflow season three is coming soon. You won't want to miss it. And we'll be getting into big questions like are we still AI optimists? And how many Way MOS will Hugh have taken at the point of season three?

Hugh Williams: 29:11

Let's see if I can get to 250. And for some reason, I feel like that's a milestone. Something like park run or something. Maybe I'll be.

Hannah Clayton-Langton: 29:16

I'm just gonna say it's like park run. Maybe we should try and record an episode from away. Oh, we should. Anyway, if you love season two, definitely subscribe to you get your podcasts. Maybe leave us a review, and you can always check us out on Instagram, LinkedIn, X. And we always love to hear from the listeners, especially the kind of topics you'd like to hear from season three.

Hugh Williams: 29:38

We'll see you all soon.

Hannah Clayton-Langton: 29:39

Bye.