Innovation: Coming up with new ideas

Summary

---

Podcast Summary: Economist Podcasts – "Innovation: Coming up with new ideas"

Date: May 12, 2025
Host: Andrew Palmer (The Economist)

---

Overview

This episode of the Economist’s "Boss Class" series explores the real ingredients of innovation—dispelling common myths and diving into the practices of three distinct organizations: Monumental (AI-driven construction robotics), Wave (AI for self-driving cars), and Google (global search giant). Through in-depth interviews, the show reveals that successful innovation is rarely the result of sudden "Eureka!" moments or trendy office aesthetics; instead, it's a messy, iterative process built on purpose, feedback, structure, and creative adaptability.

---

Key Discussion Points & Insights

1. Robot Bricklayers: Innovation on the Building Site (00:24 – 06:45)

• Field Report: Andrew Palmer visits a residential construction site near Rotterdam where Monumental’s autonomous robots collaborate with human masons to lay bricks.
  • The robots coordinate complex tasks—laying mortar, placing bricks, and even mimicking masons’ vibrations to improve bond strength.
  • Tech isn’t flawless; mistakes like dropped bricks still occur, reminding us of the trial-and-error involved in creating breakthrough systems.
• Salah Al Faji (Monumental Co-founder):
  • On "fake innovation":
    “We have a very strong aversion towards… fake innovation, which is just like colored post-it notes and bean bags. There's this kind of aesthetic of innovation which is not innovation.” (04:17)
• Innovation Drivers:
  • Monumental was founded to solve a real shortage of bricklayers, recognizing societal need as the spark for technical progress.
  • Real innovation means accepting messiness, learning from errors, and eschewing shallow “ideation culture.”

2. Wave: Iteration over Inspiration in Self-Driving AI (06:45 – 13:24)

• Key Message: Breakthroughs aren’t born from eureka moments but from years of sustained, iterative effort.
• Alex Kendall (Founder, Wave):
  • “The biggest bullshit is Eureka ideas where you just wake up and have an idea that solves things. Breakthroughs are achieved by sustained effort over many, many years.” (06:45)
  • Chose London’s challenging driving environment over easily-automated “desert cities”—testing the system where complexity is highest leads to scalable solutions.
• Adapting to Change:
  • Initially resisted but ultimately embraced language models in robotics, realizing their necessity for more human-like, interactive AI vehicles.
  • “Language gives you a great ability to introspect the model and understand how it's thinking… and build a sense of trust with it.” (11:53)
• Lessons:
  • A clear mission anchors innovation, but willingness to listen to team instincts and adapt course is crucial.
  • True “flexibility within purpose” accelerates real progress.

3. Lego: Structured Play Meets Industrial Innovation (13:24 – 22:40)

• Dan Meehan (Lego Designer):
  • Describes the "childlike wonder" central to product design—deep user research with kids shapes not just new toy concepts, but the broader innovation process.
  • “The very best part of my job is when we do kids tests because kids are so honest and they will break the model instantly.” (16:00)
  • Unexpected user behaviors (e.g., kids flying, not driving, a rover) redefine products’ purpose.
• Niels B. Kristiansen (CEO):
  • Balances creativity with rigorous supply chain constraint—“mastering total creativity on one side and then really very, very structured supply chain and huge factories on the other.” (21:50)
  • Encourages creativity via "Fabulab Fridays" (unstructured design time) and "Creative Boost" weeks, but all ideas must fit a highly organized, future-oriented roadmap.
  • “With 50% new every year, I don’t have a lot of inventory sitting. So we produce it when it's needed, when it gets on the wish list.” (18:24)
• Key Takeaway:
  • Innovation at scale depends on creative freedom and rigid discipline—a portfolio approach with strict boundaries ensures continuous, relevant breakthroughs.

4. Google Search and the AI Overhaul (23:44 – 32:03)

• Liz Reid (VP, Head of Search):
  • Discusses integrating generative AI (“AI Overviews”) into search—summarizing complex queries, offering both answers and links.
  • “People often, like, stop right there, organize the world's information. But there's like the second part about making it universally accessible and useful. And useful is a very different bar.” (23:57)
• Pragmatic Innovation:
  • Product evolution is iterative—user data and feedback drive constant refinement.
  • Not all bold ideas work the first time. Many at Google, including crowd-sourced reviews, became successes only when the time, data, and technology converged.
  • “...most of the things that I think of as being really successful at Google… were tried two or three times before and failed. Because the instinct was right, but the tech wasn't ready or the data wasn't ready or something wasn't ready.” (29:19)
  • Unexpected user questions (like, “How many rocks should I eat per day?”) push innovation in unforeseen directions.
  • On problem orientation: “Make sure you're falling in love with the problem, not the solution.” (31:01)
• Theme:
  • Innovation means obsessing over real user needs, iterating relentlessly, and mixing intuition with data—while bracing for surprises and setbacks.

---

Memorable Quotes

• Salah Al Faji (Monumental, on fake innovation):
  “There's this kind of aesthetic of innovation which is not innovation.” (04:17)
• Alex Kendall (Wave, on the myth of sudden genius):
  “The biggest bullshit is Eureka ideas where you just wake up and have an idea that solves things.” (06:45)
• Dan Meehan (Lego, on honest users):
  “Kids are so honest and they will break the model instantly.” (16:00)
• Niels B. Kristiansen (Lego CEO, on balancing innovation and discipline):
  “Maybe it's the secret sauce of what we actually can do.” (18:24)
• Liz Reid (Google, on user-centered problem solving):
  “Make sure you're falling in love with the problem, not the solution.” (31:01)

---

Timestamps for Important Segments

| Timestamp | Segment/Topic | |-----------|-----------------------------------------------------------------| | 00:24 | Monumental's bricklaying robots in action | | 04:17 | Salah Al Faji on “fake innovation” aesthetics | | 06:45 | Alex Kendall on the myth of the "Eureka!" moment | | 10:14 | Wave’s decision to test in complex UK cityscapes | | 13:24 | Lego’s design process: play and iteration | | 16:00 | The crucial role of kid testing at Lego | | 17:49 | Kristiansen on balancing creativity and manufacturing at scale | | 19:15 | Fabulab Fridays/Creative Boost: structured unstructured time | | 23:44 | Google’s AI Overviews and the evolving definition of “useful” | | 25:40 | Longer, more detailed search queries—the AI shift | | 29:19 | Repeated failures precede true success at Google | | 31:01 | “Fall in love with the problem, not the solution” | | 33:01 | Innovation themes and recap |

---

Key Takeaways

• Innovation is not about aesthetic or brainstorming culture but stems from real needs, persistent experimentation, and repeatedly learning from failures.
• Successful organizations blend clear purpose with adaptability, combining creative chaos (skunkworks, playtesting) with strict structures (roadmaps, supply chains).
• Breakthroughs are iterative, not sudden; the process is messy, user-driven, and demands constant recalibration.
• Feedback—especially from unexpected sources—is invaluable; unanticipated user behaviors and bizarre queries often open the next frontier of development.
• “Fall in love with the problem, not the solution” is a guiding mantra for innovators in every field.

---

For listeners and managers alike, this episode offers a candid, myth-busting tour of how top firms conjure new ideas—and how you can, too, by testing, iterating, listening, and truly caring about the problem you set out to solve.

Transcript

[Workday Narrator]

When you're a forward thinker, you don't just bring your A game, you bring your AI game. Workday is the AI platform that transforms the way you manage your people, money and agents so you can transform tomorrow. Workday, moving business forever forward.

[Alex Kendall]

The Economist.

[Andrew Palmer]

Tell us what we're looking at.

[Salah Al Faji]

You're basically seeing the robots building and our robot.

[Andrew Palmer]

I'm on a building site and wearing a hard hat. I've never felt more handy. I'm here because Salah Al Faji is showing me his vision of the future. Al Faji isn't a builder, he's not an architect. He's the co founder of a Dutch startup called Monumental. It makes AI driven bricklaying robots and they work together.

[Salah Al Faji]

They're like coordinating. Like one is laying mortar and then the other one is pushing down the brick in there and they grab the brick from another robot, which is the brick supply one, and that one can just drive off, grab more bricks, come back.

[Andrew Palmer]

Teams of four wheeled robots are working alongside human masons on a new residential property development about 25 minutes from Rotterdam. The house is probably two stories high.

[Salah Al Faji]

Three house, yes.

[Andrew Palmer]

See, I'm using words like story as though I'm like completely for ages. A supervisor from Monumental, who has a bit more building experience than me and also knows how to count, is on hand to supervise the robots and intervene if needed. But the supervisor doesn't control the robots. They act autonomously. So the arm with the mortar is now positioned above the brick and it's going to stop. One robot squeezes mortar onto a brick. Kind of looks vaguely disgusting. It's running along the length of the brick. Okay. Then the arm moves back. Another picks up the brick and places it onto the beginnings of a wall. The other arm, which is holding a brick, comes and places it on top. Placing a brick is not as simple as it sounds. Camera vision and a software representation of the wall ensures that it goes in precisely the right place. The robots have also copied a trick from human masons, a specific way of increasing the adhesive connection between brick and mortar.

[Salah Al Faji]

So what it's actually doing is it's actually like pushing and vibrating it into the mortar. There are sensors into the end effector because it's actually. You don't place a brick on the mortar, you actually need to push it down. And what mesas do, we discovered, is actually vibrate their hands a bit.

[Andrew Palmer]

Monumental was founded to solve a real world problem, a shortage of bricklayers. There's a long way to go before its robots can fill that gap. They can't build interior walls, they can't do soldier coursing, which as everyone knows, is a row of bricks laid standing up with the narrow edge facing out. They hit problems, things go wrong. Like bricks being dropped mid task.

[Salah Al Faji]

That dropping worries me, actually.

[Andrew Palmer]

You've seen that before.

[Salah Al Faji]

I've not seen this before. Not like this.

[Liz Reid]

It's just like people.

[Salah Al Faji]

Journalists come in, everything goes wrong. Disaster struck.

[Andrew Palmer]

Bloody journalists. Despite the mishaps, Dutch families will soon live in houses part built by machines. Even seeing a few bricks being laid in a kind of robotic choreography is hypnotic. This must make you feel.

[Liz Reid]

Yeah.

[Andrew Palmer]

This makes you feel. Yeah, yeah, of course. It's amazing. That is amazing.

[Salah Al Faji]

And like when you have a streak of this for like an hour or two, it's just like magical.

[Niels B. Kristiansen]

Yeah.

[Andrew Palmer]

If you were asked to conjure up an image of innovation, what do you picture? Hammocks, ping pong tables and whiteboards. People saying there's no such thing as a bad idea. Despite the existence of Hawaiian pizza and golf, Salah Al Faji, whose robots really are at the cutting edge, has little patience with all that nonsense.

[Salah Al Faji]

Just like culturally, we have a very strong aversion towards like, I just call it like fake innovation, which is just like colored post it notes beanbags. There's this kind of aesthetic of innovation which is not innovation.

[Andrew Palmer]

I'm Andrew Palmer, the management columnist for the Economist. The performances and practicalities of the workplace are what I write about each week. In the first season of this podcast on how to be a good manager. I learned how to hire people, how to run teams and how to motivate employees. I also discovered how important great bosses are and how rare. Managing well is extremely difficult, but it is a skill that can be learned. So I'm back with a new set of questions for the bosses of today and tomorrow. You'll hear tips from Levi's, Novo Nordisk Supercell and more of the world's best performing companies on how to negotiate, how to present, how to take decisions and more. In this episode, we'll look at the holy grail of the innovation. Every manager has been asked to brainstorm product ideas. Job candidates boast about their ability to ideate. The one thing you know about AI is that it's going to mean doing new things in new ways. So to find out what proper innovation looks like, the reality of it, not the aesthetics, we'll meet three very different firms. Wave is a deep tech company trying to make self driving cars a mass market reality. Lego is an established industry leader and an innovator. At scale. Google is threatened by disruption. It faces the challenge of overhauling its flagship product to make the most of AI. Will learn that innovation does not involve throwing a word association ball to your colleagues and wishing you were dead. It's more like the building sites where Monumental is training its robots. People know the overall goal, but the process is often messy, the pathways frequently unclear. From the Economist, this is Boss Class Season 2 on this episode. The ingredients of innovation.

[Alex Kendall]

I think the biggest bullshit is Eureka. Ideas where you just wake up and have an idea that solves things. Breakthroughs are achieved by sustained effort over many, many years.

[Andrew Palmer]

Alex Kendall is the founder of Wave. It's a startup which makes AI software for self driving cars.

[Alex Kendall]

And it's a single AI model that can learn to drive these different vehicles in different environments, safely around all kinds of scenarios, from urban to highway driving, to sun to rain, and even some examples in the snow.

[Andrew Palmer]

Like Monumental, Wave is an embodied AI company. But it's much further along in its journey. In 2024, Wave raised over $1 billion in funding, the largest ever investment in a European AI firm. Its tech is now being adopted by some very big names. The Japanese carmaker Nissan has announced that by 2027 its system will use WAVES self driving software. For a long time, however, Waves approach made it an outlier. Other self driving firms were writing rules for what a car should do in specific when the middle lane is empty, hog it, that kind of thing.

[Alex Kendall]

The largely Chinese and Silicon Valley esque efforts were focusing on creating the infrastructure and hand engineering the rules to be able to make these systems behave. And that produced incredible pilots and demonstrations. But I just was convinced that that wouldn't be the way that this technology would be able to scale globally. It's not a classical robotics problem, it's not an infrastructure problem, but it's an AI problem.

[Andrew Palmer]

If you're wondering why it's an AI problem, here's the explanation I think about.

[Alex Kendall]

Every time we go driving, the scenarios are different, the weather's different. I was driving yesterday and a lollipop person came out on the pedestrian crossing.

[Andrew Palmer]

For those across the Atlantic, a lollipop person is a crossing guard. In Britain they're required by law to carry an enormous lollipop which they lick whenever someone crosses the road in front.

[Alex Kendall]

Of them and was guiding the traffic using hand gestures and things like this, or just scenarios that are very, very hard to tell the car what factors to look for. But of embodied AI is that it can learn patterns in the data that are more complex than we can Hand engineer. The beauty is that it can scale very efficiently.

[Andrew Palmer]

The idea of a self driving system capable of making its own decisions has become much more orthodox now. But when Kendall first formed this conviction as a PhD student at Cambridge University eight years ago, it was audacious.

[Alex Kendall]

Finishing my PhD, I went and spoke with many automotive companies, big tech, self driving companies, and I was given feedback that this would never work. It's unsafe. Why don't you just use it for perception technology? I certainly had the conviction there that no, this would be how things work. This is where I want to devote my time. And look, it's certainly not been a linear or exponential growth towards this point. There's been a lot of ups and downs and a lot of points where any rational person would have looked at this and said, this is a fool's errand.

[Andrew Palmer]

Working on a breakthrough technology like this requires a mixture of optimism, bloody mindedness and flexibility. Kendall likes to talk about tackling the hardest problems first to minimise the risk of hitting a roadblock later. That explains one of the biggest early decisions he made, to move Wave from Cambridge to London.

[Alex Kendall]

We didn't go to a desert city where the sun always shines. You've got clean weather, you've got wide boulevard roads, but we went to this medieval city of transport network with lots of merging and I mean, we've quantified this. It's got seven times more cyclists and jaywalkers than a city like San Francisco. Of course the weather is varied and it's one of the hardest places to drive in the world.

[Andrew Palmer]

Good to know. Britain still leads the world in some things.

[Alex Kendall]

Starting there forced us to build scalable technology and also doing so with constraints like just a single GPU on the car and cameras. Not relying on all of this infrastructure and that forcing function forced us to build technology that could innovate and ultimately be more efficient, perhaps by orders of magnitude compared to our competition.

[Andrew Palmer]

If Wave's vision has been consistent, its path has been adaptable. Technologies change, thinking develops. Three years ago, a few of Kendall's team told him they thought it would be worth investing in language models for robotics to improve the car's efficiency.

[Alex Kendall]

This was before the world had woken up to the power of language models. And I thought it was absurd. I was at the time looking to keep the company focused and I was like, no, no, no, no, let's stay on task here.

[Andrew Palmer]

The team persisted and they convinced me.

[Alex Kendall]

That language would be a key part of the future of robotics for a number of reasons. Like, if you think about the promise of embodied AI. It goes beyond a robo taxi service where you're sitting in the car and it's like you're on invisible railway tracks and you just get in and go. But actually we can build a system that's a chauffeur experience that can interact with you, ask you how you're doing, what kind of driving you'd like.

[Andrew Palmer]

Schoolboy racer, please. With a donut at the end.

[Alex Kendall]

Language gives you a great ability to introspect the model and understand how it's thinking. You can ask it for feedback or ask, why is it doing a certain decision and build a sense of trust with it, or delegate a task that maybe it wasn't originally going to do and ask it to behave in a new way.

[Andrew Palmer]

Kendall trusted that his team saw something he couldn't and Wave incorporated language into its models. The idea that a single system will be able to cope with whatever is thrown at it was given a boost last year when the company's tech was tested on American roads for the first time.

[Alex Kendall]

It was incredible. On the very first day we were driving in the us, I was there, I jumped in the car, going to San Francisco and we saw the car learn these new behaviors. So in the uk, of course we don't have four way stop signs, you can't turn right at a red light. We drive on the left hand side of the road. In the us, it learnt all those behaviors so quickly and it was able to demonstrate, demonstrate those on the very first day of testing. I mean, that was remarkable for me.

[Andrew Palmer]

Wave has taught me a couple of things about innovation. One is the importance of a clear mission. Alex Kendall believed in his idea and stuck to it. The other is that conviction shouldn't lead to rigidity. Innovation is not linear, it zigs and zags. This blend of fixed points and flexibility leads us on to our next company. And like Monumental, it's also in the business of bricks.

[Dan Meehan]

So there's this lever on the back that you just fire the car. It's so fun.

[Andrew Palmer]

You just smashed this thing and this Formula one car came across and hit our mic. That's, that's good. One of Dan Meehan's latest creations has just collided with our recording equipment. All in the name of play. Meehan is a designer for Lego and responsible for its Formula one and Space Rangers. He is exceptionally, unashamedly enthusiastic about Lego, whatever form it comes in.

[Dan Meehan]

We made a handful of bricks. It was 12 or 14 of these bricks out of a 4.5 billion year old meteorite. For me, it was just A wonderful project because I felt like a child again.

[Andrew Palmer]

Feeling like a child again is pretty much the mission statement at lego. I've come to the toymakers headquarters in Billund in Denmark to learn about how they create must have toys for kids and big kids. When I met him, Mian was already working on products slated for launch in a couple of years time. I asked to see them, but wasn't allowed to or even to visit the floor where the designers worked. New products are kept tightly under wraps. LEGO no go. The products he did show me had launched or were just about to. Meehan talked me through the design process for the space brief. It started with research into what drives kids passion for space. To find out, he interviewed children across five countries including America, China and Germany.

[Dan Meehan]

And it was actually really interesting to find out that a lot of our preconceived notions of yeah, kids want X actually they were thinking very differently about space, particularly from their parents. So they were thinking much more about it being about beauty and wonder and imagination and less about the traditional maybe Apollo era generation of striving to be an astronaut.

[Andrew Palmer]

Meehan tries to understand what makes kids tick. He speaks to their families too. They're often the ones paying, after all. Once he's done his research, it's back to the leadership team with some proposals.

[Dan Meehan]

And then they say yes or no, depending on how well we deliver the information and the story.

[Andrew Palmer]

And then the designers start designing, sketching and visiting LEGO's vault of old products for inspiration.

[Dan Meehan]

So it's just something to get an idea out your head thinking I think it needs to be about this big or I think it needs to be this color. And then there's a process of refinements and taking that through peer review.

[Andrew Palmer]

The real test of a new toy comes when it gets into the hands of users.

[Dan Meehan]

The very best part of my job is when we do kids tests because kids are so honest and they will break the model instantly.

[Andrew Palmer]

Kids will also use models differently than the designers expected.

[Dan Meehan]

There was a six wheeled space rover and for the whole of that playtest, that vehicle with six big wheels and they had suspension and it didn't drive for the whole play test it flew. So the young boy that picked it up, he just flew this thing around the room collecting little aliens and putting them in the back of the space truck. We asked about the six wheel vehicle because there was a lot of time going into the mechanism because it had full suspension and it had six big wheels, you know, which could we take those elements and you know, change them for Something. And we said, do we need to add, like, the boosters? Do we need to put bigger wings on it? He's like, no, it's cool just as it is. And it flies. I was like, okay, no changes to the six wheel robot, right? And that's why I love it, because we spent so long making sure it was a really cool vehicle that we tested, driving over the desk and playing on the floor, and a lot of sort of, you know, action play that we did with it to make sure it stood up to the rigors of Play. We never flew it.

[Andrew Palmer]

Talking to Meehan is a bit like meeting the Tom Hanks character in the film Big, where a youngster inside an adult's body creates hit toy after hit toy because he can see things from a child's perspective.

[Niels B. Kristiansen]

We have a very structured way of being very innovative.

[Andrew Palmer]

And here's dad. Nils Kristiansen became the CEO of Lego in 2017. He knows the value of innovation. Half of the company's revenue each year comes from new products.

[Niels B. Kristiansen]

If you talk about having your head on the block as an innovative company, there's probably few companies that are as dependent on being creative and innovative all the time as we are.

[Andrew Palmer]

Why is that? Is it the nature of kids as consumers?

[Dan Meehan]

Yes.

[Niels B. Kristiansen]

Yeah, they really like something new. So we need to be there. We also need to be super relevant to what happens right now. If Wicked launches, then, and that becomes really important, then legos with Wicked. If Formula One is really creating excitement among kids, then there's a strategic partnership with Formula One and the LEGO group and we try to push into that.

[Andrew Palmer]

But Christiansen also knows that unbridled creativity can lead to chaos. So a significant part of his job is channeling the enthusiasm of the designers. Big Kids need boundaries too.

[Niels B. Kristiansen]

It's not up to any designer just to design any new brick they would like to. We need to produce the bricks, and we want to produce the bricks as close to consumers as possible. I mean, we make billions and billions and billions of bricks every year. And they need to be produced just at the right time to be packed into just the right box just before the consumers need them. Because with 50% new every year, I don't have a lot of inventory sitting. So we produce it when it's needed, when it gets on the wish list. So how do you know now allow somebody to be super innovative and have all the wild ideas while you restrict the complexity, on the other hand, so you can actually pull it off. Maybe it's the secret sauce of what we actually can do.

[Andrew Palmer]

The Secret Sauce involves giving people like Dan Meehan the chance to let their imaginations occasionally run wild.

[Niels B. Kristiansen]

Once a month we have this Fabulab Friday where all the designers, they can do whatever they want. Faber Lab, Fabolab Friday.

[Andrew Palmer]

Okay.

[Niels B. Kristiansen]

Basically means they can do whatever they want. I mean, whatever skunk project they have or whatever play thing they need to play out or do they just use that so it's 5% of the working time where they can totally freely work outside any project they have because we need them to do that.

[Andrew Palmer]

Meehan had not mentioned Fabulab Fridays to me, but he had talked about another period set aside for out of the box thinking. Come to think of it, at Lego, most things are out of the box.

[Dan Meehan]

We have what's called a Creative Boost week a couple of times a year where the whole design. Org will take a week to work on something. There's a free Creative Boost week, which is early in the year. And that is, as the title says, where if you have an idea, you have a week, make it, and then you present it to the whole design organization. So we have a conference room where we set up little posters and you build your model and you take it and everyone comes and everybody looks. And there's been great ideas that have come through that. We then have a more focused creative booth week where maybe we have an idea like LEGO Space, and we think, hey, we'd really like to make six to 10 products at a variety of price points and sizes and complexities, and we need some help. So therefore, that's when you're saying, hey, space fans. Hey, anyone who's got a burning desire to make something, come and join the team.

[Andrew Palmer]

So far, so fabulabulous. But at Lego, Niels Kristiansen says every idea also needs to fit into a structured process.

[Niels B. Kristiansen]

So at the end of the day, it finds its way into exactly the portfolio representing all the different age groups, all the different themes and partners we have. We work together and just hit the portfolio we think is right for 2026.

[Andrew Palmer]

Okay, all right. So that's your lead time.

[Niels B. Kristiansen]

That's the only time in actually making it happen.

[Andrew Palmer]

Yeah, because you mentioned a film like Wicked or, you know, so you have a. You have a sense of. Part of this planning is the kind of the Hollywood.

[Niels B. Kristiansen]

We have more than a sense. We have total insight on what's going to happen.

[Andrew Palmer]

Total insight. That's me told.

[Niels B. Kristiansen]

I think it's just interesting that despite being super creative innovative, we actually make decisions on what we think will be hot in 26. We do that by the, by the end of 24. So it takes a lot as an organization to really know our bets and make those decisions, be super clear on where we put our eggs.

[Andrew Palmer]

Yeah. So the constraints at the front end on the designers are kind of that partnerships, the calendar that's coming at the back end. It's the modular materials that have to be made. And within that they can play as they wish.

[Niels B. Kristiansen]

Exactly. I think that mastering total creativity on one side and then really very, very structured supply chain and huge factories on the other side and making that gel in a way, I think that's quite special for us.

[Andrew Palmer]

As I left, Christiansen went to fetch what I assumed would be his card. But instead of an oblong bit of paper, he handed me a little Lego figure. We're looking at a little at a Lego representation of Niels. It says Nils B. Kristiansen on the front. I mean, this is your business card.

[Niels B. Kristiansen]

Right.

[Andrew Palmer]

It literally is the email address and phone. Phone glasses are there. Yeah. Pretty accurate. Yeah, it's good. I love it.

[Niels B. Kristiansen]

You could have argued it should have.

[Andrew Palmer]

Longer legs, but you are quite tall.

[Niels B. Kristiansen]

But that's back to complexity.

[Alex Kendall]

Right.

[Andrew Palmer]

Lego is very different to wave a monumental. The startups are trying to make a brand new technology work. At the toy maker, innovation has been industrialized into a potent potpourri of playfulness and planning, creativity and calendars. But there are echoes between them. A clear sense of purpose, whether building walls or driving cars or making bricks. A focus on feedback and a mixture of iteration and new ideas. If these strands connect firms at the start of their journey and titans of their industry, what about a company which has to innovate in response to the challenge of a new technology?

[Workday Narrator]

When you're a forward thinker, you don't just bring your A game, you bring your AI game. Workday is the AI platform that transforms the way you manage your people, money and agents. So you can transform tomorrow Workday, moving business forever forward.

[Liz Reid]

The breadth of what people come to for search like continues to leave me in a sense of awe.

[Andrew Palmer]

Liz Reid knows what you Google, though not in a creepy way. She's the tech giant's head of search.

[Liz Reid]

If you go back to our mission of, like, organizing the world's information, people often, like, stop right there, organize the world's information. But there's like the second part about making it universally accessible and useful. And useful is a very different bar. Right.

[Andrew Palmer]

Reid's task is to get the world's most popular search engine prepped for the era of generative AI.

[Liz Reid]

You can have A medical book in a library that you can check out and it can be available to you, but you may not understand any of it. So great. It's accessible, but who cares, right? And so we constantly have this question about how can you make information more useful to people? And so what we thought about a lot with AI overviews is what are the types of questions that are either not really possible or they're really hard for people to get a good understanding after, and how do we change them?

[Andrew Palmer]

AI overviews are the most visible manifestation of Reid's work on AI in 100 countries. And for many search queries, Google's large language model now generates a summary answer at the top of the page. The feature is changing the sorts of questions that people ask.

[Liz Reid]

What you see with generative AI is that we'll often talk about people asking longer queries. And the question is, why do they ask longer queries? Well, they ask longer queries because they often put more of the pieces of their task in the query. Right, okay. So instead of it just being, okay, I'm going to search for dog friendly campsites, and I'm going to search for kid friendly campsites, and I'm going to search for affordable campsites. I'm like, I need an affordable campsite that is available in this time and allows dogs and will be okay with my kid in a stroll. That's actually the question they had.

[Andrew Palmer]

Why not just search? How can I have the worst possible holiday? The obvious problem for Google is that a good AI summary risks fewer people clicking on the links that power its revenues.

[Liz Reid]

This question of the ecosystem sometimes has this myth that users want either an AI overview or a link, and that's actually not what we see. What we see that actually a lot of people like about AI overviews is the fact that they can get both.

[Andrew Palmer]

AI overviews first went live in America in May 2024. But the story began long before that with Google's early work on large language models.

[Liz Reid]

Different teams sort of had demos or little projects together, and then at some point we sort of started to pull them together and we had more of a sense of what the product is. And I think that's often the case with innovation, where, like, people will be like, aha. There was this like one aha moment.

[Andrew Palmer]

Some of us haven't had an aha moment since 1985.

[Liz Reid]

It wasn't like there was nothing and then there was an immediate aha moment.

[Niels B. Kristiansen]

Right?

[Liz Reid]

You're scratching the surface, you're finding out things.

[Andrew Palmer]

There were multiple versions of the overviews.

[Liz Reid]

In the months that we were building it, it went through many iterations. You would figure out, this felt great, this felt too hard for people, this was confusing, this was too long, this was too short. Search is really interesting because people have an expectation that they can get information very quickly. And so one of the things you're trying to do with AI overviews is figure out that right blend of ensuring that if you came in with a question, you can get to that answer as quickly as possible. And also people like learning a little bit more than they necessarily asked.

[Andrew Palmer]

Right.

[Liz Reid]

And so what's that right balance, even within the response of the clarity. When should you elaborate? How do you help people have ideas of what to do next?

[Andrew Palmer]

As ever at Google, user behavior told the team what was working and what was not.

[Liz Reid]

If you put something bad at the top of the page, this will not cause people to use search more often. This will cause people to come to search less often. Okay, so that can be a signal. How do they engage if they're not happy with results? They often try again. They issue another query. So if you're shrinking the number of times that a user has to, like, issue a query and then try again and try again, that's usually we've seen years and years over. That's a sign that people are happier.

[Andrew Palmer]

The questions people asked also guided the team's efforts. And some of those early questions caused the AI to give answers so weird that they were newsworthy.

[Liz Reid]

Some of those questionable, bizarre results centered around the AI overview telling users that you can eat a certain amount of rocks a day for your health. It's healthy to do so. To be honest, if you asked me before, are people going to ask us, how many rocks should I eat a day? I don't go like, yeah, that seems like a question on top of people's minds. But then they ask that question, you're like, oh, I need to go do something about that. And this is something that's been true well before. Generative AI in search, right? You have a known sort of area where you think you can improve. You open something up, and then people start asking new questions in that space, and then they surprise you with additional questions you never thought of. Nobody searched for opening hours 20 years ago on Google because their belief that they would get opening hours was basically nil, right? So they just didn't ask that question. And then you started to do it. And then people asked more and then they were like, yeah, but what's the opening hours for the breakfast menu? And you're like, oh, my gosh. Okay, this is a new question.

[Andrew Palmer]

Reid is trying to anticipate a fast moving future, but she's keen to emphasize that answers may be found in the past.

[Liz Reid]

I try and be careful, especially because Google's had a long tenure about. Like when somebody says, we've tried that before, most of the things that I think of as being really successful at Google, or at least in the search and map space where I've worked in, they were tried two or three times before and failed. Right, okay. Because the instinct was right, but the tech wasn't ready or the data wasn't ready or something wasn't ready.

[Andrew Palmer]

Like what? What do you have in mind though.

[Liz Reid]

Right now on Google Maps, we have quite a lot of reviews from users and photos from users. We tried that for years before we could get any traction. And then two things changed. The combination of notifications and understanding where a user was. So if I go to you, Andrew, I'm like, hey, Andrew, would you like to review all the restaurants that you've been to in the last two months? You're like, oh, now I have to think of the restaurants and I have to search for the restaurant one by one. Oh my gosh, this is a lot of work.

[Andrew Palmer]

Not really. I had about 800 pret a manger sandwiches, but imagine I had a more varied diet. And Reid's point is that an old idea can suddenly have its moment.

[Liz Reid]

And we know that you went to the restaurant yesterday and we can alert you and say, hey, would you like to write a review for that restaurant? Ooh, that's a lot less work. Right. You just click the notification. You click five stars. Whoa, you've given a rating. And so the bar of friction was really high for us before, and we couldn't do that. And then suddenly the bar of friction for the user got way lower and we were able to interact with users.

[Andrew Palmer]

Data obviously matters to the iterative process Reid describes. This is Google, after all. But she says there's still room for intuition.

[Liz Reid]

I do think you have to mix it with a set of instinct because you're charting new territories. Right. So you can't just look at data before. I think one thing that stuck with me that one of my UX designer peers years ago told me is like, make sure you're falling in love with the problem, not the solution.

[Andrew Palmer]

Falling in love with the problem, not the solution is a nice aphorism. I can imagine repeating it to people as though I came up with it myself.

[Liz Reid]

So if you're really looking at what is the thing you're trying to unlock, then you want to be obsessed with that. And so you don't want to just focus on, here's your first idea to do it. You want to get focused on testing does the solution actually solve the problem for that? I think a lot of the instinct, the question I sometimes push people on with that is like, what is the hypothesis behind the instinct? You can't always measure everything, so you can't always have data. But the instinct is like, this is hard for people, or watch somebody do this and see why it's difficult, or if I could unlock this, this thing would happen.

[Andrew Palmer]

Liz Reid's advice reinforces many of the messages from the other people I've spoken to. Try things out, get as close to the customer as you can and be prepared to get things wrong.

[Salah Al Faji]

I'm just trying to think of our very, very first pilot we did. We didn't even know what to look for. So we focused a lot on wall quality, obviously, because that was important.

[Andrew Palmer]

Back in the Netherlands, Salaar Al Faji is an evangelist for this approach. Being out on the building site, seeing what his robots do in the wild is the way Monumental learns and improves.

[Salah Al Faji]

And then we realized we actually missed two or three key things, including the staining of the bricks, just like visually their color, it sounds so stupid as I say it, right, but like, you're just like focused on like, where do the bricks go, what do they look like? And then we realized like the way we built the wall, the way we laid the mortar, even if you cleaned it up afterwards, like half the bricks were like stained and had mortar on them, which you could not remember.

[Andrew Palmer]

There's no one way to innovate. Lego is churning out new products with a two year lead time wave and Monumental are moving towards a founder's vision. Google is overhauling the world's most popular webpage in midair. But there are common themes. One, forget the 1980s earworms. Breakthroughs very rarely come from aha moments. You often have to work towards something for years and you don't know when the market will be ready for it. Two, Iterate and innovate. A portfolio approach pays dividends whether you do it formally or not. Give people space to work on big new ideas as well as improvements to existing ones. Three, test and learn. Whether it's a building site or London streets or playing children. Get feedback and be ready to change course. There's a phrase that I've found useful for at least the past 10 minutes of my career. Fall in love with the problem, not the solution. Next week I'll look at the aspect of office life that causes people the most stress. Public speaking.

[Alex Kendall]

Beautiful.

[Liz Reid]

Give me one more. I'm going to not look at you this time.

[Andrew Palmer]

In this season of Boss Class, we're releasing an extended interview to accompany each episode. On this week's Meet the Boss interview, I talk with Liz Reed of Google and find out more about how people search, why dog food matters to the firm, and what three year olds can teach all of us.

[Liz Reid]

They ask whatever question comes to mind because they don't worry about how hard it is to answer. They assume you know everything.

[Andrew Palmer]

And coming up on the rest of this season of Boss Class.

[Liz Reid]

I've seen many leaders and managers over my career that get so stuck in the here and now as opposed to let's look around a corner.

[Andrew Palmer]

And I think what the jazz band does is have an alternative for dealing with complexity. Things are going to go wrong.

[Alex Kendall]

The question is, what happens next?

[Andrew Palmer]

What's the next note? So what's the secret of a good negotiator? You just have to be able to shut your mouth and listen well. To listen well, you'll need to be a subscriber. Right now you can sign up for half price. Just a couple of dollars a month. Search Economist Podcasts plus for our best. We'd also love to get your thoughts about what you've heard and what you'd like to hear on the show. Email [e-mail] with the subject line Boss Class. Boss Class Season 2 is produced by Alicia Burrell, Sam Colbert and Pete Norton. Our sound designer is Weidong Lin. Original music by Darren Ng. The series editor is Claire Reid. Our executive producer is John Shields. I'm Andrew Palmer. This is the Economist.

[Workday Narrator]

When you're a forward thinker, you don't just bring your A game, you bring your AI game. Workday is the AI platform that transforms the way you manage your people, money and agents so you can transform tomorrow Workday, moving business forever forward.