The Secret To Innovation - Professor Vaughn Tan

Dr. Vaughn Tan is a strategy consultant, advisor, and researcher.

He currently leads Southeast Asia executive education for University College London, where he was part of the founding strategy faculty in the School of Management.

He is the author of The Uncertainty Mindset (Columbia Univ. Press, 2020), a book about using uncertainty as a strategic tool for designing high-performing innovation organizations.

Vaughn works with the leadership teams of corporations, startups, governments, and NGOS to develop their innovation capacity and ability to manage uncertainty.

His executive education practice focuses on building programmes that develop participants’ leadership capacity, strategic vision, and innovation ability. He speaks regularly at conferences and for leadership teams on uncertainty, strategy, innovation, and the future of work in a time of AI.

He is currently a Fellow of the Singapore Government’s Centre for Strategic Futures, an Rethink Food Executive Board member, and an advisor on diversity and inclusion matters to Wellcome.

CHAPTERS:
00:00 TRAILER
00:53 Understanding Risk vs. Uncertainty
03:30 Navigating Uncertainty in Decision Making
09:05 The Importance of Embracing Uncertainty
14:46 Learning from High-End Cuisine
17:06 Vaughn's Favourite Example
22:07 Designing for Productive Discomfort
28:58 How Desperation By Design Works
35:11 Singapore As An Innovator
39:07 Should Singapore Re-Invent Its Education System?
45:37 Beyond Credentialism In Adult Learning
52:23 Reframing Singapore's Economic Strategy
57:53 Closing Questions

Keith 00:00:00

Right now as we speak, the US is experiencing the liberation week. Donald Trump introduced a slew of tariffs for all their trading partners and there is a sense that he has injected a lot of uncertainty into the global economy. By the same time, a lot of people are saying that this is very risky as well. And part of my question that kind of bubbled up as I was looking at this behavior was:

There seems to be a difference between uncertainty and risk. What do you think is the difference?

Vaughn 00:32:00

There absolutely is a difference between uncertainty and risk. We tend to use the two words almost interchangeably. Whenever there's an unknown, usually in the future, we think of it as something we don't know what's going to happen. But the words we use to refer to it are either uncertainty or risk, as if they mean the same thing.

When we use the word risk, we often think about how to manage this unknown thing. We turn to risk management, cost-benefit analyses, and expected value calculations to decide what to do. The problem is risk and uncertainty are actually two different things.

In a situation of risk, you don't know exactly what's going to happen, but you know almost everything that you need to know about all the possibilities. You know all the possible actions you can take. You know all the possible outcomes that might result. You know precisely and accurately the connection between an action you take and an outcome that might result. As a result, you can do things like cost-benefit analysis and expected value analysis.

You point out that Liberation Week is here with all these unknowns. But ask yourself: Do you know all the possible things that Trump could be doing? No. Do you know all the possible outcomes that might result from things like a tariff war? No. Do you know the probabilities accurately and precisely that connect an action that the US takes and an outcome that might result? No.

That's not a risky situation. You almost don't know anything about what's going to happen. That's what I would think of as true uncertainty. Risk is something that you can quantify with precision and accuracy. Uncertainty is something that you cannot quantify, that you don't know enough about to quantify.

This isn't a novel idea. The economist Frank Knight came up with this distinction a long time ago, and lots of people know about it. Yet people still use these two words interchangeably in a conceptually loose way that makes thinking about uncertainty and risk very difficult and usually not done very well.

The simplified version is: in a situation of risk, you know almost everything there is to know about what you don't know, and in a situation of uncertainty, you don't. You should not mix the two up because they are fundamentally different things as we've just discussed in relation to tariffs.

Keith 03:02:00

What's the cost of mixing it up?

Vaughn 03:04:00

If you assume that an actually uncertain situation is just risky, and then you apply risk management techniques to making decisions in that situation, you will frequently make the wrong decision.

Let me give you an example. Remember back in February 2020, very early in COVID? Big cities in China had shut down. We were beginning to see this weird disease. We didn't know how it spread, how fatal it was, how likely it was to transmit when exposed, or who had been exposed.

In that situation, sometime in the middle of February 2020, the World Health Organization came out and recommended against any kind of travel or trade restriction. The reason was that they had done a rigorous cost-benefit analysis about the costs versus benefits of travel restrictions.

But how do you do a rigorous cost-benefit analysis if you don't know almost anything about what the disease is, how it spreads, how quickly it spreads, where it's coming from, or what it is? You can't.

As it turns out, their decision not even to suggest pausing travel for a few weeks to figure out what was happening led to the world we all lived in—very long shutdowns because the disease spread very fast before we knew how to deal with it.

The consequences of treating an uncertain situation as if it's merely risky are that you make the wrong decision. Another example is banks and financial institutions that assumed complex derivatives were predictable and understandable until they realized very late in 2008 that you cannot actually predict what's going to happen with these because they're too complicated and interconnected.

Keith 05:09:00

In that case, how should one think about uncertainty?

Vaughn 05:14:00

In a situation of risk, you're trying to optimize. You know so much about the things you don't know that you can figure out: "If I take this action, that outcome has this probability of occurring and I value that outcome in this way." This means you can optimize and choose the probably correct action in a comfortable, predictable sense.

If you're in a situation of uncertainty, you can't do that. You have to fundamentally rethink how you deal with the unknown. One approach is to take smaller bets and take more of them. Another is to think about learning over time what the situation looks like.

In a situation of risk, you tend to figure out the right approach and then take a big bet because probably the bet will be right. In uncertainty, you can't take a big bet. What if you take a big bet like the World Health Organization did on recommending against travel restrictions, and it's wrong? The consequences of one big wrong bet are very significant.

I think we need a fundamentally different way of thinking about it. It's not about optimizing a singular monolithic action. It's about taking lots of well-designed, informative actions that are like little experiments, learning what to do so that you figure out the parameters of the uncertainty you're facing.

Keith 06:35:00

So if you were in the World Health Organization back then, you would probably say we should try to pause travel for a few weeks?

Vaughn 06:44:00

Yes, taking a few weeks off would be very disruptive and there would be costs of doing that. But if you're in a situation where an entire very big Chinese city had shut down completely before they decided to recommend against restricting travel or trade, I think with hindsight, we should have at least shut down travel for a couple of weeks just to see what was going on.

We could have used that time to update our prior understanding. I would have suggested experimenting with smaller interventions that would have a cost, but not as much as some really big pronouncement. You're not saying shut down travel and trade for three years, even though that almost happened. You're saying shut it down for a couple of weeks, then review and see. Let's try other things, maybe smaller things. And if those small things appear to work or give us more information, that's a good way to approach this uncertain—not risky—situation.

Keith 07:44:00

You remind me of the analogy that back then the ministry task force used: just a circuit breaker. I don't know whether it's the right analogy, but the frame of thinking of small adjusted bets using that decision to update the kind of information model that you have.

Vaughn 08:03:00

Yes, for sure. And in reality, this is how humans learn. When you are growing up as a small child, you're not taking a single bet on a single method of learning. You're just in the world doing your thing, experiencing stuff, and the things you experience update some internal model about what the world looks like and how you act inside it. This is how humans learn naturally.

I don't know why we aren't thinking about applying something which is such a natural part of how we understand the world to a world that is clearly not risky, but instead increasingly uncertain.

Keith 08:40:00

But it seems like the world today likes to shun uncertainty. The more mathematical your financial model is, the more you feel certain about things. And there is this inclination that with more tech, we can escape uncertainty. But you probably argue that uncertainty is crucial. In that case, why is uncertainty actually important? Why should we learn to embrace it?

Vaughn 09:05:00

I think there are two parts to this answer. First, there's a reason why it's so tempting to try and say that everything is modelable, predictable, optimizable—because that's comfortable. I suspect we are evolutionarily opposed to the idea of not knowing what's going to happen. From a physiological perspective, there's a stress reaction.

You feel stressed out and a bit frantic when you confront something unknown. We don't like it cognitively or in an affective sense. You just feel bad when facing uncertainty.

The reason for this is presumably that back in evolutionary time, if you were entering into an uncertain situation, it was actually existential. If you were walking from someplace to your cave, you could either take the path you know well, where you know all the animals and dangers, or you could take a shortcut through a dark forest with unknown dangers. You don't go into that kind of uncertainty because it's existential.

But nowadays, the kinds of uncertainty we face usually aren't like that. Our evolutionary instinct to be incredibly stressed out by uncertainty is no longer adaptive for us. So we should probably change how we think about uncertainty because it has different implications for us now.

The second part is that if we want to do anything new, by definition, there is no option other than to do something which is uncertain. Risk, formally defined, is when you know almost everything there is to know about what you don't know. In that situation, you are never doing any true innovation because you already know what's going to happen.

The only time you have true innovation is when someone does something that people say won't work or that nobody will want. We have lots of examples of this. The best example is probably the iPhone. First, people said this weird interface wouldn't work—nobody would understand how to use it. And when it finally came out, many people in the phone and computer industry said nobody would want it. Then, less than 20 years later, everybody's using a smartphone.

The innovation wasn't just hardware; it was an innovation in understanding what people want. If there is no uncertainty, there is no possibility of doing something new that addresses new demand.

You see this again and again in every kind of innovation. The only way new things happen is if you're in an uncertain situation about what kinds of actions are possible, what kinds of outcomes might result, what kind of causation exists, and what things people value.

That's why uncertainty is important. Not only is it no longer adaptive for us to treat all unknowns as something uncomfortable that we think of as optimizable and risky, but if we care about doing new things—because new things help us adapt to our increasingly weird world—we really need to understand that uncertainty is essential for doing anything new.

Keith 12:46:00

It reminded me of the classic Henry Ford phrase: "If you ask the consumer what they want, they'll just ask for a faster horse" so you'll never get a car. In a sense, that frames how we should think about uncertainty—to think of a future which is different from ours, to embrace that uncertainty. That's the only way you can get any form of innovation going.

Vaughn 13:10:00

I'd actually never heard that phrase from Henry Ford before, but it makes a lot of sense. Generally, when you're trying to do something, you're trying to figure out how to get as much certainty as possible that what you're doing is right. That's fine if what you're trying to do already exists and you're just deciding whether you should be doing it or not. But if you're trying to do something brand new, where would you go to find out whether the brand new thing actually works? It doesn't exist yet.

That's a crucial insight. You cannot do anything really new and figure out whether it's right to do by looking in the world and asking people if they want that thing.

There's also another kind of uncertainty that nobody really talks about, but I think it's very important: the outcome that you're imagining. If it doesn't exist yet, very few people can think about whether it's valuable.

Before women had the vote, the default position was that women weren't citizens who were valuable enough to vote. Before women had the vote, very few people thought women should have a vote. There were lots of protests against the possibility of women's suffrage. And now if you ask people whether women should have the vote, the answer is obviously yes.

Our value systems—what kinds of outcomes we value—those things change too. There is uncertainty about them, and that's what creates the possibility for different ideas about what's valuable. I think it's a really important kind of uncertainty for us to have.

Keith 14:46:00

When people think about innovation, they think about iconic brands like Apple and Tesla that you previously cited. But they don't think of high-end cuisine, which is something that you spent years working on to figure out why it requires innovation. I'm curious, why did you start with high-end cuisine as a place to study from?

Vaughn 15:10:00

Well, the honest answer is it's more fun than studying microchip foundries or cars. But there is also a technical answer. If you want to study innovation in a company that makes microchips or semiconductors, the cycle for innovation is long. Going from an idea to a plan for a semiconductor to actually producing a prototype chip to launching it can take years.

If what you're trying to understand is what kinds of group processes, structures, and behaviors make it possible for people to ask questions that have never been asked before and answer them well—that's the innovation problem—do you want to study things with a very long cycle time? Or do you want to study things that can happen in a couple of hours or days?

Because your own time as a researcher is limited. If you study something that takes two or three years for a full cycle of innovation to complete, you only have one example if you spend two or three years watching it.

If, on the other hand, you study a bunch of cooks trying to come up with a new idea in food, a new way to use an ingredient, a new way to combine ingredients on a plate, or a new story to tell about a plate—some of that happens in a few hours, some in a few days. If you spend six to eight weeks with them, you can see 40, 50, 60 cycles and start to see patterns.

So that's what I tell people, which is true. But the reality is it's also more fun.

I think it's useful to look in unusual places that are informative about phenomena you care about. Innovation is happening in all sorts of places. The fact that we only look at obvious, legible places like big tech companies gives us a very warped idea of what innovation is and how it's done. Looking elsewhere gives you other perspectives that generate alpha, if you want to call it that.

Keith 17:06:00

When I was looking at your case studies, many showcased both the uncertain environment that these restaurants operated in and the ingenuity needed to develop either a product or process innovation. Were there any examples that particularly stood out to you that changed the way you think about innovation?

Vaughn 17:32:00

That's a great question. I think there are maybe two things that were most interesting.

One was about hiring. If you've ever done any hiring—I used to be on the hiring committee at Google—you always have a detailed requisition for the person that says what you expect them to do. You recruit people who seem like they're a good fit, and when you're interviewing, you're trying to figure out if they're as good a fit as they look on paper. When you choose who to hire, you're hiring for the best fit to this requisition.

That approach has made sense to HR professionals for decades. But if you're doing innovation work and you don't know what the work will have to be in advance, how do you write a scope of work? You can't. The moment you write it, it's out of date.

If you use that to hire people for innovation, you are selecting people against a requisition that is already outdated when you wrote it. By the time you hire them—in two weeks if you're lucky, but usually more like three months—it's very out of date. Even if it wasn't out of date on the day you hire them, by the time they're in the company for three or six months, it's even more outdated.

This way of hiring people, which almost every company uses, makes no sense at all based on my research. What these R&D chefs were doing in their companies was not being hired against a scope of work at all. They were being hired against the idea that there are some things they absolutely must know how to do, but it's a very small part of their actual work.

Let's assume that maybe 30 or 40% of your work is stuff you must know—for a chef, maybe how to clarify a stock, which every junior chef eventually learns. That's a baseline. But beyond that, most of what you have to do, you have to figure out with the team. You have to try things out. You have to figure out what we need that we don't even know about. You have to show by trying it out that you understand what we need and prove it to us. If we think it's good and you like to do it, then that becomes part of what you do.

This turned into the idea of open-ended roles that you shape through a process of negotiated joining. Nobody really does this in conventional business, even though if you're trying to run an innovation company or a company that makes new products, you really should. In the few companies I've worked with to help redesign their hiring process, it dramatically changes the efficacy with which you hire people in innovation capacities.

Innovation capacities are one type of function where you do uncertainty work a lot. But really all leadership roles are like that too. As long as the business or operating environment is changing unpredictably and frequently, you have to do things you weren't expecting, and your job role changes all the time. You should be hiring with negotiated joining into open-ended roles for sure.

The other insight from looking at these R&D chefs was how good they were at doing things they didn't know how to do and might fail at. I don't know about you, but I'm terrible at doing things I don't know how to do. I hate the prospect of failing. It's an active effort to get over that hump.

But many of the people I was interviewing and working with in these R&D kitchens didn't like failing, but they were very comfortable with doing things where they might fail. They were okay with the discomfort that comes from not knowing what you have to do and how to do it yet.

After looking at those chefs, as well as working with startups and small business founders, that single characteristic—the ability to be uncomfortable but still productive, the ability to see that discomfort from uncertainty as something you cannot avoid but must embrace and learn to deal with—that's a big differentiator. And it's something which is lacking in many corporations and public institutions.

Keith 22:07:00

What about the environment that allows them to be uncomfortable yet productive?

Vaughn 22:13:00

I think it's less about the environment and more about what they have chosen to do themselves to learn how to deal with the physical sensation of discomfort.

Remember I said we've probably got an evolutionarily evolved stress reaction when we face uncertainty. When you face the unknown, your stress hormones rise—cortisol, adrenaline—giving you a physical sensation. That's probably legitimate if you're in a fight-or-flight scenario. You want those stress hormones so you can act very fast.

But do you want that to stop you from doing something that creates a stressful situation when that thing might help you learn something new or create a new product? No.

So what you need to do as a person trying to learn how to be productively uncomfortable is to learn how to tame that very bodily response. And what these chefs were doing was practicing.

One example of a similar system that many people use is progressive overload in resistance training. If you're lifting weights and have never lifted before, you don't immediately do a chest press with 80 kilos. You start with what you can do. Once you can lift 15 kilos, you slightly overload and build capacity. Then you rest for the capacity to build up. Then you overload again and again.

Dealing with productive discomfort means learning how to do that kind of progressive overloading with your exposure to uncomfortable situations that help you learn.

These people were frequently and deliberately putting themselves into situations that were uncomfortable, but where if they managed to get through the discomfort, they learned something new.

Many people in my book come from unusual backgrounds and then entered cooking because they were interested in it. If you come from biochemistry research and want to enter a professional kitchen, it's very uncomfortable. Everyone knows how to do things in kitchens, they know how to be in a kitchen. There's a lot of tacit knowledge nobody teaches you. And you come in with this other valuable knowledge, but you don't know how to be there. It's very uncomfortable.

It's like going from one school to another—you don't know the language or who the cool kids are. But if you manage to adapt and make new friends, you have more experience than anyone in that new place. This is what was happening with all these people. They were learning how to be uncomfortable by being progressively more uncomfortable and by designing their discomfort so it was more likely to result in learning rather than just random stress.

Keith 25:10:00

There is this idea of design that you also talk about—"desperation by design"—in which the leader throws the team into something beyond what they can handle.

Vaughn 25:24:00

I would zoom out from the culinary context and think about the general principle. One thing, again about comfort, is the idea of a comfort zone. It's a zone in which everything you do, you know how to do reliably, you understand the resources needed and the processes for working with others. People understand. It's easy.

The problem with that comfort zone situation is there is no growth. You may become more efficient, but you don't necessarily rethink what it means to be effective. The only time you learn as a team is when you push outside the comfort zone—when you're asked to do something you don't know how to do yet.

If you push too far, the team breaks down. Imagine a team designed to run a restaurant. They're very good at it. Now you ask them to run a conference for frontier thinkers in food systems. They've never done anything like that before. There's a possibility it requires skills they don't have and cannot acquire in time, causing the team to break down.

But if it doesn't break them, it forces the team to learn something brand new. It forces them to absorb information and people they don't currently have because they realize they need skills they lack.

It forces them to learn to do different things with each other. If I used to be a sous chef who knows how to cook and pass my dish to the sauce chef and then to the pass for plating and garnishing, now I also need to learn how to talk not only to a famous chef but also to a famous food agroecologist.

That kind of learning, that change of process, and the idea that it's okay to not know what to do—these are the benefits of pushing people beyond their comfort zone, putting them into a situation of discomfort and maybe even desperation.

The reason you need to push into desperation is because when people feel everything is going to be okay, they never work as hard and aren't willing to let go of comfortable things. That prevents learning. If you're not in the discomfort or desperation phase, you're not willing to do whatever it takes to get past that phase.

If you look at a team that has been put into a "desperation by design" situation, at first they're just annoyed they're doing something they don't know how to do. At some point they realize that to get it done, things must change. That becomes desperation.

This relates to John Kotter's model of organizational change where you have to unfreeze an organizational structure, change it, and then refreeze it. Unfreezing usually takes a lot of work, but sometimes it can happen just because you created a desperation event calibrated not to break them down, but just enough to unfreeze the situation so they can change before refreezing. The desperation is a crucial part.

Keith 28:47:00

How does one begin to think about unfreezing or begin that desperation project? Obviously you talk about the progressive overload of starting with what you think they can handle and pushing them slightly beyond.

Vaughn 28:58:00

The thinking process behind desperation by design is not just to assign a task that the team doesn't know how to do. Many leaders do that—they assign stretch goals with no understanding of how the team can achieve them. That often leads to serious overload, team collapse, and relationship breakdown.

Desperation by design is actually a much more carefully managed process. There are a few things you need to do.

One is to ensure there is external commitment to the goal that causes desperation. It can't just be some fake exercise. There has to actually be a business consequence. Don't just create a task for a team to learn something new if the result never becomes part of business process or a product that gets sold.

If you don't have a real task, people don't commit. If you have a real task, but it's not committed to either people outside the team or to actual customers, there's no accountability. You can't just say, "Let's do this real thing, but we won't tell anyone about it because if we fail, that's okay." The moment you tell someone outside whom you trust or respect that you're going to do this thing, you're committed to doing it. That can push you into desperation as well.

But perhaps the most important things with desperation by design are: One, as the leader, you must have a very close connection to the affective state of your team. When you propose something that pushes beyond the comfort zone, the non-cognitive reactions—the stress reactions—tell you whether you're pushing too far. When you work with your team frequently and consistently, you start to understand when they react and when they overreact. You can calibrate whether you're pushing them too far.

The other really important thing is, like with resistance training, you have to give people time to rest. Doing something which is desperation by design is very hard. It takes a lot of energy—not just physical energy but cognitive and emotional energy to rethink how you do things and change something comfortable, which is stressful and emotionally difficult.

You have to give people time to rest after that. If you do, then it's almost like a muscle for dealing with discomfort. If you build that muscle by doing something desperate and uncomfortable, and then give time for that muscle to rebuild, the next time you do something beyond their capacity, the new thing can be further beyond their capacity than the first thing. You gradually see a buildup of capacity if you do those four things.

Keith 31:57:00

The three points you made about it being tied to real outcomes, having some kind of accountability mechanism, and having that effective relationship between the leader and the team—that is really illustrated clearly in the high-end restaurants you talked about in your book. Because when you're doing product innovation or process innovation, it leads to maybe a better entrée or a better main course for the patrons. Were there any favorite examples where you saw this desperation by design fully fleshed out?

Vaughn 32:35:00

The best example, the one that's in the book, is a restaurant in Copenhagen called Noma. It became most prominent around the time when I was joining them—I was hanging around in the kitchens a lot. They were a restaurant very good at telling a story about where they're located in Copenhagen, Arctic Circle cuisine, cooking and eating food using techniques and ingredients from very close to where they are.

That's what they did for 10 years and became famous doing it. Then one year, 2011, the guy who runs it, René Redzepi, decided not only should they be a good restaurant, they should also help the food industry rethink what it means to do food well. This meant running a conference, bringing people together, sharing ideas.

His restaurant doesn't do conferences. None of his team had ever done a conference before, but he decided to do it in just three months. I think he came up with the idea in June, and the conference was in September.

That was the first example I saw of: "We don't know how to do this at all, but the leadership team has decided and consulted with us that we should do it. We don't know how, but maybe we can try." It was quite desperate. They didn't know how to organize it or whether they could get people to come, but they did it.

Over time, the fact that they did it meant that every single one of these symposiums—it was called MAD, the MAD symposium—pushed further away from comfortable topics in the industry.

In the food industry, for a long time, there was a lot of drug use and mental health problems that nobody wanted to talk about. But one of the MAD symposiums eventually focused entirely on these issues.

If you start from a point where you do something uncomfortable you've never done before, and you keep pushing the team so they become more able to do uncomfortable things they've never done before, you keep climbing this ladder. Eventually, they now have a school. They've turned it from a conference they'd never organized before that was very small, about new ideas the food industry needs to talk about, into a school that teaches people in the food industry how to change the food system. Now they've got MAD Academy in Copenhagen. And all that started from doing something difficult but doable that caused desperation.

Keith 35:11:00

And in that sense, they demonstrated not just effectiveness, but also efficacy.

Vaughn 35:17:00

Yes, for sure.

Keith 35:41:00

It brings me to a question about institutions and governments. Singapore is one of the model nations that many countries look at today and say, "Wow, it's very effectively run." There is also this idea that we were innovators in many ways—the way we started out, the institutions we built, the philosophies we adopted. We didn't take the free market approach, but we didn't exactly embrace communism wholeheartedly.

But at the same time, we've become so successful, so effective, we've kind of frozen in our own way. My question would then be, do you think Singapore has that trap or might face that trap of becoming so effective that we kind of lose efficacy?

Vaughn 36:10:00

Yes, absolutely. I don't know as much about Singapore history as I should, but I know a little bit because I grew up here.

I think we can interpret what happened from the 1960s until the 1980s-90s as a situation where things were actually pretty desperate. It wasn't clear in the 1960s whether we would survive as an independent nation-state. This is part of our founding narrative.

If you're in that situation, you're willing to do things you wouldn't be willing to do if you're comfortable. We were willing to take pretty big singular bets. The first generation of political leadership were all brilliant. When I was younger, I was much more critical of what I saw as a repressive regime. I actually don't think it's that repressive at all now.

But one thing that is definitely true about that first generation of political leadership is they knew there were difficult choices to be made. The answers were not certain. You don't know what the right answer is to the bet you take. Should you invest in upskilling your population, or should you try very hard to make a country able to deal with regional dynamics rather than influences from further away, as we did?

None of these questions—including infrastructure and economy questions—had correct answers, and all were difficult decisions. That is the kind of strategy under uncertainty that I think is extremely important, which they did very well back then.

But I think they were able to do that because it was very clear to everyone that Singapore was in a stressful, somewhat desperate situation. There was no certainty about whether we'd survive. They didn't have to convince the population that was true because it was obvious.

We're now many decades after that, and things are very comfortable. So now you need to convince the population, the electorate, that things are uncomfortable in order to be able to take more risks.

That may not seem like something you should do because things are comfortable enough to not have to take those risks. And not all the things we're asking people to do are risky in the "we know everything about what we don't know" way. Many things we should be trying to do are also uncertain—things that will help us learn whether there are possibilities that have never been done before or haven't been done here before that might be good to do.

I think the dilemma we face is we are too comfortable, and it's very obvious that we're comfortable. When you're comfortable and not used to being uncomfortable—cognitively, emotionally, economically—you don't tend to do things that are different. That's probably a big problem for us.

Keith 39:07:00

So now I'd like to put you in a bit of a spot. One thread in your writing and work is an emphasis on education. For example, the restaurant does this learning by doing, the chefs are constantly learning, and they embrace uncertainty through learning. Learning plays a huge role.

In Singapore, the Ministry of Education is focused on educating the young, teaching them to embrace or to operate in an uncertain world. If you had the ear of the Minister of Education today and they asked for advice saying, "It's clear that we're living in an uncertain world. I need your help to revise the curriculum or shape the system better," what advice would you give?

Vaughn 39:59:00

Great question. One that's obviously very close to the things I'm interested in.

I think from my external view, the Ministry of Education is trying to do many things. They're teaching critical thinking, computational thinking, and what they call 21st century competencies. That's clearly an area where there are efforts.

But very frankly, whenever you have a system where almost all of your ability to do what you want ends up being tied to something easily quantifiable, like a test score, that washes away all the other stuff.

One thing we should probably do, although I can't see how it's possible given the infrastructure we've already built up, is create ways for people to do highly desirable things like getting into a top university in Singapore without focusing so much on test scores.

Some ways to do this—that may not work here, but just as ideas—would be to have a small population of people who could get in not randomly, but based on things that are explicitly not test scores.

I think we can start to think more creatively about encouraging people to demonstrate success in ways that aren't about conventional, easily legible performance.

Another thing we could do that's a bit more radical is to think about what we can do inside our pedagogical framework to introduce people to uncomfortable ideas more. That's not an explicit objective of our education system.

I went through the system all the way through JC in Singapore. The way we framed learning for the most part was: there are correct answers, the correct answers are in a framework you get tested on, and the arbiters of correctness are teachers.

The only reason I didn't fully believe that was because I had amazing teachers all the way through who said, "This is not the right way to think about it." In reality, there are some facts, but interpretations of facts and what things you value are diverse. That means correct answers can only generally be about facts. They can't really be about interpretations. Interpretations have to be argued into existence. You have to be able to justify them.

If we had an education system that was better at teaching people to do that kind of uncomfortable consideration of whether or not you believe something, why do you believe something—in philosophy, you would call this epistemic reasoning—that would be a good thing. We should never call it epistemic reasoning or epistemology in schools because that would turn people off, but we should be teaching them to question.

If we say something, for example, provide an interpretation of why Singapore became independent in 1965—do you believe it? Where is the data coming from? Do you believe this data? Is there missing data that would give us a different interpretation? That's just for history. You can do this for almost every subject.

If we taught that instead of teaching people to remember facts and then regurgitate them in exams, which I know they're not trying to do, but that's just how it works—if we explicitly said, "Let's not do this, let's do it this other way instead," maybe we'd get a little bit further.

I think also, to the point of progressive overload, having other ways for people to be exposed to uncomfortable things is probably a good thing. How that looks for different age groups will be very different. For adults like you and me who are well-versed in living in the world, exposure to uncertainty can be something you do as a game.

We often need prompts to force us to do something uncomfortable but designed to expose us to new ideas, new ways of thinking. Maybe we need to think about what those look like for younger children as well. That's something I've been thinking about a little bit. I don't have answers yet, but it seems like an important thing to try.

Keith 44:52:00

It seems to me that once we hit adult learning, there's almost a reversion to credentialism. Some of the things you talked about, like when we use exams to test a student's aptitude, that washes away other maybe more important but seemingly inconvenient things.

It translates a lot into adulthood. Because when people think about adult learning in the workforce, they talk about upgrading skills and staying relevant. The way people frame it is, "Let me go and do some course, get a piece of paper, and then I can say I have a skill now." Why is that the wrong way to think about learning? And would there be a better alternative?

Vaughn 45:37:00

It's not necessarily the wrong way. What I would say instead is that of the ways you can upgrade people or give them the ability to do more things if their previous skills have become less valuable, credentialing is probably one way, but shouldn't be the only one.

Traditional education provides several things: a network where you meet peers and learn to work with them, which is cool; actual learning of content; and then of course, the credential as a signaling mechanism.

Nowadays, especially for continuing education—I do some of that for UCL—the credential often doesn't seem to map very much to learning new ways of thinking. It doesn't often translate to changing how people approach work.

One thing we could do to get away from credentialism in adult education or re-skilling is to change how we hire. Instead of looking so much at the list of things on someone's CV that make us think they're a good fit for our job scope, we could do something like negotiated joining on open-ended roles.

How could you implement this with people late in their career trying to change industry? You could say, "I can't hire you full-time yet, but I want you to come in for two to six months on a pro-rata full-time salary. While you're here, show me what you can do."

You would still do an interview or two upfront to determine if they're people you want to work with, but most of the learning about whether they can do the job, and most of their learning about what they need to do, should come from actually doing the work, rather than from lines on a CV.

The idea of negotiated joining and open-ended roles is about helping an organization and a person figure out what they do in a situation where work can't be programmed in advance. It also can create space for people changing careers to show that their prior experience might be relevant in ways that you as the employer don't recognize. That seems like the more effective way, even though it looks inefficient and messy, to give people opportunities to try new things instead of just incentivizing them to credential up.

Keith 48:32:00

What prevents companies from trying this? I assume it increases the friction and transaction cost of joining the company.

Vaughn 48:49:00

One barrier is that most hiring goes through HR talent organizations. They have best practices evolved from business school research and looking at what other companies do. It appears frictionless, but if you ask anyone in an organization about the real cost of hiring people who aren't a good fit, it's substantial.

You hire someone, they know they should stay at least a year and a half or it looks bad on their CV, so they stick around. They're not very productive. You want to get rid of them, but can't really. They want to leave, but think it's not a good idea. They act as a drag on the company.

So there is a hidden friction to hiring the wrong way. Often credential-based hiring doesn't get you very good people. It gets you people who are "good enough," but does it get you people who tailor what they do to what you need based on capabilities you don't know they have? Usually not.

The first thing I'd say is that you may be overlooking the fact that your existing apparently low-friction hiring process is actually very high-friction, just in a way that's longer-lasting and less obvious.

The other problem is that most HR people follow best practices that don't talk about this approach at all. If you're an HR practitioner running HR in your company, a good way to think about changing this is to look at the kind of work you're hiring for.

If you're hiring for work that is very programmatic, where you understand exactly how it needs to be done and how to measure success, you should probably hire the conventional way. But for anything to do with product innovation or operating in an industry where things are changing dramatically—and this applies at all levels, not just senior roles—you should consider a different approach.

Even a customer service person in a market where consumer demand is changing a lot can't be expected to succeed using a static script. They have to be able to learn what customers want as it changes.

If you're hiring for this kind of uncertainty work, you should probably be hiring differently. Negotiated joining into open-ended roles sounds messy and frictional, but it's actually very easy and low cost. It just requires changing how you think about hiring. And for the right potential hire, it's also a more engaging way to join. You'll automatically filter for the right people based on who is interested in doing it.

Keith 51:53:00

Singapore has built itself up as a financial hub and high-end manufacturing hub. The traditional pitch from an EDB officer might be that Singapore is the gateway to Southeast Asia. With that, we've developed startups that look at reaching the region and trying to scale up as fast as possible. Should we frame ourselves in a different way?

Vaughn 52:23:00

Up until now, our strategy has been to bring in large companies with established business models that want to have a Southeast Asian or Asian location to operate from. That's basically EDB's function—find potentially good companies to locate their Asian headquarters or R&D facilities in Singapore.

I think our strategy for economic development has been very subtly but profoundly affected by this idea that we need to go big and go with already successful business models. I don't think this actually works for the Southeast Asian context. Here's why.

A lot of the companies we look at as being successful tend to come from very large markets where it's possible to have a huge total addressable market. The US has about 280 million people, so you can build a very big company that reaches massive scale just inside the US. Similar with China and to some extent India.

If you look at Southeast Asia, we appear to have big countries like Indonesia. But Indonesia isn't one single country in many ways—it has many languages and is extremely geographically spread out. Southeast Asia is actually a region of relatively small markets.

In this context, trying to build companies from Singapore—tech companies or any other kind—that have some massive total addressable market is a dream that will likely never become reality because the market doesn't exist that way.

Instead, what Singapore should be doing is something that East Asia has been doing for a long time: build conglomerates composed of companies that are individually quite large, but never have the kind of total addressable market we're aiming for. They have markets big enough for the size they need to be.

Singapore has a bunch of policies to encourage venture capital for tech startups. The economics of venture capital always require enormous total addressable markets. You need a unicorn to pay off your portfolio. This is fundamentally incompatible with the reality of Southeast Asia.

If the government wanted to encourage a tech ecosystem in Singapore, one thing it could do is encourage the size of tech startup that a VC would never fund, but that a private equity firm might want to buy—cashflow positive very quickly, addressing existing needs in conventional industries, and intentionally aiming for a right-sized total addressable market for parts of Southeast Asia.

Singapore is one of the smaller markets here. But if you're doing insurance or finance, you shouldn't be trying to build a unicorn-sized thing invested in by VCs. You should be trying to build something that's cashflow positive, solving existing problems that, for example, the insurance business in Singapore needs to solve. That becomes a business that generates cash that a private equity company might want to buy.

It's a fundamentally different approach to thinking about what kinds of tech startups and companies we should be encouraging in Singapore—smaller and much more oriented at solving existing problems rather than creating new markets that don't exist yet, which require a lot of VC subsidy to build.

Keith 56:11:00

You make a convincing case, but a VC might say it is possible. I've interviewed Magnus Grimeland who founded Antler, and his view is that if you're able to exploit the different comparative advantages of each region or country, you can build something that solves problems across the board. That's probably true, but the success rate might be much lower than in the US, which is why after a decade of money pouring into VC-funded startups, we looked at the results and saw there weren't that many successes.

Vaughn 56:54:00

Magnus's point makes sense. It is possible. In fact, it's plausible that the competencies or needs across different markets in Southeast Asia have some correlations. But the reality is the VC game is already one where, if you're lucky, for every 10 companies you fund, nine don't work at all, and one, if you're lucky, pays off the whole portfolio.

If you reduce the odds even more, you have to ask yourself: is VC the correct conceptual model for capital allocation in Southeast Asia? My answer is no. If you're super lucky or incredibly insightful and able to have even the same returns profile as traditional VC in China or the US, fine. But even that model is clearly not working very well. So why try to do it here when there's a different model that might work much better?

Keith 57:53:00

I have two more questions for you. These are my closing questions that I try to ask everyone. First, what book do you recommend that everyone read? And second, what piece of advice would you give to someone coming out of school and entering the workforce?

Vaughn 58:09:00

The book I recommend reading is quite strange. It's called "Seeing is Forgetting the Name of the Thing One Sees," and it's about how you can continue to do things over your entire career and do them very differently while still trying to do the same thing over time.

It's actually an intellectual biography of an artist I love, Robert Irwin, written by Lawrence Weschler. I won't say any more about that. It's a very strong recommendation. There's no good reason why you should read it, but it's an amazing book.

As for advice for people coming out of school—the earlier in your career you are, the more uncertain things you should try to do, because that's when you have the least downside from doing something that fails.

Everything I've said before feeds into that, but that is absolutely what I would say. It's very hard to see this happening widely, but there are some people who advocate for it. More people should be advocating for it, especially in schools. Schools should be telling people to do something weird while in school or just after. What's the downside? Don't do anything illegal, but do something weird.

Keith 59:31:00

With that, thank you Vaughn for coming on.

Vaughn 59:33:00

Thanks for having me. This was fun.