The future of entrepreneurship

0:00This is Stanford Engineering's The Future of Everything, and I'm your host Russ Altman. I thought it would be good to revisit the original intent of this show. In 2017, when we started, we wanted to create a forum to dive into and discuss the motivations and the research that my colleagues do across the campus in science, technology, engineering, medicine, and other topics. Stanford University and all universities, for the most part, have a long history of doing important work that impacts the world. and it's a joy to share with you how this work is motivated by humans who are working hard to create a better future for everybody.

0:37In that spirit, I hope you will walk away from every episode with a deeper understanding of the work that's in progress here and that you'll share it with your friends, family, neighbors, co-workers as well. It is quite important that we be thoughtful about AI regulation and that it's not just people at the table that are the big players that are shaping those regulations in order to make sure that we're winning the AI race, not just with the current generation, but also what creative, innovative entrepreneurs are going to do with AI in the future.

1:17This is Stanford Engineering's The Future of Everything, and I'm your host, Russ Altman. If you want to shape future episodes of the podcast, please write a review, Tell us what you think. And of course, give us a rating. We like 5.0 if we deserve it. So please rate and review. It helps the show. Today, Chuck Easley will tell us that entrepreneurship is key for fueling the economy, but that it requires a context of support from institutions in regulation and in culture. But with those regulations and cultures, people with good ideas can succeed. It's the future of entrepreneurship. Today, we're continuing our feature called The Future in a Minute.

1:55At the end of my discussion with Chuck, I'll ask him a few rapid questions. He'll give us some short answers. Before we get started, a reminder to rate and review. That's a way to help shape future episodes. Tell us your thoughts.

2:14When we think about entrepreneurs, we often think about that single hero founder. They have a great idea, they build a big company, and it's great. And of course, they are heroes. But they require a context for the success of their companies, an institutional support through regulations that are beneficial, and through a culture that supports the creation and the growth of their company. Without that context, lots of entrepreneurs with good ideas will fail. Well, Chuck Easley is a professor of management science and engineering at Stanford University, and an expert on the conditions under which entrepreneurship can thrive.

2:51He'll tell us what that context is and how he's studying it so it can be exported around the world to help create good businesses everywhere and help the economy of all nations. Chuck, why have you decided to focus your research on the conditions that favor entrepreneurs and entrepreneurship? Sure, it's because entrepreneurs are really the drivers of the modern economy, especially high tech, high growth entrepreneurship. And so that's why I've chosen this focus. I also think it's one of the most fascinating, interesting things you can possibly study. But of course, I'm biased. Great. So I know that kind of a high level description of your work is, and that you even have written about is the institutional environments and how they shape entrepreneurial outcomes.

3:42So that's a great reason for me to set some kind of basics here, which is the first thing is, can you define for us, for people who don't think about entrepreneurship, how would you like us to think about it? What is it? Sure. I guess there's a broad definition and there's a more narrow definition that I tend to study. The broad definition is the pursuit of opportunity without regard to resources currently controlled. So I love that broad definition because it can apply to so many things. You can be entrepreneurial in a large company. You can be entrepreneurial in academia and government. And of course, you can be entrepreneurial through starting companies.

4:20So that's the more narrow definition that I tend to study in my research is someone forming a company, often a high tech company, but not always. I also study entrepreneurship in emerging economy contexts. Great. And then the other term I want to define is institutional environments. Give us a sense of what you mean when you're looking at these institutions. What kind of institutions do you look at and what are the kinds of things, and this is going to be the bulk of our conversation, what are the kinds of things that these institutions do that either promote or hinder successful entrepreneurship?

4:54Sure. Yeah, I love the question. So you can broadly think about institutions in two buckets. There have been various classifications, but basically it boils down to policies and culture. And so the policies, you can think about various levels of government. There's all kinds of policies that turn out to influence entrepreneurship. So we can talk a lot more about that. On the culture side, people often talk about Silicon Valley as being this unique culture where failure is, if not tolerated, even prized for the learning that comes out of it. So that tolerance for failure, that willingness to take risks in pursuit of opportunity, there's various dimensions on the culture side.

5:36So both of these, I've been looking at how they shape not just the rate of entrepreneurship, but also the types of companies created. Great, great. Okay, so let's get started in thinking about how do you approach this question? These are huge questions. And so let's start out by saying, like, what is a sample investigation that you might do or one that you've done that kind of illustrates how all of these things come together to create interesting new insights about entrepreneurship? Yeah, sure. So I'm usually looking at one specific institution at a time, often a policy change. So let's take a couple of examples that we can talk about.

6:22you know, people often look at the barriers to entry. So that was a lot of the initial work on how institutions affect entrepreneurship. So the number of regulatory steps that it takes to start up a business, for instance, or the financial barriers to entry that might be there in setting up a company initially. So these are very important in determining the rate of entrepreneurship. But what I got more interested in was this kind of form of high tech, high growth entrepreneurship that we pioneered here in Silicon Valley around Stanford and how that was starting to spread around the globe as people started to catch on that this was, again, the game for driving economic growth in the modern economy.

7:07So I got interested in barriers to growth, barriers to failure. So one example of a barrier to growth, I got started collecting one of the first high-tech databases of high-tech startups in China in collaboration with some partners at Tsinghua University, regarded as the MIT or Stanford in China. And so China had some industrial policies that, you know, this was going back to the 90s that really favored both foreign invested firms and state-owned enterprises and that put at a disadvantage on private entrepreneurs. So they made a reform, but this is going back to 1999, kind of studying the early history of the creation of the Chinese startup ecosystems.

7:56They made some reforms to that, which put these different types of firms on more of a level playing field and made it easier to scale and to grow private startups. So it's kind of pro-competitive. Is that a fair way to kind of summarize it? a more level playing field. So, you know, for example, people talk about antitrust these days, a lot of interesting work on that around some of the big tech companies, or the dawn of the internet being able to use the telephone lines for data transmission. So there's various institutions, various ways that, you know, large companies can capture regulatory institutions and put small startups at a disadvantage.

8:41And so if the regulators are careful about this and set up the industrial policies in a way where the private entrepreneurs are not at a disadvantage in growing and scaling their companies, then what we found was, unsurprisingly, you get more high growth startups. You also get more people with more education, from the more elite universities to get involved in entrepreneurship because these folks have a high opportunity cost. They make a good salary at a big company. And so for them to justify taking the risk on starting an innovative new venture, there's got to be that reward at the end of the day where they can scale it up to a significant level and grow these companies larger.

9:26And so we were able to document that this kind of industrial policy change does result in more high quality, more highly innovative startups being created. So I do want to get to the other elements of your work, but this is such a juicy one that I just want to pause because as you know very well, there's a very active debate right now about innovation in AI. Because as you also know very well, there are these big players who have right now seem to have captured a lot of the AI market. They've certainly captured a lot of the hardware, the data centers and the talent. And so there's a question about whether we're in a good place with respect to ambitious, you know, new entrepreneurs who have exciting ideas.

10:12So this seems both for the entry and for the growth to be a very contemporary problem. So is that a fair statement? And what are your thoughts about the current status of the institutional, both regulatory and cultural aspects with respect to AI startups? Yeah, great question. And there is actually some evidence out there that's making people concerned that the rate of entrepreneurship is potentially declining as a result of these kind of factors that you're talking about with access to large data sets, access to compute resources, and so on that large companies have that might be resulting in this kind of not level playing field in the AI arena.

10:55So I think that's why it is quite important that we be thoughtful about AI regulation and that it's not just people at the table that are the big players that are shaping those regulations in order to make sure that we're winning the AI race, not just with the current generation, but also what creative, innovative entrepreneurs are going to do with AI in the future. so that we're getting talent to flow into the startups that are going to create the next generation of AI technologies. So in terms of this very interesting dichotomy you set up about institutions as either regulatory or cultural, do you see one or both of those as the main opportunities for kind of getting us out of this potential fix?

11:41And what are the kinds of things that might be considered? And I know that this is complex and people have to, there's going to be negotiation aspect to it. But what are the kinds of things you might imagine being done based on either historical precedent or just your experience that might help this situation? Yeah. Well, I think the original dot-com boom is a great example of doing some things right and carefully. The FCC at that time was carefully thinking about the fact that the large telecom companies were at an advantageous position in terms of the telephone lines and access to the technologies of the day for data transmissions.

12:27And they put in place thoughtful regulations that made a more level playing field that enabled all those dot-com startups to be able to access those phone lines for data transmission as well. And so we need to be thinking about, what are the comparable things these days for AI and for the current generation of technologies. So some things that come to mind, social network, data portability across platforms. So there's one thread throughout my work that looks at how platforms may be the new regulators. A lot of that early work was looking at government regulation. But these days, we can think of platforms as the new regulators.

13:12Their algorithmic rules are kind of the equivalent of policies. They're creating the markets, ad placement or search engine rank algorithms. These can unintentionally sometimes, if we're not thoughtful about the regulations and how these are set up, can reinforce some of the either socioeconomic inequalities or reinforce the unfair monopoly position. of the large tech companies. When you talk to leaders of these big tech companies, do they understand why it might be in their interest to encourage these smaller players, or are they all about eating their young? It really varies. Going back to the culture thing, there's a different culture, different organizations.

14:07Intel was famously very thoughtful about this, that they wanted to be growing the overall ecosystem because that was going to drive demand for their semiconductors. And so they consciously realized that, you know, they were early setting up a corporate venture capital group and things like that. And so I think some of the companies are thoughtful about it, but it's very tempting when you have a platform, when you start to understand the network dynamics and the data advantages on a platform, It's very tempting to start to make the playing field unlevel and start to do anti-competitive actions.

14:49And so I do think there's some role for executives to be informed about playing the long-term game. But there's also an important role for regulators as well. And I'm struck when I am exposed to some of those leaders that they're really more worried about the competitors. Like, you know, like, you know, like OpenAI is worried about Anthropic or Gemini and Google. And so when you talk about level the playing field, their first thought is not let's help spur innovation in the little guys. It's like, is that going to give an advantage to my big competitors that allows them to kind of dominate? Yeah, exactly.

15:25And that's why you can't just rely on the market forces. So the third area that I know you're looking at and we can talk about, I want to talk about a little bit is that this issue of misinformation and trust and policies that kind of that are kind of a little bit like connected in ways that people might not expect. And you've written about this. So tell me about how companies might be involved in misinformation and not even realize it or not even be in control of it. Yeah, well, so this is a fascinating phenomenon, and it's only being fueled even to the nth degree by generative AI and the advances in AI technologies that are making it dramatically easier to create misinformation.

16:11It's going to be very difficult, if not in the present, in the very near future, to distinguish video, audio, of course, text. Yes, I'd like to assure any listeners or viewers that this is real and that this is really Russ and really Chuck. You'll just have to believe me. We promise. Yeah. So we started to see this coming a few years back. And I've got to give a lot of credit to my collaborators, current PhD student, Wajihah Ahmed. She was coming out of computer science math background at MIT and was interested in misinformation and starting the PhD program here. And I sort of said, you know, in so many words, hey, I study entrepreneurship.

16:55Like, you know, what do I know about misinformation and identifying misinformation? And so she said, well, let me go back and think about it a bit. And, you know, it's this interdisciplinary kind of collaboration that really drove the insight. And so she had the insight initially that, hey, a lot of these misinformation sites are really like little mini entrepreneurs. that many of them don't care about politics. They're putting stuff out on all sides of the political spectrum, but just for the economic incentive, because they can make a buck off of it. And so we started talking about, well, what data could we gather to actually take a look at this?

17:34And so this resulted in a Nature publication that came out around June of 2024. And what we did was we gathered one of the first large-scale databases of what ads are appearing on what kinds of news sites. We matched that up with a great collaborator, NewsGuard, a nonprofit organization created by journalists across the political spectrum. They rate news websites on a set of nine journalistic criteria. We brought in a couple of other collaborators here at Stanford, Eric Brynjolfsson and Ananya Sen. And so we started to analyze, OK, whose ads are appearing on misinformation websites and whose ads are appearing on trustworthy news sites?

18:15And what we found was shocking because like 80 plus percent of companies across industries, including many universities that we hold dear, are actually placing ads on misinformation websites, essentially sending advertising dollars. Wow. To incentivize the creation of this bad content. And we're not talking here about shades of gray on the political spectrum. Like with these nine journalistic criteria, we're able to isolate things that are demonstrably false, negative impact on society. We all know from the pandemic and from misinformation about climate change and so many other things how damaging this can be.

18:53So we took it a couple of steps further. We then wondered, OK, we see how prevalent this is. We see how many how much in advertising dollars are going to these bad actors on the Internet. And we want to know, do customers actually care? Are they going to punish brands potentially? Who appear, who appear, yeah. Ads appear on these sites because, you know, these things get shared across social media. We know that misinformation from previous research is shared faster and deeper across social networks. It spreads more quickly because it's more controversial. It's more salacious. Right, right. Juicy.

19:29It's more juicy. More juicy. Yeah. Yeah. So, so, and when we talk to folks, you know, in addition to the quantitative work that we did, we also did a number of customer interviews. And it's like everyone's pointing fingers at each other. Like that, the advertisers say, oh, the platforms handle this problem for us. And then you talk to the platforms and they say, well, you know, the brands don't really care. And you go back and talk to the brands again, they say, well, the consumers don't really care. And so everyone's, it's such an intermediate, disaggregated supply chain for digital ads. It's just a mess.

20:02So we did a survey experiment, incentive-compatible survey experiment with the consumers, offering them gift cards and then giving them information, accurate information about whether those brands' ads appear on misinformation sites and the role of the platforms. And we found that it's like having a product recall or an environmental disaster. customers are willing to punish brands for this. They don't like brands ads appearing on misinformation websites. They recognize it's harmful. They're willing to give up about a third of the gift card value for$8. Second choice. So the consumers do care.

20:41Then we did a survey experiment with the executives to figure out, do they hold accurate beliefs about whether companies ads are appearing on misinformation sites? Are they doing this intentionally? Right. And are they interested in solutions? Because technologies could be developed and are developed to keep companies' ads off of these bad sites. But what we found was, unfortunately, there was a lot of misinformation about misinformation amongst the executives that are in charge of making these marketing decisions. Only about 20 % thought their company's ads were appearing on misinformation sites when the actual number was about 80%.

21:18Yes. And when we give them accurate information, they do become interested in solutions. So just so I can kind of make sense of this, I can imagine that this is what's happening. Tell me if this is right. I want to get my products and services, my advertisements, in front of as many eyeballs as possible. That's the whole goal. And the platform says, yep, we can do that. We'll get you guaranteed eyeballs. Then there's algorithms that really don't worry about the content of the websites. They're more just saying how many eyeballs on this website. And then as you said, the juicy stuff gets a lot of eyeballs.

21:55So of course, the way to deliver on that promise to the advertiser is to put it on these sites. And so every, as you said, it's disintermediated, but actually everybody is doing exactly what they promised they would do. It's just that when you step back, it wasn't what people were expecting. Yeah, that's right. So let me go a layer deeper if you're interested. So, you know, people want the lowest cost per click. That's understandable. And as you say, they want as many eyeballs as possible. But, you know, AI is increasingly machine learning is involved in these algorithms. If you're if you're a big brand, you're advertising across 10 ,000 sites on the web, you're not going to go and negotiate a contract with each of those.

22:37So AI algorithms are algorithmically placing these ads where they're going to get the highest ROI. But brands do also care about brand safety. If you're a telecom company, you don't want your ads appearing on a site claiming that 5G causes cancer. Various things like that. So brand safety is important to them. And they would be willing to sacrifice some cost per click, many of them, in order to preserve brand safety. And so if the platforms realize that there was this economic incentive, they could develop these tools or regulators, just like we have information transparency on food labels to know what's in your food.

23:18So they could capture that value. So that creates an incentive for the platform to have new services that are like the cleaner version of, you know, you'll pay more per click, but the quality of those clicks will be exceptionally good. Yeah, the wonderful thing is this is a market-based mechanism. So what nobody wants with misinformation, nobody wants the platform owners policing what's misinformation and what's not. And those experiments have not been that successful. Those experiments have not been that successful. Reminding people to double check the accuracy, the demand side solutions have not been that successful.

23:53Nobody wants the government regulating what's misinformation and what's not. So the nice thing about this is an information transparency intervention, just like food labels or miles per gallon on automobiles, could fill those gaps and create a market-based solution that would dramatically reduce. I mean, we're talking about like out of every$2.16 that's sent to legitimate news websites, a dollar is sent to these misinformation platforms. That's the estimate. This is The Future of Everything with Russ Altman. We'll have more with Chuck Easley next.

24:36Welcome back to The Future of Everything. I'm Russ Altman. I'm speaking with Chuck Easley from Stanford University. In the last segment, Chuck told us about entrepreneurship and the critical ingredients for success of entrepreneurs. In this next segment, he's going to tell us specifically about the race in the semiconductor industry and some of the unexpected fallout in terms of the effects of that CHIPS Act on entrepreneurship in the U.S. and entrepreneurship in China. He'll also tell us about taking his ideas globally to places that are not traditionally associated with successful entrepreneurship, but where his methods and his findings are having an effect.

25:13Don't forget, at the end of the episode, I'll ask Chuck our future in a minute questions. It'll be a few questions and he'll give us a few answers. I wanted to start out in this segment asking you about some recent work you're doing on semiconductors. This is very much a race. It's a race between the US and other countries, in particular China. And I'm wondering, and I think of it as a big, settled industry. So the idea that somebody who's an expert on entrepreneurship is looking at this is very interesting. Tell me why entrepreneurships play into this big battle about semiconductors. Yeah. Yeah, well, semiconductors are so important.

25:52I mean, the future of AI, the race to develop all these cutting edge technologies that increasingly have not just commercial value, but also national security implications, rests on the future and the pace of innovation around semiconductors. And so that's why I got so interested in it. I've been doing work on entrepreneurship in China, as well as entrepreneurship in the U.S. And of course, being here in Silicon Valley, the state of the semiconductor industry is of interest. So the Chips Act came along as one of two key policy tools that the U.S. has been exploring to try to onshore more manufacturing of chips so that we can have a secure supply chain.

26:43And so I started to wonder, what impact was this having? As we were talking about the beginning, you know, slanting the playing field towards the big incumbent companies. Yeah. You know, the nascent semiconductor industry was developed by the startups right here around Silicon Valley or around Stanford. And so I saw the CHIPS Act as mainly benefiting these large incumbent players. And perhaps that's necessary. You know, there's different perspectives on this, I acknowledge. And so the big intels of the world, the big massive companies do have an important role to play in manufacturing chips. But this is a moving target.