Most People See the Black Dress. Stephen Schwarzman Sees the White Lint.

Lex Fridman opens with a deliberately hard question: what does Schwarzman believe that many people he respects disagree with him on? The world's transparency, Schwarzman says, has shrunk the list. What remains, and what he put in *What It Takes*, is a bias toward doing something consequential, something quite large and unique if you can. Operate in that space and success has huge impact; the prospect of success lets you recruit people who want to be part of it; and those large opportunities are pretty easily described. Not everybody likes to operate at scale. Some people prefer small things because they are emotionally meaningful. Those, he says, are life choices rather than commercial ones.

Fridman asks what good and bad come with going big, including for happiness. Schwarzman refuses the mandate. You do things that make you happy; everybody is different; some people have the talent to play pro football and others just like throwing the ball around. Going big, for him, is both impact and joy, and it is also, practically, easier to succeed. Catch a huge cyclical opportunity and you can usually find some place inside it that works. Prosecute a really small thing, get it wrong, and you do not have many places to go. Large efforts concentrate human resources, get people excited, and generate the financial resources to pay talent. Taken together, he calls it a virtuous circle.

How do you know an opportunity is the one to go big on? Schwarzman does not dress it up as mystic intuition. It is pattern recognition, which first requires understanding the patterns and the changes that are happening—observational, data, or simply listening to unusual things people have not said before. He has always tried to describe it as seeing a piece of white lint on a black dress. Most people disregard the lint and see only the dress. He always sees the lint, and he is fascinated by how something got someplace it is not supposed to be.

The discrepancy does not even need to be big. If something should not be in a constellation of facts that used to make traditional sense, focusing on that one discordant note is usually a key to something important. Find two of those notes and you usually have a straight line to some place you have not been. When you figure out that things are changing, you have to be able to describe it, because it is not an odd intuition. It is focusing on facts, almost like a scientific discovery. Then you describe it to other people in the real world, and they tend to do absolutely nothing about it.

Humans are comfortable in their own reality. If there is no particular reason at that moment to shake them out of it, they will stay in it even if they are ultimately completely wrong. He has always been stunned that when he explains where they are going, what they are doing, and why, almost everyone just says that is interesting and continues doing what they are doing. You need a huge data set. Before AI made data fashionable, he had been doing this most of his life. He is not a scientist, let alone a computer scientist. You hear someone say something, or you observe something that simply does not make sense, and that is when you really go to work. The rest, he says, is just processing.

Fridman notes that in AI, "pattern recognition" usually means identifying the dress—the pattern that is always there—while Schwarzman is talking about outlier detection, the rare thing models are often bad at. Could that be put into code, or is it still deeply human? Schwarzman will not pretend to domain knowledge. In his own case, most people do not see any of that. He used to assume it was motivational. He also thinks it is hardwiring: what you are programmed to be looking for. What happens every day is not, frankly, that interesting. He does a bunch of that too, because that is normal life. He has always been completely fascinated by the stuff that does not fit, and by a related pattern: determining what people want without them saying it, seeing the missing piece they do not know is missing, and delivering it.

Growing up, nobody in his family raised their voice in anger or otherwise, and that quiet, he has said, taught him to listen. You have to listen really intensely to understand what people are saying as well as what they are intending, because those are not necessarily the same thing. People mostly give themselves away no matter how clever they think they are, particularly if you have the full array of inputs. Look at the face. Look at the eyes, the window on the soul. It is very difficult to conceal what you are thinking. Look at facial expressions and posture. Listen to the voice, which changes when the topic is comfortable or not: speaking faster, amplitude higher.

Most people just give away what is really on their mind. They are not that clever; they are busy spending their time thinking about what they are in the process of saying. Observe that not in a hostile way but in an evocative way, and just let them talk for a while, and they will more or less tell you almost completely what they are thinking, even the stuff they do not want you to know. Once you know that, the rest of the conversation is easy. They have already told you everything you need to know. Getting to a conclusion, or to an area of common interest, becomes an enormous advantage compared with walking in, hearing something, and believing what they are saying. There are so many different levels of communication.

The practical use of that listening, in the book and in the conversation, is figuring out the biggest problem bothering a person or a group, coming up with a solution, and presenting it. There are two kinds of meetings: chance encounters and meetings you know are coming. The second is the easy one. Pretend you are them. What is on their mind in daily life? What are the big problems they are facing? If the person is the president of the United States—this one or any one—the press is already reporting what is on their mind. You know what they look like, what they sound like, what they are focused on. Take the biggest unresolved issue and come up with a few interesting solutions that have not been out there.

In the early 1990s he was invited to something at the White House, a big deal for a person from no place. He had met President Bush once before, because Bush's son had been in Schwarzman's dormitory and they had crossed at parents' day. He thought through two or three ways to approach the issue of the moment. He was separated then and had brought a date. He and the president went over in a corner for about ten minutes. Later the date asked what they had been talking about all that time. There was something going on in the world, he had thought about different ways of approaching it, and the president was interested. Of course he was interested. There did not seem to be an easy outcome. Conversations of that type work, Schwarzman says, once somebody knows you are really thinking about what is good for them and good for the situation. It has nothing to do with you. You are in service. Then they trust you and tell you other things, because they know your motives are basically very pure.

Chance meetings require different skills and no advance planning. Ask what they have been working on, what they are thinking about, whether anything has been particularly difficult. If they trust you, most people will tell you, and then you have to go to work instantly. Almost everything going on is already out there; people in interesting situations are playing in the same ecosystem with different roles. His running example is the owner of a pro football team that loses all the time—New York, he notes, specializes in those. Under a salary cap every team pays the same amount for its players, so chronic failure is almost always the quarterback, the coach, or the general manager who hires talent. Start asking questions about that and they are typically very happy to talk, because they have not solved it and in some cases do not even know that is the problem.

Fridman observes that for smart people it is hard to escape their own ego and their own problems, which is what empathy of this kind requires. Schwarzman claims a competitive advantage: he does not think he is so smart. The truly smartest people Fridman knows say the same thing. How does he stay humble? He has not changed much since his mid-teens. He was raised partly in the city and partly in the suburbs, on what he still calls middle-class values. Why would he change? That is who he is. The accoutrements of the rest of life have to sit on that foundation, because if you start losing who you really are, who are you? He has never had the desire to be somebody else. He does things now that a middle-class kid from Philadelphia would not have done; the life has morphed. Integrity of personality lets him stay in touch with people from that background, look at new rooms through the same lens, and keep plowing ahead without real adjustments.

Fridman brings up effective altruism and asks how Schwarzman sees philanthropy's goal. Schwarzman does not think about philanthropy the way people expect. He looks at solving big issues, addressing big issues, starting new organizations to do it, much like Blackstone grows: not by taking the original thing and making it larger, but by continually seeing new things and building those, marshaling financial and human resources. In the investment business they find something new that looks terrific and do it. In what other people call philanthropy, he looks at other opportunities to help society, starts something new, marshals people and a lot of money, and at the end of that creative process somebody typically asks him to write a check. He does not wake up and say, how can I give large amounts of money away.

He looks at issues that are important for people. Sometimes he does smaller things because they are important to a person: an unfairness he can relate to, money given anonymously. That, he says, is a miniature version of addressing something really big. MIT was the big version. He helped start the new School of Computing because he saw a global race in AI, quantum, and other major technologies, and thought the United States could use more competitive enhancement. He gets to China a lot, and travels more than a regular person, and he can see the need to have control of these technologies so that when they are introduced the country does not create a mess like the internet and social media—unintended consequences hitting freedom of speech and the functioning of liberal democracies.

On AI, the relatively few practitioners who really knew what was going on had enormous differences of view. By accident he knew a bunch of the famous ones, and he could ask one why this was a force for bad and another why it was a force for good. How do you move the technology forward and still keep a grip on the bad side, including workforce disruption that could happen much faster than the Industrial Revolution, so that the really good things—great things, not merely good—are allowed to happen? To him this was one of the great issues facing society, and the number of people aware of it was very small. He accidentally got sucked in, saw it, and thought: this is mega, globally competitive and also about protecting and benefiting society.

The right thing they figured out was to double MIT's computer science faculty and create the first AI-enabled university in the world, a beacon to the academic research community and a way to make the U.S. university race more robust. If MIT raised a lot of money and doubled its faculty, you could bet other universities would do the same. At the end it would be great for knowledge creation, great for the United States, great for the world. He likes to do positive things other people are not acting on, that he sees first for whatever reason. At the end of the situation somebody says, can you write a check to help us, and the answer is sure, because if he does not, the vision will not happen. It is the vision that is compelling. Fridman adds that Stanford and Carnegie Mellon are watching; the seed is meant to ripple.

What is the long-term hope for MIT's college over five, ten, even thirty or fifty years? Predicting knowledge production is hard. MIT has unique aspects globally; one of the big academic surveys had recently ranked it number one in the world. Leave aside whether you are number three somewhere else: in the great sweep of humanity, he says, this is pretty amazing. You have a remarkable aggregation of talent. Where it goes, you have to be a scientist to have the right feel. What matters is critical mass, and he splits it into two buckets. The first is scientific advancement. If the new college can serve as a convening force inside the university, or help coordination and communication among people, that is an absolute good.

The second bucket is AI ethics, in a way equally important. If the science side creates blowback, so that science is a bit crippled because society's reaction to knowledge advancement in this field becomes really hostile, then you have lost the game in terms of scientific progress and innovation. In a perfect world MIT would serve as a global convener. You need the research universities. You need the companies driving AI and quantum work. You need governments who will ultimately regulate certain elements. And you need the media to be knowledgeable and trained, so that an overreaction to one situation does not go viral and shut down avenues that are perfectly fine to be walking or running down. Discordant information, not even necessarily correct, pushed around society can produce a really hostile regulatory environment. Those four drivers have to be integrated. If the new school can be helpful there, that is a real service to science and to MIT.

Fridman asks how society gets a healthy, factual, exciting discourse about science—especially in a fear-filled field like AI—through media, social media, and the internet. Schwarzman thinks that is a little above his pay grade. Trying to control social media to make it do what you want appears to be beyond almost anybody's control. The technology is being used to create what he calls the tyranny of the minorities: two or three people on a street corner, it does not matter what they look like or where they came from, united by one issue, whose job is to enforce their views on the world. In politics, people manufacture truth and throw it all over. Sometimes they are just hired to do that. You think it is one person; it is a front for a point of view. This has become exceptionally disruptive, dangerous, and undercutting of liberal democracies' ability to function, and he does not know how to get a grip on it.

He was really surprised at the College of Computing announcement. Famous scientists, some involved with the invention of the internet, almost every one of them got up and said, I think I made a mistake. As a non-scientist he had never thought he would hear anyone say that. They had thought connecting everyone and moving knowledge instantaneously would be really cool. He does not know that anyone thought about social media coming out of that, or the actual consequences for people's lives. He had just seen a national-news story of a younger person who killed himself after being ridiculed on social media. "This is dead. This is dangerous." He does not have a solution, other than this: going forward, you cannot end up with this type of outcome. Using AI to make this kind of mistake twice is unforgivable. At least in the West, and in parts of China, he finds people quite sympathetic to AI ethics, to what gets introduced when, and to cooperation inside a country, an industry, and globally to keep the technology a force for good.

Since 2007 Schwarzman has had relationships with senior Chinese leadership and an interest in understanding modern China. Fridman, originally from Russia, notes that Americans are told a narrow one-sided story. Schwarzman starts with scale. China is physically huge, with a billion-three people, and the character of the people is not as well understood in the United States. Chinese people, he says, are amazingly energetic. If you are one of a billion three, you have to be focused on how you make your way through a crowd of a billion 2.99999 other people. Another word for that is competitive. They are individually highly energetic, highly focused, always looking for some opportunity for themselves, because they need to: there are an enormous number of people around. They will try to find a way to win for themselves.

The country is complicated because it does not have the same kind of functional laws as the United States and the West. It is controlled through a web of relationships you have with other people, and the relationships those people have with others. If somebody three levels above you, who is in effect protecting you, gets knocked, you become a floating molecule, tethered only one or two layers up, and that layer is about to get affected. Getting people to change is not that easy, because if there are not really functioning laws, there are only relationships. Sign up for a major change and something is changing in your life; there will not necessarily be all the same people on your team. That is a high-risk enterprise. When you suggest doing something differently you are lining up forces. In the West you usually talk to a person and they figure out what is good for them. In that sense, he says, it is easier to make change in the West, the opposite of what people think. Once the Chinese system adjusts to something new, everybody is on the team. It is hard to change them, but once they are changed they are incredibly focused in a way that is hard for a more individualistic culture.

For a more technical audience he offers one fact. A few weeks earlier, one of the top people in the government had told him that every schoolchild in China is going to be taught computer science. Imagine 100 percent of those children. That does not mean every one of them will be good at it, any more than everyone in the West who takes math or English writes books or poetry. Schwarzman himself evolved to the third grade in math and is still doing flash cards. Pause on the idea that everybody in the society would have some ability to program a computer. Fridman finds it incredibly exciting, and thinks it should give the United States pause. Investing in youth and the education system, he says, is where everything launches.

The American system, Schwarzman replies, is complicated because it has over 3,000 school districts. China does not worry about that as a concept. They make a decision at the very top of the government, and that is what will happen. The United States is handicapped by distributed power in education, even if people in that area think it is great. What percent of American children have computer science exposure? His guess, with no special knowledge, is five percent or less. If the other major economic power is at 100 percent, the U.S. is at five, and computer science is the future, then the country is purposely or accidentally handicapping itself, and the system does not allow a quick adjustment. Issues like this, he says, are fascinating, and if you are lucky enough to go to other countries and learn what they are thinking, it informs what the United States ought to be doing.

Fridman asks how America establishes AI leadership from a federal perspective, given the 2019 executive order and the hoped-for ripple from MIT. Schwarzman thinks you have to get the federal government in the game in a big way. The technological leap is going to happen with or without the United States; it really should be with us, and it is an opportunity for another moonshot-kind of mobilization. He thinks the appetite is actually there. What is getting in the way is poisonous politics. Go below the lack of cooperation that is almost the defining element of American democracy in Congress, talk to individual members, and they get it. They would like to do something.

Another part of the issue is money. The country is running huge deficits, trillion-dollar-plus deficits. How much money does the initiative need, where does it come from, and who is prepared to stand up for it? If it involves taking resources from another area, the political system is not real flexible. If you are creating an initiative you need, trying to get money amid trillion-dollar deficits could in a way be easy—what is the difference of a trillion and a trillion a little more—but it is hard with the mechanisms of Congress. What matters is that this is not an unknown issue. It is viewed as very important. There is almost no one in Congress who, when you sit down and explain what is going on, does not say we have got to do something.

Fridman asks the impossible question. Schwarzman did not endorse Donald Trump, but after the election he gave him advice, which Fridman sees as a good thing no matter who the president is, and he has taken a lot of criticism for it. How do you have a healthy conversation with government about science and technology without it becoming politicized? Schwarzman goes back to a president before there was a moonshot: John F. Kennedy, from Massachusetts, saying in his inaugural not what your country can do for you but what you can do for your country. A generation grew up with that credo. Sometimes you do not need to innovate; you can go back to basic principles. Americans have GDP per capita around $60,000, not equally distributed, but big. People have an obligation to help their country. He does that, and he takes grief from people who project on him things he does not even vaguely believe.

The position, he says, is quite simple. He tried to help President Obama, a good guy from a different party. He tried to help President Bush, also a different party. He does not care that much about the parties. He is a big donor for the Republicans, but what motivates him is the problems the country is facing and whether he can help people get to an outcome that will stand any test. The filters and the hostility now are unbelievable. At Yale in the 1960s his four years held the Vietnam War, Black Power, the sexual revolution with the birth-control pill, and the drug revolution—more stuff in a four-year period than any generation—yet there was not this kind of instant hostility. People who believed something different lived together and respected the other person. That change, he argues, has to come from the leadership of major institutions. Leaders cannot be bullied by people against the classical version of free speech, open expression, and inquiry. That is what universities are for, among other things: Socratic method. If people will not sign on to core principles—people are equal, speech can be heard, no enormous shout-down, subtle or out loud—they do not belong at those institutions. They are violating the core principles. Change takes courageous people willing to lay that out for their institutions and for society as a whole. He believes it will happen. It needs the commitment of senior people.

Fridman is taking a first step into building a startup and asks for advice for the thousands, maybe millions, of first-time entrepreneurs. Schwarzman says it is a rough ride. You have to be psychologically prepared for things going wrong with frequency, and prepared to be put in situations where you are asked to solve problems you did not even know existed. Renting space is not really a problem unless you have never done it: you have no idea what a lease looks like, you do not even know the relevant rent in a market. Everything is new. Everything has to be learned. An entrepreneur starting, unfortunately, does not know much of anything. It is important not to be alone, because it is overwhelming, and you need somebody to talk to other than a spouse or a loved one, because even they get bored with your problems.

Get a group. Jack Ma told him Alibaba was at a financial death's door at least twice. People think it was just Jack, because he became the public face and the driver, but a group of people who can give advice, share situations, talk it through—that really matters, and not just for small details like renting space. It is also the psychological burden. Most entrepreneurs at some point question what they are doing because it is not going so well, or they are screwing it up and do not know how to unscrew it, because everyone is learning. It is hard to learn with 25 variables going on; miss four big ones and you can really make a mess. Either an outsider who is really smart, or other people working with you daily, is the difference. Most people who have not had experience believe in the myth of the one great person who creates the outcomes. Look back over a lot of the big successful tech companies: almost every one of them had two people. Google was that. Early Microsoft was that. Apple was that. People have different skills and need to play off other people. Do not head off as a lone wolf and find you cannot invent all the solutions, or you make bad decisions on certain types of things. This is a team sport. "Entrepreneur" means you are alone, and that is the myth. It is mostly a myth.

Where, in the bigger picture, does he put love, family, friendship? Ultimately all journeys are alone. It is great to have support. If your job is to make something work, that is number one, and you are going to work at it with superhuman effort. People do not become successful as part-time workers. If you are prepared to make that 100 to 120 percent effort, you are going to need support, and people in your life who understand that this is really part of your life. Sometimes they do not, and that is a source of conflict. Involve the right people—dating or spouse—in your life, but do not burden them with every minor triumph or mistake. They actually get bored with it after a while. You have to set up different ecosystems: home life, love life, children as the enduring part, and on the other side the unpredictable nature of this kind of work.

What he tells younger people at the firm, usually around a first child, is to go away with their spouse at least once every two months, to some lovely place with no children and no issues—once a month if they are energetic and clever. Escape the craziness. Reaffirm your values as a couple. You have to have fun. If you do not have fun with the person you are with and all you are doing is dealing with issues, that gets pretty old. You cannot protect the fun element by hanging around the house dealing with more problems. You have to get away and reinvigorate the relationship. Whenever he tells younger people that, the scales fall off their eyes: they had not thought about it; they were so enmeshed. An entrepreneur has to do the same thing. You cannot let relationships slip because you are half overwhelmed. Fridman closes by reading Schwarzman's own lines from *What It Takes*: it is as hard to start and run a small business as a big one; you suffer the same financial and psychological toll; so if you are going to dedicate your life to a business, which is the only way it will ever work, you should choose one with the potential to be huge.