Goldman Sachs Chairman on Why Finance Adopts AI Differently | a16z

Anybody who's investing, you know, you're doing two things. You're trying to make money for yourselves and your clients, and so you're trying to get out there and take risk, and you're also trying to be a risk manager, and you have to do both.

Once the present turns into the past, everybody's a genius. Most of what we do with respect to risk is not so much predicting, it's a lot of contingency planning.

Not because it's smarter than us and it's going to turn us into pets, but because we don't have the ability to test whether it's right or not.

I know. You know what, it's a funny thing. You would think that you see something and you're activated to tweet about it. Me, I said, 'Oh, gee, I haven't tweeted for a long time. Let me find something to tweet about.' That's more it. And also being in the risk management business, I always know that everybody keeps doing that and eventually you get cancelled because you do something, you step over some invisible line that nobody knew about. And so I realized that from a risk-reward point of view, it's all ego and no real value other than that. But that was saying, you know, when you retired, your grasp its draws.

I remember when I was doing it, what's his name from... You know, I got this, I said to you, when I retired, no unrestrained... I am freed from the restraints that I had because I started, you know, I did this at Goldman. And I realized that I was playing a dangerous game because I was being snarky with the president and I had all those back and forths with Sanders and Elizabeth Warren.

No, that was... Yeah, that was real. But it wasn't, you know, it wasn't like I was hungry. So every, you know, I always used in moments of crisis like that, I always tried to be disarming. And you know, everybody was... and by the way, it was very sensible to duck down under the desk. I mean, it was a line, you know, realized we were pretty close up and I was just... it wasn't that thoughtful on my part. It was just that I was like... it was like being in a movie and I was like enjoying watching it. And you had all these guys who were wearing tuxedos. Suddenly they had little, you know, they had pistols in their hand and there were guys in full tactical gear and they all ran in and they all were on the stage with their guns facing outward of course because that's where the threat would have come from. Then suddenly, you know, a guy tugs at my leg and he said, 'You really should get down.' And I said, 'You're really right.' I said, 'This is like when I get into an airplane, this is another time that I'm glad I'm short.' But I was watching it and then I saw what everybody was doing and I didn't see a lot of panic. I didn't see any panic really, the people under this, which was a sensible thing to do. But again to break the moment I looked down and I said, 'By the way, are you going to finish your salad?' And as I said, it was kind of funny at the time.

Yes, I was. You know, somebody said, 'Gleman, you're very good in a crisis and that's why you go out of your way to create them. So, just so you can give you an opportunity to be good in a crisis.' And I would say that my normal resting state is to not be resting. So, I tend to be a little bit wound all the time, but I don't get especially wound. In fact, things slow down for me. I'm used to seeing things like that. They're in slow motion, and I become very sensitive to what the people around me are thinking and trying to, you know, get them... most of the time like at Goldman and in most of life in a crisis time, the really important thing is just to get people to do their jobs and to stop being, you know, don't be frozen and don't submit to the chaos.

I don't know. I wouldn't have predicted that about myself, but I've now gone through, you know, we had the crisis of the century roughly every four or five years. And it's always that way. But, by the way, it doesn't mean I like crises and I wouldn't go out of my way to volunteer to be in one. It's just that when it happens, I generally have confidence that I'm not going to get discombobulated. If I'm going to get discombobulated, everyone is going to get discombobulated before me. That's how I... and so I've done that. And by the way, that taught me a lot about the people that you need to rely on because you can't really tell... you know, I mean, not to coin a phrase, but you can't tell a book by its cover. And I went through the financial crisis and we had people, and thinking one in particular who was a great athlete, terrific guy, real man's man, you know, did rodeos on the weekend and he was terrible. And then here I am the co-president of the firm, you know, here I am trying to teach people how to... and trying to say, you know, you have to breathe. And then there were people who didn't look like they could walk up a whole flight of stairs and they were really good. And so, you know, just people, you just don't know. And that's why my advice when you pick board members... and by this is a very... I'm turning something that's generic into a very narrow thing. I think a good place to go is find people who've already gone through a crisis because to me, people who look like and sound like they'll get through it, it's not really... I'm not sure how much of a correlation there is to the reality of it, but when somebody's gone through a crisis, I think that's your best bet.

In South San Diego in Chula Vista, like 10 minutes from Mexico. Mom was a public school teacher. Dad worked in retail in Mexico. Very far from, you know, Cambridge. And Harvard really changed my life, right?

He had a motorcycle, so it was a little bit easier.

Exactly. But Harvard definitely changed my perspective on what's possible. I always say I learned more from my peers than I did from my classes. I'm just curious if you had a similar experience.

Amazing. To get out of Brooklyn.

Right. Yeah, that started, that went in a different direction after I left. That was an outgrowth of the merchant bank, totally nurturing a business and then somebody said, 'Gee, we shouldn't be just a private equity firm here, we should be our own strategic...' and that's how it was done.

Well, I was an acquiree so I don't know what they thought at the time. I subsequently found out what they felt about it. It was a disaster and it was a little bit like Columbus sailing, trying to find the Indies and instead finds America. It turned out okay but for different reasons. They discovered something but not what they intended to discover. So they ended up getting a bit of an entrepreneurial culture that they didn't know they were buying. But certainly at the time, this was in the early 80s. It was a moment of great... in high inflation. That inflation, the manifestation was higher commodity prices, precious metals. Gold had only been recently freed up to be owned by individuals who wrote, 'Hey, gold had been... we've been on a gold standard that evolved. It's hard to transport back to that time.' But the business of J. Aron & Company was kind of a sleepy business except it erupted in a positive way at the end of... before Volcker came in and clamped down on inflation, highly inflationary period. And of course the savvy Wall Street guys at J. Aron extrapolated the value of the firm at the peaky part of its thing and sold itself to Goldman. At the same time, DLJ, which was an investment bank at that time, bought Ackley and Solomon Brothers and Phibro got together. So, it was in the air that the Wall Street firms needed a commodity arm and Goldman Sachs got J. Aron. Now, J. Aron had a different culture. It was, to the extent that this is kind of all lost now because all these firms have kind of blended and you wouldn't know the difference. And at the time Goldman was kind of an 'our crowd' kind of a firm. It was a Jewishy kind of firm. So was J. Aron but very different. Goldman was kind of the upper echelon crowd and J. Aron was more of a kind of a street guys. Goldman recruited from the Ivy League and people with MBAs and J. Aron just recruited people and the first, the entry-level job for most of the life of J. Aron was the best job to get was the driver for one of the traders. And literally, and it was kind of almost like mafia-like in a way and that's how you rose in the organization. And I had gone through college, went to law school, took myself and my loans into a law firm and worked there for about four or five years. And like a lot of other people at that time, I wasn't doing... I was doing well at the law firm, but it wasn't necessarily for me in the long term. And I looked for jobs, I knew nothing about. I interviewed at a lot, being in New York, what do you go into when you're done with... you become a consultant or go to Wall Street. I said, 'I'll go to Wall Street, I will bestow myself on them. They should be so grateful to have me.' I knew nothing about it. And of course I got a job nowhere. And the only place I got a job, including, by the way, Goldman, where I didn't get a job. And the only place that offered me a job was J. Aron and Company, the small commodity trading firm that I had never heard of. And they hired me as a precious metals salesperson. And right around that time, they were acquired by Goldman, which is how I got into Goldman.

It hasn't shifted, you know, the vehicles have changed but the kind of judgments and the perspective that you believe... look, we were at Goldman and anybody who's doing this business and yourselves, anybody who's investing, you know, you're doing two things. You're trying to make money for yourselves and for your investors and your clients. And so, you're trying to get out there and take risk. And you're also trying to be a risk manager, which is, you know, it's almost like you bifurcate yourself and say, are we... I know we want to take risk, but let's go into risk management mode and let's consider, are we diversified enough? Are we overly committed to this? Are we managing it well? And that's kind of a different head that you have to bring. So, and you have to do both. And by the way, we get challenged on both sides. Sometimes things go badly and people, the pleasure-pain principle works and people don't want to take risk. But yes, we're paid to take risk. So, you have to take risk. So, what do you want to do? And you have to exhort people and sometimes shame people into taking more risk. And sometimes you have to get them, okay, we're not talking about what risk we want to take. Let's go over our portfolio. I'm sure you do portfolio risk. And saying where are we overly exposed? What contingency plans would we have if X, Y, or Z or W or G happens? What can we do today to mitigate the adverse consequences if any of those things happen? And when you go around the table for those meetings, you're not so much interested in what people think about the future, where things will go. You just want to know, forget about what you think the likelihood and probability of something happening is. What will you do if it does happen? And what can you do today to mitigate the consequences of that in advance at a very low cost today? You know, buying insurance is very expensive when everybody needs it. When everybody needs it and when the problem is dramatic, when the hurricane is coming and it's on its way, it's very expensive to buy insurance for your oceanfront property. But, you know, in the middle of winter when it's the furthest thing from your mind, it's a lot cheaper. And so, what can you do? And so we did both those roads. I'd say what I might have had an orientation towards was the risk management part because I could find the cloud around any silver lining. You know, my wife yells at me, she'll buy something new and I'll notice, 'Isn't that a chip on the lower part of something like that.' I think my wiring was always to be a little bit fatalistic, a little bit nervous, and a little bit looking for stuff that could go wrong. And but it turns out that I had a kind of an appetite for risk. Not... yeah. Now, that's a little bit different than saying I was good at risk-taking, but I could live in a risky situation. I didn't shrivel up. So, you know, I ended up having to do both things. But we have a lot of risk-takers. And I would say that the biggest challenge for management is the risk management side which is really getting people to refrain from risk which is about a third of the time when you're in that business. And by the way, not the most important part but probably the bulk of the time is getting people to take more risk when they don't want to.

Yeah. Because you get singed, right? And you don't want to do it, but we're paid to put money in the right place. And so, you just can't be afraid.

Well, that I'm honored by that. But yes, I always think of myself more as a tour mentor than a mentor. He said some amazing things which I want to come back to. But he said one of the maybe cultural thumbprints from his perspective that maybe J. Aron left on the firm was a culture mark-to-market. One of the other things he said was you were a manager that understood losses. So you weren't afraid of them. You would often, to your point, kind of encourage people to lean in. The other thing was he said you were incredibly good at gathering information from the organization. You were both very approachable so people wanted to come to you and then often when you were doing maybe an audit of a division you wouldn't just speak to the head of the division you'd speak to like the number two. I did but I don't want to undermine but I always did... on that score I tried to make it so everybody felt comfortable talking to me. One thing I never did, if somebody was calling to tell me something that was bothering them, that they saw an opportunity or a challenge, I never said, 'I already know about it.' Because I never wanted anybody to self-censor later and say, 'Oh, he must have heard about it from someone else.' I didn't... I want... if a junior guy was telling me something and three people up the letterhead told me the same thing, I would sit and listen. First of all, you find out a lot about the person who's telling it to you also. So, you're not just learning the content of what he's saying. You're learning a lot about the messenger. But secondly, I didn't want anybody to have an excuse to not tell me stuff. So, I listened to a lot of redundant facts and circumstances. So I thought about that a lot. And about taking losses, you learn that the first day. Sure. I mean, of course, everybody... and I'll tell you one other thing is very important on the loss side. People can lose money. You could lose money because somebody's stupid or you could lose money because somebody's wrong. Smart people are wrong. Smart people tend not to do stupid things, but they tend to be wrong. You know, the old saw about the best hitters in baseball make out two-thirds of the time and that kind of stuff. But it's very important not... when something goes wrong, it's just, you know, when something is not right or somebody loses, it's very important not to treat somebody who's wrong like they're stupid. And people make a mistake because the big fault of risk management or bosses or managers is they let after-acquired information seep into their judgment of what they would have done at the time. And that's... you have to be very careful about that when you evaluate people and you engage with people. You have to show an appreciation of what people have done in the fog which always exists. Because none of us know the... by the way, most of us don't even know the present. The present is a mess of things. How who can sort that out? But once the present turns into the past, everybody's a genius. Nobody voted for Nixon and everybody got... and yet he won in a landslide. It's like everybody remembers things differently.

Yeah. So, you know, again, when pundits come up and said, well, I know this or that. So look, when anybody tells me, 'Oh, I knew this.' And I said, 'Well, if you were so prescient, tell me what happens next.' I always... 'Oh, well, it was easy then.' And by the way, when somebody's telling me about the certain future, I say, you know something, did you know that we would be doing this today or that the AI... if you didn't know those things, how why are you so sure that you know the future? People don't know this stuff. I'd say most of what we do with respect to risk is not so much predicting and not so much forecasting. It's a lot of contingency planning. And if you're a good... and you go around the table, what could happen? Don't tell me about the probabilities or the improbability, what could happen? And again, we said this before, what are you going to do about it? But the act of going through that thing makes you so alert and on it. When things get triggered and you have a plan, you get off the mark so quickly that people think you did anticipate it. But what you really did is you heard the gun go off before anybody else. You know, I don't know why I use sports analogy. I'm not the best sportsman in the world, but I know that in track and field, if they shoot the gun off, but you leave within a tenth of a second after it, they call a false start because your reaction time is at least a tenth. So you're not allowed to anticipate a start. And they'll call it for... I said I want everybody to be called for a false start because they hear the gun so much quicker than anybody else they get off the mark. And so that's the exercise you can do. Now some people are more intuitive, some people see things, but what I really think for most people is that they thought about that, what if X and Y and Z could happen, if this happens this is what I'm going to do.

Oh my god, we were always... the technology was always changing everything. And by the way, in a lot of what... not everything, but a lot of things in finance it's winner-take-all. You know, if you had a risk, you know, an execution system that communicated digitally back to the floor of the exchange, you wanted your computers a half a block closer to the exchange than everybody else's because the milliseconds mattered. Not only mattered, it was win-or-take-all. You got everything. You got the offer or you hit the bid and other people were left looking at your dust for that. So, you were always competing for the best technology in a winner-take-all situation. By the way, a lot of life whether people realize it or not is win-or-take-all. I can see, you know, I know the opportunity set and the challenges and the anxiety people have about the current thing and I think obviously I still invest and I still transact in the market. I think about that too but I would say that no one is a better adopter or pays more. Except for obviously the hyperscalers themselves who want to be the providers of the technology but in terms of use of the technologies, the financial area wants to be on top of it. And it's so interesting, by the way, there's a lot of... you end up in a lot of cul-de-sacs, you'll end up going down bad paths because you just don't know. And you have to do this and I know that everybody's talking about looking for cost savings but we always had to do things twice. We had to use the system we were confident in and then simultaneously run the new system we had high hopes for. We didn't have a high level of confidence in our business and as a regulated company that we were. We weren't allowed to have mistakes. By the way, that's another schism between the Valley and finance. So you could be... I looked at Robinhood, great company but early on they declared that they had government-insured accounts that weren't government-insured. They had some slip-ups and a lot of apologies get made. You could do that. We weren't allowed to do that. We had to be right. We had to run things 50 times and had to be perfect the last 49 times before we could go that way. So we would always have the plan that we would... the technology that we knew worked, inefficient as it was compared to the new, and simultaneously run. And when we got confidence in the new system, we implemented that and then there was a newer system that we were also beta testing at the same time. So technology in the first instance always augmented our costs, never detracted from it, but as we went from one lily pad to another, things got better and more efficient. But we were always testing new stuff and always geared towards it and always very anxious about what would happen if somebody trumped us on something. By the way, in addition to execution capabilities and things that go to the efficiency, we also... our risk systems. By the way, we had a huge technological advantage because of what we invested in early on.

Yeah. I don't know if I deserve... I accept... whether or not, you know, except for blame, I accept anything that comes my way. But we did have very good early stage risk models. By the way, SecDB, which was a kind of a risk management system that we had and it's kind of modular whereas other things were kind of rigid, you know, we can always change things, was so good and so flexible that I think the system must be between 25 and 30 years old and the core of it is still implemented. The only thing I know like that and I once tweeted this out because the battery... I still have my HP-12C calculator. And the battery went out and I know that battery must have been there for 22 years. I think I owned that device for like 40 years. And I looked at that and I said, you know, I never thought of this before, but what consumer device is still tracking after 40 years, not only are people still using, but looks like it could have been designed last year. It's an amazing thing. Well, our SecDB was kind of like that. It was also... but it wasn't a consumer device, but it was good like this. So, I have a lot of admiration for design that really, you don't expect design to stand the test of time. So, fashion doesn't. But this does. I'm telling you, the original iPhone looks like an old product to me.

You know, one of the things I think was unique about your career as well is you spent half of your time at the firm kind of pre-IPO in a partnership and half the time post-IPO. Very relevant to your entrepreneurs.

Yeah. I mean maybe you could describe kind of what were those principles pre-IPO and people really credit you also for kind of carrying that culture forward. Ashoke said this as well which is like we don't have a partnership but it still feels... it's a partnership. So let me just say the differences people aren't alert and... in a partnership... now we're a big firm, you're dealing with small firms who want to become big firms and some of them have become big firms. But there's a really big difference between a partnership culture and a corporate culture by sometimes by necessity. And it was really to go public and I'll tell you we can go into that direction but we had to go public. But one of the big impediments to going public was the fear that we'd lose our partnership culture. Now what do I mean by partnership culture? Partners own the firm. The employees there, especially the senior people, are your co-owners of the partnership. To the extent that you're a senior partner, a lot of it is by consent of the governed. When you're looking at your senior card, they don't just work for you. They're not just subordinates. They're your co-owners of the business. They have certain expectations that come from that. For example, their fortunes rest on the success of the whole enterprise, not just their narrow silo. If you work for Amazon in the retail area, are you really raising your hand asking questions about AWS? But if you owned it, you care about the whole. So one thing, they own the whole. They care about the whole. They expect as owners to have a lot of information about the whole. They expect to have influence about the whole. They expect that any sudden moves by the senior partner is socialized with them. They expect to have input into that. They expect the process to be slow enough for them to have that influence and input. And you have to have a certain amount of discipline when you're managing that if you want to perpetuate that. And I'll get to why you want to do that. And so you have to socialize things and maybe your decision-making... lightning bolts don't come from your fingertips. You're trying to make suggestions. Maybe you slow things up and you hear complaints and maybe you actually don't do things that you want to do or you table it for another time when things could be more revealed.

Well first of all, generally when you're on top people want to get in line with you and sometimes they can't, they just think you're wrong. So socializing and talking in advance had the benefit of just enlisting support from people who otherwise might be neutral who just... not because they're sucking up but just naturally they want to... they're pliable. And then you had to honor the fact that they felt like owners. Now, why do you care when they feel like owners? Because you get a much more stable organization. They feel attached. They feel committed. Even people who've been there for a few years take that away with them. And people who've been out of the firm for a long time still self-identify as ex-Goldman. By the way, how we treat... one of the examples of part of that kind of ownership. We treat our alumni very specially. Goldman has an alumni office. I put in an alumni office where I spoke to Alison Mass. Alison Mass is a partner and she runs our alumni office where we do things with people who have been out of the firm for 20 years.

Oh, no. He goes on to like... yeah, he hasn't been at Goldman for 35 years or something like that. Where does that come from? Again, it's how you know... it's a lot of times it's crazy to expect a kind of loyalty if you don't show loyalty. It's crazy to expect commitment if you don't show commitment. I would say leadership, my predecessor did, my successor does. The challenge of Goldman Sachs, we had to go... I mean, I can get into this. We needed to go public, grow the balance sheet. And when they repealed Glass-Steagall, once upon a time, the lenders were separate from the investment banks and the investors. That got repealed. And all of a sudden, people who gave advice could now implement the advice by financing it. So, we had to... if JP Morgan was going to become an adviser, we had to become a good lender. And a good financier. So, it meant that we had to have a bigger balance sheet. Couldn't run that on impermanent capital of a partnership. And so we had to go public. But the big anxiety was we'd lose the partnership culture. We went public basically in an instant legally, but it's taken 25 years to get it done in a way that it wouldn't undermine the partnership culture. So we do those things that make it partner-like. We have partnership elections. We pay people based upon how the whole firm does. If your area does particularly well, you'll know it in your compensation. The most important thing in compensation is how does the whole firm do? And so you get people who are bankers sourcing investment things for the merchant bank. You have investment bankers who would like us to represent their client on an auction. And there are three other investment bankers who represent three different potential buyers. And you have to pick one. And we sort it out together collectively what's the right place for Goldman Sachs to be or maybe we should represent the seller or maybe we should be a buyer ourselves. How do you decide that and you explain it and you do it, you let everybody have their say, now what should we do here and you convince people that if they throw in with the enterprise as a whole and sacrifice in the short term they get to use the platform and exploit it for their professional career and their personal career. So you got to get the... it's like I use as a metaphor, you know the metaphor of the 800-pound gorilla in the jungle gets his way. I'm the 800... but what if you have 20 800-pound gorillas? 19 have to say, 'Excuse me, after you.' And how do you get them to do that? And that's a bit of the art and we did that, the firm did that. By the way, there were other things that we had to do in terms of reforming to make a public company. In a private company, your company... sure, you care, your partners presumably, everybody cares about making money for their investors and their clients, but as far as you're concerned, you don't care whether you make money smoothly. In 5% higher increments every year. You can have three in a 10-year cycle, you can have three fantastic years, make no money for five years, and lose money two years. And it could work out well. In a private company, you care about the E, the earnings. In a public company, you care about PE. And if you have volatile earnings, your shareholders don't like that. They reward you with a lower multiple or they punish you with a lower multiple.

And so we shifted a lot of that to off-balance sheet vehicles. And by the way, that means you have to do more of it. Because instead of earning hundred-cent dollars, you're earning twenty-cent dollars with lower risk. And a higher P/E and a higher return on equity as a result. But that took some time because you didn't want to lose the people who do that.