BlackRock sent its team a memo secretly written by ChatGPT—and one major critique emerged

Goldstein discussed his team’s efforts, along with one major learning about ChatGPT, in an interview at Fortune‘s Future of Finance conference in New York City on Thursday.

“(Generative AI) is an incredible force with regard to how you need to think about operating a company, the talent strategy, and just the overall strategy…it’s a whole new world,” Goldstein told Lee Clifford, executive editor of Fortune.

Below is a transcript of their interview, which has been lightly edited for clarity.

Lee Clifford: Good morning, everyone. Rob, you made it here in the nick of time, braved some awful traffic. Let’s dive right in. 

Robert Goldstein: Okay, a little traffic is good, because the city is sort of back. 

Lee Clifford: I like it, absolutely. The city might be too back. Okay, so you guys founded an AI Lab in 2018. You have been at this for a little while. Tell us where is it? What does it do? What have you learned? 

Robert Goldstein: Sure. And I think that when people talk about AI, they forget that the AI Lab at MIT was actually founded in the 1950s. So, a lot of these ideas, these concepts, these methodologies are actually not brand new. What’s brand new, for the past few years, has been this confluence of this more data than anyone could have imagined even a few years ago. And the amount of compute capacity in the world and how you access it has just fundamentally changed. Couple of years ago, even less, when ChatGPT started, what I would call a Gen AI revolution, the world changed again. So, we had started our AI Lab in 2018. We’ve been using these methodologies really for a couple of decades. And what the purpose of the AI Lab, which is literally like right on the border of the campus of Stanford, and we have a lot of people who are affiliated with the math department, the electrical engineering department of Stanford who were effectively running the lab. And the vision at the time was a very simple vision, which was, we have a lot of things we’re solving for, we have a group of people who are solving for them using a variety of methods we’ve been trying for many, many years, and let’s get a group of people who are more focused on natural language processing optimization, more focused on the math than the finance. And let’s throw the same problems at them and see how they would ultimately solve those problems. And we’ve had great, great success with that. What changed a couple of years ago, and I think at the most basic level, you could think of it as everything we were doing really was centered on numbers. Now, what’s happening with Gen AI is that you’re able to extend your problems and solutions to be more about words and language. And that unlocks what I believe will be the most significant technology, innovation, evolution, revolution of my 30-year career, because the kinds of things you could do have just expanded greatly. And I’ll give, can I give one quick example?

Lee Clifford: You may. 

Robert Goldstein: Okay, so the quick example that I would give is we had a presentation to BlackRock’s board several months ago, it was either October or November of last year. And the presentation was on Gen AI and what our strategy associated with Gen AI is. And as part of that presentation, because we normally would write a memo, as part of the board materials, we had the idea of let’s actually use ChatGPT to write the memo. So, we took what our strategy document was, and we fed it into ChatGPT with a very simple prompt. And that prompt was,“write an executive summary,” I think it was 500 words, maybe 1000 words, for the board explaining our strategy, some of the risks and so on. So, it gave us a memo. And then we gave that memo to a bunch of people internally to read. And what was most amazing to me, is and we have some tough graders within BlackRock, let me be very clear. As we gave the memo to those tough graders to read, everyone had comments. And the comments were typically things like, I hate the tone. The comments were like, I think, “you’re selling yourself short.” The comments were like, “what about this?” No one realized it was actually written by a computer. Couple of people when we told them it was written by a computer, they said, I don’t like the computer’s tone. And I’m like, well, you should take that up with the computer. And then what became clear is that even something like the tone, the computer will be able to modify its tone.

Lee Clifford: Be a little nicer.

Robert Goldstein: Or you feed it 10 other memos with the right tone, and you say, give me the tone of those memos. So, it’s just those are the kinds of problems you would have never thought of a computer solving three years ago, writing the memo describing your strategy. So,the unlock here is a generational unlock. And particularly through the lens of a COO, it’s an incredible force with regard to how you need to think about operating a company, the talent strategy, and just the overall strategy of how you engage with and communicate with clients, how you process information, to generate investment ideas. It’s a whole new world.

Lee Clifford: Let’s talk about strategy a little bit. One thing I’m hearing from a lot of companies is suddenly you need to reevaluate your strategy every three months, every six months, it’s no longer something that, you know, gets evaluated every couple of years. What are you finding? 

Robert Goldstein: Well, I think is a starting point. Without question, everything is just happening faster. So, if you take a step back, I think this is another, we would think of technology, innovation, and consequence, maybe an unintended consequence. But everything is happening faster in this world and a big part of it is because one of the fundamental ingredients for change is information flow. And the information flow is just accelerating, accelerating, accelerating. So, I think we view strategy as an incredibly fluid assignment. But what’s interesting from a BlackRock lens is that the strategy has been largely unchanged, because from our perspective, we’re really looking at what are the challenges that clients have with their whole portfolio, with the whole investment portfolio. And the core of our strategy is really fulfilling that whole portfolio requirement. And if anything, many of the forces that are happening, including the forces that we’re describing now, are just accelerating the client requirements, to work with people who could do more and more things for that whole portfolio.

Lee Clifford: When you look back over your 30-year career at BlackRock, what are some of the things that you think are unique about your company and have made it so successful over the past 30 years?

Robert Goldstein: We only have six minutes. I think I could talk for a year about my 30 years. But importantly, I think that if you look at the history of BlackRock, and if you look at what BlackRock is today, and even as we were being introduced, we spoke about BlackRock, we spoke about the Aladdin platform, we spoke about the ETF platform, the iShares platform that we have. And what’s interesting is that if you think about the asset management industry, particularly if you go back 30 years ago, 20 years ago, 10 years ago, you could argue even today, it really is organized much more around the service providers, the asset managers, and then there’s this expectation that the clients have to put all these service providers together. So even if you look at BlackRock and our size and scale, it still is under 10% of the market. The industry is incredibly bifurcated, which leads to the client having to do a lot of work. And the strategy that we’ve been employing, at the most basic level, is that the clients are our compass, for where we take the company. So, the clients are building these whole portfolios. Those whole portfolios involve taking investment strategies that seek Alpha, investment strategies that are designed to provide exposures through replicating indices, public markets, private markets across a whole array of things. Then you need to understand the risks in those portfolios, you need to understand asset allocation, you need technology to do the processing. So, our strategy has always been how do we be able to solve the client’s whole problem or any building block piece or the technology infrastructure? Now, if you look back when we started to provide Aladdin, people said, why would an asset manager provide technology? When we bought Barclays global investors and we, as part of that, had the iShares leading ETF franchise, people basically said, you need to be an active manager, or you need to be an index manager. They would call it back then a passive manager there could be nothing further from the truth. You can’t be both. It was religious. And we just said, why is it religious if clients are taking all these pieces and putting them together? Isn’t the industry, an industry designed to serve its clients? And I think if you look at the past three decades, the industry is just shifting more and more, and I like to think BlackRock is leading in that shift, to how can you really service the whole portfolio? And how could you be able to talk to clients, where even if they buy one Lego piece, one building block from you as an individual product, you recognize it’s in the context of that whole portfolio. And you can talk to them about that Lego piece in the context of the whole portfolio. 

Lee Clifford: Okay, well, we’re here at the Future of Finance Conference and now it’s time for some bold predictions. We’re sitting here 10 years from now, what is going to look different about the whole finance industry?

Robert Goldstein: I think that BlackRock was founded. And from the earliest days, there was a thesis that at its core, the asset management industry, is an information processing industry. And I think the past three decades have been if you look at it as a line, it’s been going like this towards that vision. I think the line is going to go like this. I think that the asset management industry, and the broad business of investments in aggregate, is going to just increasingly become an information processing a technology business, there’s going to be mass personalization, that clients are going to be able to experience there’s going to be a mass ability to look at all the data that’s being created. And remember, if you look forward as a prediction, I’m not smart enough to know if it’s 5x, 10x, 100x, 500x. I just know it’s lots of x more data in the world that asset managers will need to understand and process, and these businesses are just going to increasingly become technology businesses in every way possible. I think it’s also quite clear that the lines between the public markets and the private markets, the lines between where you are within a capital structure and a capital stack, all of these lines are getting more blurry. So, the portfolio of the future will clearly have public market exposures and private market exposures, it will clearly have fixed income and equity. But it will clearly be thought of more as a series of exposures than the convenience technology of being able to label things a certain way. And assume, as a simplifying assumption, assume that that label behaves a given way.

Lee Clifford: Last question, you’ve said tech a lot. I haven’t counted, but it’s many time. Is it changing who you’re hiring? What type of background are you looking for now? 

Robert Goldstein: That’s a great question. We within Blackrock roughly a little over a quarter of the company would be people who are technologists, that is something that we’ve been doing for quite some time. But it’s interesting, because I feel like we have more and more conviction that we need more technologists. We also have more and more conviction, we’re a firm that has a very large analyst class every year that we hire out of university. We also have more and more conviction that we need people who majored in History and English in things that have nothing to do with finance or technology. And it’s that diversity of thinking and diversity of people and diversity of looking at different ways to solve problems that really fuels innovation.

Lee Clifford: As a history major, we have to stop there but as a history major, I love to hear that. Thank you so much for joining us.