So we'll start with Shakar. Shakar, professor of business school, seeing a lot of different companies here. What are you seeing out there in general? I'd like to talk through that, and then we'll come to a publicly traded company.
You know, John really summarized a lot of what we're seeing out there, which is companies intrigued by the problem and then trying to find a way to work out what is their strategy relative to this. We've seen this before, we've seen it when the internet first came along, we saw it when mobile came along. There are certain patterns to it that we get a sense of. I think one question I'd like to ask you is: imagine that you are walking on a beach, and as you walk along you see this iPad-like thing lying there, and it's broken. But as you get to it, it gives you a little message and it says 'I'm GPT 10.5, and unfortunately I fell down but my power supply is gone. GPT 1 is going to have the ability to keep powering itself up. If you pick me up and connect me to your USB, get me the power, I will give you the power with your company to change customer service to 5% of what it costs now, to change marketing, to change your strategy, to change engineering, to dramatically change who you are as a company and how you compete.' The question to you is: would you hook it up? Would you pick it up and would you connect it? Think about that for a minute. I did this as a little poll in a conference where we had a little bit more time, and approximately 40% of the people said they wouldn't. I said why, and they couldn't give me a good reason. I asked them how many of you who would do it would take it to your board and say 'This is the most amazing thing ever, we need to get into it, we need to do something with it,' and most of them wouldn't. There was this fear. The biggest thing that I'm seeing in terms of the restriction of people and the ability for people to embrace this is the same problem we have in almost any new technology, and that is primate behavior. We're all basically primates where fear drives what we do. We're concerned about uncertainty. In that process, AI has all the potential to change things dramatically, but it also has the potential to have some spectacular failures. So what do you do about that? How do you move forward with it? I've thought about it as: on one axis you can say what is the AI actually doing. You can say it's assisting me. It could be summarization of notes, meetings that you work through, code co-pilots. That's often put in the co-pilot kind of category. Then there is the notion of augmenting, which is it is doing things that you could never do with any reasonable economics, like language translation, personalization of every message, avatars. These are all technologies that exist today at industrial level abilities. But the final one is transformation, and what it can do is change the business model dramatically. We had Sam Altman in class a couple of years ago before he got careful about what he said, and he was pretty open. He said what they see happening, the problem they'd like to deal with, is when one company owns the universe, because if you work through the recursive loops and just work through the math of it, there is very little, because everything we do is cognitive. While that's an extreme statement, he meant it as an extreme statement, it gives you a sense of the direction in which things are going. I'm going to end with three thoughts about where this is all going. One: most of the time when you have these big technology changes, the first effects are all negative on society, on the company, except for a few people who come out. The Industrial Revolution is often used as an example. What people forget is that when the Industrial Revolution happened, for the first 80 years, three generations at that time, the average worker's wages went down, the average standard of living went down quite dramatically, and the institutional structures all supported capital versus labor. You had this readjustment of society, but you also have the readjustment within companies that occurs. Be careful for this jagged move up, move down thing because people are trying to get their heads around it. The second thing: the way that companies will transform is more likely to be through termites and not tornadoes. Small parts of your value proposition will get eaten away, and then it's harder and harder to compete. At each point, the incremental decision is not going to be worth it or there will be a lot of risk. Most corporate systems don't have a good way to manage risk that you cannot quantify. By the time the tornado comes or the building collapses, it's too late. A large part of this process is getting the mindset. When we ran an internet company, I visited a retailer who said 'Amazon can't compete with us.' I said why, and they said 'Because we have a website too. People can come in, order all our products, we can ship within a day, we do all the functions that Amazon does.' What he didn't realize was that Amazon thinks differently. The trajectory they're moving on is completely different. That company is out of business now. Amazon dominates everything. It's about the way of thinking, about taking learning and moving it through that process. The final thing: so far, most of the discussion of AI has been around process, how to make things better by using AI. The big change is that it's going to be about people. Your mother knows you really well because she knew you from childhood, she knows the way you react to fear, to incentives. AI is getting there. The context windows from a company called Magic AI now are 100 million tokens. Three years ago the largest context window was 3,000. When you have the context window, you know everything about a person and you can predict what they're going to do, what they're going to think. That has been the holy grail of business. We're going to move from words to tasks where agents do everything in the process. These are very radical changes likely to occur over the next 2 to 3 years, not the next 10 years. Then there's going to be a long group of laggards. The way you organize the company becomes critical. Thank you, Shakar. Shakar is a professor, he's also a board member at a publicly traded company, very enthusiastic about this.
Melissa, so tell us who WEX is and how have you guys managed your alignment and discussion on AI?
Sure. WEX is the global B2B fintech company, and our purpose is to simplify the business of running business. We have over 600,000 customers globally. We're behind the scenes, we help people get paid and make payments in a lot of ways that you probably experience indirectly. If you have an HSA account or an FSA account, we're probably the backbone to that. We sit behind a lot of integrated payments happening within travel payments as well as vehicle-related payments. Really happy to be here. Shakar has been an amazing board member. We've talked a lot with companies, John and I, in the last 12 months, and it's so clear how critical getting alignment at the board is. Lots of discussions, lots of consultants running around, lots of ideas.
How did you guys approach that and think about this journey? We've been very impressed with how much you guys accomplished, but it started somewhere.
It did. To put this in context, when I first became CEO 11 years ago, every day I would get an article from Shakar, and the article was about companies failing. About the fifth day of getting these articles, I called him and said 'Is this lack of confidence?' We spent some time talking about it, and really the message was around complacency kills. That message has really sat with me for a long time. I'd like to think of myself as a really thoughtful risk taker, wanting to get ahead of these trends. That played forward in 2022. Shakar and I and the board had been talking quite a bit about AI, but we had a lunch. At that lunch, we sat down and you could see the zeitgeist of AI starting to occur. We were ahead of the curve, we were using AI before we were talking about ChatGPT becoming a thing. We looked at each other and we both believed that this was something that was really going to be impactful. We decided to make a big push and to go all-in. What all-in meant was making this one of our core strategic points, setting up a few things across the company that were really important. I wanted to make sure we had goals around productivity, but really starting with a customer perspective, making sure what we did had customer impact, and that we were going to educate our employees and bring them along this journey. That was the pivot point. We thought about it as being akin to security. You think about our technology committee, that's where we spend a lot of time, on information security and AI, because both of those things are foundational to technology companies and really foundational to the success of the company in the future.
Thanks. And Shakar, what did you see in that process? Was there a critical meeting? Did you have to get consultants? Was it a collective process? Was it an offsite? Was it articles? The whole management team.
I think it starts with board composition. You need some people in the board who understand what's happening and are willing to go sort of far out. Then you need some other people at the other end who are saying 'We need to meet our quarterly numbers, we need ROI, reputational risk is a really big issue.' But you need the fabric by which people can actually talk to each other. What I saw at WEX was this willingness to talk and to be open about what could happen, and the willingness at the management level to bring outside speakers, outside people in who would give different perspectives about where things were going. Today, you find people like Marc Andreessen, who says this is going to solve every problem in the world immediately, and then you have Yudkowsky and people like that who say we'll all die. The range of options is as broad as it could possibly be. You need people coming in and saying, given that we don't actually know what's going to happen, what is a reasonable approach during this time so we don't fall into a trough? Having that trust and that set of different voices becomes really critical. Of course, you need a CEO who is willing to stand behind it and back it, because AI is one initiative, but there are many others that companies can have. It's got to be put at the top of the list because it's fundamentally about learning.
We'll try to get a few comments in here and still stay on track, finish up at 9:40. Melissa, tell us a little bit about your personal journey and advice you have for others. As you know, it takes a long time to get something on the agenda for a board. More meetings are planned out. Let's do something in six months, something comes up, we wait another six months. Probably less than 90% of the companies we deal with have had meaningful multiple board conversations. How did you see this right away? Did you have to go through this process, and what advice do you have for other companies, other CEOs?
