‘Biz leaders must find ways to use AI to deliver value to consumers’

AI is top of mind for all sectors. How do you see the future developing?

What we have seen so far is the digitalization of the B2C (business-to-consumer) sector. Think of books, travel, films. Google, Netflix and Amazon, which have changed the lives of consumers. The next frontier is the digitalization of B2B (business-to-business) industrial companies – from tractors to cars and aircraft engines.

The fundamental difference in what we have been able to digitize so far in the B2C sector is what I would call pure information sectors – like Google, or physical products, for example cameras, which have disappeared and been overtaken by mobile phones. In industrial B2B companies, the physical product, such as a tractor, will never disappear. But the value will migrate from the physical product to data and AI. Therefore, this is the most exciting frontier, because the total GDP in the world is 100 trillion dollars, of which only 25 percent is in the B2C sector, which is digitalized. And as you digitize this $25 trillion, think of the value we have unleashed.

Imagine if we digitized the remaining $75 trillion. Why didn’t that happen? There are three main reasons. With one technology you can digitize the B2C sector, via the mobile phone. But you can’t digitize a tractor with a cell phone. You need sensors, computer vision and IoT. All these technologies are out there, but their costs are prohibitive and it will take some time for costs to come down.

Tesla has shown that we can digitalize cars. But many other sectors still need to be digitized. The second reason is that in the B2C sector, an accurate recommendation of 80 percent is good enough. Suppose Amazon gives you a recommendation for ten books, and you only liked eight. It may be uncomfortable, but you can live with it. But an airline can’t get 80 percent recommendations from Rolls-Royce because if you fly the plane miles high, that would be disastrous. The third reason is that traditional AI models could only analyze structured data. Although industrial sectors also generate structured data, most of the data they generate is unstructured and qualitative – for example an image or the sound of a machine.

But GenAI is just what the doctor ordered. When you put these three together, you can develop large industry language models that are 100 percent accurate.

Will the AI ​​revolution be as groundbreaking as, say, the Internet revolution of the early 2000s, or will it be even more of a game changer?

Without question. It will be a thousand times more transformative than the Internet. You see, we’ve seen three fundamental turning points in technology. Web architecture was the first. It only decentralized information. Secondly, there is app development. However, apps are fragmented: different apps for different purposes. In the whole agent age that we’re entering now, it’s being able to learn from these different agents and then solve your problem. It can coordinate all fragmented sources of information: some may be in my wearables, some may be in my finances or, for example, information from a hospital. It’s able to coordinate all of that and give me real-time recommendations.

That’s why this is huge. This is bigger than the steam engine or the industrial revolution. This is definitely bigger than the decentralized web and app architecture.

Is AI a bubble or does it really exist?

Many people are wondering if we are seeing an AI bubble since we have spent and are still spending billions. AI has been going for five years now and what we have done is build the infrastructure. Because all major language models are essentially infrastructure. The next five years will show whether there is a bubble or not. Only when companies build applications will it create value, and the investment in AI will pay off. Again, go back to the Internet. We spent a lot of money developing it. But the value was created by Netflix, Amazon, Airbnb, Uber… That hasn’t happened yet in AI. So the next five years are important. If AI is to pay off, business leaders must go a step further and think of intelligent ways to use infrastructure to deliver more value to their customers. Then AI will pay off.

AI moves at warp speed. While you have a generation of managers who learned in an era when everything remained static. Accounting or marketing principles would change only incrementally. But now things are moving fast. Is academia retraining itself?

Let’s talk about higher education. I believe AI will transform all industries. Even GenAI’s current capabilities have the potential to transform higher education, not to mention how quickly this technology is improving. We need to use AI intelligently to provide more education to our students. The position I would therefore like to advocate is that we should have a bifurcated approach to AI in education. By that I mean that some parts of our education should be an AI-free zone, where the fundamental knowledge is taught. If you study computer science, you need to know how to code. Even though GenAI can code, you have to teach them how to code. It’s very important. In the same way, in a writing course you have to write yourself. If you don’t, I think you’re losing a very valuable skill. I think we need to use AI very generously, but as teachers we can still add a lot of value to students. Because things like judgment, persuasion, communication, connecting the dots… AI can’t do it.

Let me give you a quick example. I’m going to do an experiment where I put the assignment questions for my classes on ChatGPT and provide the students’ answers on ChatGPT. Read that, think about it, and come to class. If I can’t add value to what ChatGPT told you for 90 minutes, you don’t need me. If you replace thinking with ChatGPT, I will expose you in five seconds because your learning is superficial. The point I’m making is that ChatGPT is an information processing tool, but humans have a tremendous ability to judge, make choices, and convince people to accept your choice.

A basic skill is knowing how to add. Recently I went to a store in the US. I bought five shirts, each cost $55. That day the cash register broke down. So the man behind the counter, his brain stopped working. Because he was given a calculator from the start. So he never knew how to add, took a piece of paper and slowly added $55 plus $55. I told him it was $275. He looked at me for a bit, but then returned to his paper and pencil and slowly added something. After five minutes it came to $275. And he looked at me and said, ‘You must have a PhD from Harvard.’ I said: that’s me! But that’s beside the point. He didn’t know the basic logic of addition and multiplication.

Therefore this is what I say. We need to separate fundamental skills. Let them use GenAI for information processing. But there is a higher level of intelligence that humans have, namely AI-resistant skills that only humans can do. Therefore, I would say that higher education needs to embrace this divided thinking when using AI.

Do you see companies retraining their employees to use GenAI?

The companies that are leading are absolutely doing this. A Honeywell, a John Deere and a Siemens are retraining their employees to be ‘digital first’. It is absolutely important to be digitally first and digitally savvy. But not all companies do it. That is why these are the companies that will take the lead.