We’re increasingly curious about the future of artificial intelligence.
The AI boom of recent years has outpaced the ability to understand or predict the future. While AI is expected to unleash a new age of productivity, questions remain about what it means for humans to live in an AI-driven world.
Technode spoke to Zack Kass on Thursday, Global AI Advisor, Former Head of Go-Market at OpenAI and former Global AI Advisor at OpenAI. He shared his insights from the BEYOND Expo.
The interview was edited to ensure clarity and brevity.
Q: In the past, every major technological breakthrough has led to the replacement of some jobs. It seems that AI is on a whole new level. It’s threatening to replace so many white-collar positions at once. The new jobs that it creates don’t seem able to keep up. It feels like societies are struggling with how to respond, especially as the global economy is already under pressure. What do you think about that? Do you think that this time around is different? If so, what should we do about it?
Kass: I think we need to acknowledge a few things. First, it seems unlikely that we will see true full automation in all industries and economic sectors. This is for a variety of reasons. Many jobs are protected by regulations and compliance, but many others are not yet fully automated. It’s true, however, that many jobs will be automated at least in part very soon. We don’t even know what it will look like. It probably fundamentally alters the nature of much of the work we’ve done.
We should also take into account the fact that we do not yet know how many jobs AI will create. AI has the potential to reduce the cost of providing a quality service through automation. On the one hand, this will relieve us of many jobs, but on the other it will increase its purchasing power. The net result is a new demand for goods, services and other products. I don’t believe we should underestimate the economic impact of people feeling relatively wealthy.
We are likely to see a new demand for bespoke works, as we do whenever we undergo an industrial revolution. There’s the automation of some work and then there’s the demand for new labor that people consider to be luxurious or bespoke.
It is unlikely that chefs will be fully automated. However, the likelihood of more people wanting to eat out in restaurants is higher. We don’t yet know how the labor market is going to play out.
We will also observe a world where there is a lot of economic abundance. We have automated so much of our work that we don’t sit around anymore asking ourselves, “What about our jobs?” From an economic perspective, I don’t think we’ll find ourselves saying we’re sorry we can’t do any more work because we’re now starving. I think that we will say it’s a pity we can’t do any work because we don’t understand what our purpose is. If you ask me, the biggest risk of all this automation is not an economic crash. Automating our work will have a profound impact on the economy.
What I expect is that, for a time, we will live in an environment that is very confusing for us. It will be much harder for us find a sense of purpose and identity, because we have found it for so long in our jobs. This is a great opportunity for the human race, as at any time, there are two billion people on Earth who want or need something to improve their quality-of-life. I am very excited to provide these goods and services soon. Everyone else is also curious about the meaning of work if it is not to earn a living. We’re about to find out.
Q: What exactly do you mean when you speak of near-zero costs for products and services? And how can we interpret this vision?
KassAll that we see, all that we do, everything was scarce at one point. There are very few exceptions. Freshwater, food and electricity were all once scarce. We’re now living in a world where everything is available. Many people wear stylish shoes here, not to cover their feet but to express themselves. It’s a post scarcity reality. Shoes are so cheap that luxury versions exist solely for self-expression.
We live in such abundance now that we barely notice. Many people don’t realize how abundant things have become, because many goods and service are now available at near-zero cost. Take water. Most of us today drink water without thinking. It wasn’t possible for most humans to do this until plumbing and irrigation were invented. Food is no different. Food was rationed for millennia until we developed methods of preservation. Energy was scarce too. Even after electricity was discovered, its use was limited. You can now charge your phone without worrying about the cost of energy.
The internet was born. I remember dial-up. It took 10 minutes to connect and you had to ask others to hang up. It was expensive and scarce. We use the internet for free today. It’s almost free. In India, for example, many people pay less than US$7 per month for unlimited access. This is what a near-zero price looks like.
Imagine where this will lead. Unmetered energy – likely via fusion – would mean that we stop thinking about our energy use completely. AI will allow us to achieve unmetered intelligence, which means no cost for computing or solving problems. In the past, solving difficult problems required gathering smart people with food, water, rest, and other necessities. Now, we plug in GPUs to solve problems.
I don’t share this to be pedantic but to highlight the power of deflationary technology: It drives costs so low we no longer think about consumption.
Q: Are you saying that AI will also become like this?
Kass: I’m saying AI has already reached a point where we no longer consider how much we use AI, which is already close to this idea of unmuted Intelligence. The machines are cheap and smart. You have a PhD as a sidekick. You can ask whatever you want, whenever it suits you. It’s pretty amazing to think about.
The next step is to solve other problems. These other problems will then drive down the price of other things. What would it look like if a factory produced a product or service was fully automated? This is probably a world in which we don’t think twice about the amount of food or drink we consume.
Q: Do You believe that AI will eventually reach a state of no return?
Kass: I’m sure. I think it’s so uncertain right now that we can’t even imagine the end state. It probably feels more concrete. I expect we’ll live in a new world with rapidly evolving scientific frontiers. AI may also have a limit. I don’t really know, but I remind people that there comes a point where the next model or breakthrough is no longer relevant. There comes a point where you don’t really need anything smarter and we’ve already automated so much.
The question I ask everyone is: How much automation do you want? This novel concept, which I often discuss, is called the societal limit. It asks the question of whether the future is determined by the machine’s capabilities or by what we want. Just because you have the ability to automate everything does not mean you will or should. What is the point where we say that this experience is unique and immutable and we shouldn’t automate? It doesn’t improve our lives. This is, to me, one of the most incredible outcomes of AI.
Q: Is AI just automation? What about its ability?
Kass: Yes, but reducing AI to automation is an interesting and clever way of thinking about what it is. AI can be so complex and scary to many people that it is hard to understand.
AI’s ultimate promise is to automate the things humans want to do. The cure for cancer, for example–10,000 people are currently studying the cure on Earth. We would all love more oncological research, but there are only a few people who can qualify for the job and land it. We’re saying that we should automate some oncological research to supercharge these people’s productivity by adding a new level of productivity to the market. If you think of it in this way, then you can sort of go over all the things humans do collectively and individual. AI offers the chance to automate these things. This is how we describe these things as economic and novel scientific gains.
It is obvious that it is more complex for many people. When people say it will replace our creativity, I ask: Do we want to automate the creativity? We sit around asking: Will AI do that? I think the question is much simpler: Do we really want to live in an environment where we can’t think creatively? No. It’s easy to see what AI does. In reality, we want AI automate their new scientific discoveries. As we try to understand what AI can do for us, the question of what it means to “be human” will become more and more relevant.
Q: What are the most important skills or qualities for young people who want to enter the AI field?
Kass: It depends on what you want to do. Many young people who are interested in AI think about it from a purely technological perspective, which is fair. I would caution anyone thinking that their technological abilities are unique. AI will offer many opportunities for those who are interested, but they must be willing to adapt and change quickly.
People’s expectations about their career often differ from the reality, especially when it comes to a field that is evolving so rapidly as AI. Initially, people aimed to work on the model level (like OpenAI), then moved to application development and finally infrastructure. This is a lot to take in. It’s clear that the AI boom has created a lot of new economic opportunities. Working in AI is now easier than ever. I began working in AI 16 year ago, when there were only about 1,000 data scientists. It didn’t matter that no one was doing it.
What people need to understand is that their perception of AI will change a lot. The idea of what people should do will change quite a lot, and it is important to be able to adapt to this.
