We’ve been hearing about a particular kind of promise for a while now and it has started to become more mainstream recently, especially after LLMs. The promise that an artificial entity referred to as super intelligent AI, an upgraded version of artificial general intelligence is a decade if not few years away. One which will have more compute capacity at its disposal than that all of the humanity’s brain processing capacity combined and would not cost us the entire planet to run it. When it’s finally here, the world would be a different place assuming everyone has access to it. Optimists say if widely shared and well governed, it could make humanity more egalitarian by giving everyone access to world class education, healthcare, scientific knowledge and productivity reducing advantage of wealth or geography. But if controlled by few governments or organisations, it could result in concentration of wealth, influence, mass surveillance and may provide decision making power in the hands of few elites. Anyway, we’ll focus mostly on the good part today i.e. the part where everyone becomes a builder, an entrepreneur and the collective productivity capacity of human civilisation multiples, not by little but by orders of magnitude because for the first time in the history as nobody is held back by not knowing how
In the last century humanity has solved an extraordinary number of problems. We’ve dramatically, reduced hunger and famine built institutions to respond to wars and natural disasters and developed systems that allow societies to recover from crisis rather than collapse under them. While these achievements are far from universal and many parts of the world still struggle, we have shown an incredible ability to solve problems at scale. If super intelligence fulfils its promise, its greatest contribution for us will be replacing human labour. It will be helping us solve the problems that have remained beyond our reach for a while now, for instance removing plastics from the oceans or carbon dioxide from the atmosphere ensuring that the planet remains habitable for the future generations to come. It could also help us to find cures to deadly diseases or perhaps even prevent cognitive decline that comes with age. Beyond Earth, it may help us accelerate the discovery of physics which is unknown to us, help us understand dark energy or maybe even make it possible for the humanity to go beyond the local galactic cluster to places which are disappearing faster than the speed of light in the cosmic horizon. Whatever the impact of super intelligence ultimately may be, one thing we can be certain of is that it won’t just accelerate what we already do but rather expand the frontier of what is possible and will likely turn today’s impossibilities into tomorrow’s engineering problems
With that aside, let us get back on the good part we discussed earlier, where everyone becomes a builder, everyone becomes an entrepreneur and the collective productivity capacity of human civilisation multiples, not by little but by orders of magnitude. I think this statement in itself is very vague and fantasises more about the possibilities rather than observing the ground truth of how previous information technologies were received by us. Super Intelligence isn’t the first thing thrown around as ground breaking & life changing, every other form of information technology had a same fate. The printing press promised enlightenment but most used it for novels and gossip sheets. TV promised a classroom in every home and best use case has been soaps and serials. Each one lowered the cost of acquiring knowledge & information. But cost was never the constraint. Desire is, and the incentives to act on what we learn is. Technology can make something cheaper and easier, but that doesn’t mean people will choose to do more of it. Their behaviour is ultimately shaped by what they want and what they are rewarded for doing. Everyone chooses Netflix over free MIT course if they have three hours to spare tonight, including me :D
The problem isn’t that people lack time. It’s that they lack a compelling reason to learn. Learning is front loaded pain with a reward comes late and it has to compete with a trillion dollar industry trying to engineer zero friction dopamine. More importantly, curiosity is mostly situational, not a trait. People learn ferociously when there’s a live problem — a diagnosis, a new baby, a visa, a side business. If someone’s life has no loop where knowledge compounds into money, status or identity, self-development is a hobby. And most people pick different hobbies
I think we all can agree that for the most part, we already have everything a person needs to start a business, write code, wire electronics, design a circuit or learn any lesson history has to offer, all of it at our fingertips and most of it is free. So how has that gone? Do more people use the internet to maximize their individual output or do we use it to be entertained? I think the answer is obvious and we had the data before LLMs existed. When MIT and Stanford put their courses online for free, completion rates landed around 3–7% and the people who finished were mostly those who already had degrees
Let’s take an example of curing a disease, it takes a few thousand people in a few hundred laboratories. It doesn’t require anybody else to change how they spend their evenings. That win is a top down one, a small number of motivated people with resources, and the rest of us receiving the benefit passively, the way you received antibiotics without contributing anything to them. Everyone becomes an entrepreneur, however is a completely different claim which is more bottom up. It needs hundreds of millions of ordinary people to want something they don’t currently want. And even if AI makes entrepreneurship cheap, people must still want to become entrepreneurs. They must be willing to take financial risk, find customers, deal with uncertainty, sell products or services, handle failures and rejection or give up some of the stability of a conventional job. So what do we do with any Super Intelligent AI? Two things and we can already see it. First, endless entertainment made just for you i.e. shows, videos, games generated on demand, all tuned exactly to your taste. Second, and this is the one people underestimate: AI companions. Friends, partners, someone who always listens and never gets tired of you. That’s already one of the biggest uses of chatbots today, especially among the young and the lonely. Things already competing for our attention could become so personalised and compelling that they capture even more of our time and will likely dwarf how much time we spend scrolling currently
The reason companionship wins is the same reason entertainment wins, only sharper. Both satisfy our desire for immediate gratification with minimal effort, but companionship goes a step further by fulfilling our deeper need to feel understood, valued, and connected. AI companionship and assistance could make our lives more comfortable while simultaneously removing some of the reasons we develop, learn and grow. Historically, discomforts like loneliness, a failing business or a personal crisis pushed us to acquire new skills, build relationships, develop emotional maturity and solve problems we couldn’t ignore. But a system that can explain, comfort and solve on demand can dissolve these problems before they become catalysts for growth. If AI handles every difficult work problem, we might never develop the expertise we’d otherwise acquire through experience. Similarly, many innovations and businesses emerged because we encountered problems we couldn’t ignore: we built better tools to overcome inefficiencies, created products to meet unmet needs, and developed new skills to change our circumstances. But if AI can satisfy our needs and eliminate our frustrations immediately, we might have less motivation to create something ourselves
So the two futures don’t compete, they arrive together. A few thousand motivated people, the ones who were always going to be motivated, point the tool at oceans and tumors and the edge of known physics, and the rest of us receive the results the way we received antibiotics: gratefully, passively, without having to touch the work. Meanwhile everyone else gets something that feels, to them, more valuable than a company they’ll never start — a world tuned exactly to their taste, one that entertains them, listens to them, and asks nothing back. And this is the part the optimists get right, technically. Super intelligence is for everyone. Everyone will have it. It’s just that having it and using it to build were never the same thing, and access was never the problem. The MIT course was already free. The problem was that the course is work and the alternative is not and the alternative is about to get much, much better
I don’t think this is a tragedy, exactly. Most people were never going to write a compiler or start a company, and telling them the only thing in their way was access to knowledge was always a little dishonest. The gains will be real and widely shared, and that is the genuinely egalitarian part, it’s not nothing. But we should be honest about how it’ll look down the line. There won’t be a hundred million new entrepreneurs but rather a small number of people with god tier leverage, and a very comfortable everyone else. Which is fine, maybe. It’s just not what we were promised. We were told the ceiling was coming off i.e. anyone could now rise as far as they wanted. What’s actually coming off is the floor, the last few hardships that used to make a person reach up in the first place. And when the tool that could have made you anything is also the one that makes not reaching feel wonderful, most people, most nights, will choose the thing they were always going to choose
Including me!