The two pressures shaping our future hobbies

Paul Graham’s advice is simple: “Make good new things.” He argues that making something is often the best evidence that we have thought clearly, whether the result is a scientific theory, a piece of music, a painting, or a useful tool.

AI will make that advice easier to follow. It will also make it harder to know whether what we make matters.

I see two pressures fighting each other.

The first pressure is the falling cost of creation. Personalized software will become much easier to make, and the cost of many hobbies will fall with it.

Someone who has always wanted to make a short film may soon be able to do it in an afternoon. Someone who likes designing board games may be able to test dozens of ideas without spending much money. Someone who wants to write music, illustrate a children’s story, or build a simple app may no longer need a large team or years of specialist training.

The same will apply to less obvious hobbies. We may make private games for our families, software for a very specific interest, interactive stories for our children, or tools that fit our own habits exactly. More people will discover what they enjoy because the barriers to trying things will be lower.

That is a good thing. Many people have interests that never become activities because the tools are too expensive, the learning curve is too steep, or the first attempt feels too embarrassing to share. AI will let more people try.

The second pressure comes from the same change.

As personalized software becomes more prevalent, our hobbies may become more siloed. We will be able to create things for ourselves with increasing precision, but those things may have little value outside our own lives. A game built around my private jokes may be delightful to me and meaningless to everyone else. A story written for my exact mood may not be something I can share. Music generated for my own taste may never become part of a shared musical world.

The cost of hobbies will go down, but the impact of those hobbies may become more fragmented.

This is the unsettling possibility. We could end up surrounded by an endless supply of things made specifically for us, while having fewer things that bring us together. Each person has a personalised stream of games, stories, music, lessons, and small digital experiences. The supply is abundant. The common ground is thinner.

That could also challenge our idea of purpose.

People often find purpose in becoming good at something, making a contribution, and seeing that contribution matter to others. If AI can help us produce almost anything cheaply, much of the work involved may start to feel trivial. We can make a song, a game, a film, or a piece of software, but the act of making it no longer proves very much. It may be enjoyable without feeling meaningful.

The problem becomes sharper when the result benefits only us. A personalised hobby can fill an afternoon, but it may not give us a role in a community or a reason to feel needed. We may have more freedom to create while feeling less certain about why creation matters.

That makes original ideas more important, not less.

When creation is cheap, production will not be the main difficulty. The harder question will be what to make that benefits other people. A good idea will need to travel beyond its maker. It might give people language for something they have felt, help with a problem they share, or show them an ordinary part of life in a new way.

This is where Graham’s principle becomes more demanding. “Make good new things” does not mean produce more things. It means make things that are new, good, and worth having in the world.

AI will give many more people the ability to create. The first pressure will push us towards richer personal lives, with hobbies shaped exactly around our own interests. The second will pull us towards isolation, as those hobbies become private experiences with no audience and little shared impact. It may even leave us asking what our work is for when the work itself has become easy.

The best creators will have to work against that pull. They will use cheap, personalised tools to make things that can still leave the silo, reach other people, and give them a reason to care.