“Our fine arts were developed, their types and uses were established, in times very different from the present, by men whose power of action upon things was insignificant in comparison with ours. But the amazing growth of our techniques, the adaptability and precision they have attained, the ideas and habits they are creating, make it a certainty that profound changes are impending in the ancient craft of the Beautiful.” – Walter Benjamin, 1935
Less than 200 years ago, a small group of people decided what counted as art. They could do this because only a few had both the talent and the patronage to create it. The vast majority of art that was created and survived to this day was therefore images of the rich and powerful.
Speaking as an artist for a change, I also know that artists are an ornery lot. They frequently stepped out of line and painted things that they were not paid for. However, even then they represented a very small slice of the population with the imaginative and physical talents necessary to create complex works of art.
With the invention of photography, however, the level of physical talent required declined. Suddenly people who were previously excluded from expressing themselves individually had the power to do so.
Initially, the costs and time necessary to do photography limited the group of people who could create images, but with the introduction of technologies like the Kodak Brownie in 1900, suddenly the masses had access to creation tools. Now, we take over 2 trillion pictures per year.
This literally changed what we can see. Visual media are far more powerful than textual media, especially when a large proportion of the population was illiterate. But even today a picture is far more dramatic than lines and lines of text like I’m producing now.
Technology democratizes access to creation. It empowers people to do things that were impossible before. This goes against the centralization theory of technology, such as that expressed by Tim Wu, which argues that technology will inevitably be gathered into a small number of hands.
In an era of tech giants, that is an easy theory to believe. AI carries the baggage of Web 2.0, where centralized platforms like Facebook and Google built empires by controlling data and attention. Companies like OpenAI and Anthropic are explicitly modeling themselves on the first generation of Web 2.0 giants.
These companies were, in turn, modeled on industrial models where massive machines produced everything from automobiles to washing machines. In the industrial age size meant power. It manifested in systems of production that dictated how we worked and what was available to us as consumers.
Over the last 40 years, this model has been undermined as production has become far more efficient. Technology has allowed relatively small startups to become successful. Of course, the founders of these startups were raised in the industrial age and very much see size as the ultimate arbiter of success.
However, we no longer need thousands of people to produce one product. It can be produced by a few dozen people or even one person with the right technology.
We have also democratized access to art. We went from glass plate photography to the iPhone in the span of a little more than a century.
When we democratize access to creation, we empower new voices. In societies where people are increasingly frustrated by the old voices who seem to only lead them down blind alleys, this is a powerful force to counterbalance the conservative elites.
I agree that sometimes in order to build stuff you have to go big. However, why does this have to be the end state? We used to need a massive factory to do injection molding. Now I can create custom versions of the same thing in my living room with a 3-D printer.
Until recently, the biggest barrier to makerspace innovation was the complexity of design software like AutoCAD, which was needed to operate CNC machines or 3D printers. Now, I can get an AI to generate models for me with a text prompt.
As I play with the relative capabilities of large and small AI models, I recognize the power of brute force when it comes to transformer technology. At the same time, I see a narrowing of the gap between the smaller models and those that require huge server farms.
Of course, those huge server farms come with immense cost, and only large corporations and governments can shoulder that level of investment. With that, comes control. They may control the smartest models in the world, but where does the true power lie? It lies in the creative inputs they use to train the models.
I have argued before that the power of the PC revolution was in its democratizing force. It spread technology over a much wider group of people and every one of those people had a brain and experience that were unique. Kevin Kelly called this the Hive Mind. Diversity is a bedrock of innovation.
Even with today’s biggest models locked behind corporate walls, I can now accomplish things that were once impossible for me. I can run powerful models locally on my Apple silicon Mac and use the big models for a limited set of problems. This is why small, open models are the future, not massive server farms.
This is tremendously empowering for humans. AI has made me a more efficient teacher. It has made me a more efficient creator. And the only real limits are in my imagination. And this is only the beginning.
… of a three part series on our transition into an AI-driven maker world.
Part II: Making Opportunity With AI
Part III: A Maker World