November 2024
Image: "Robot garden".
Ideas, like dollars, are meant for other people. You may hold onto some, for a while, but ultimately they exist to be exchanged. Their value is only realized when given to somebody else, in the hope of a positive outcome.
This essay is an invitation to discuss an idea: SOL75 and the potential it holds. If it resonates with enough people, it might come to life. If nobody cares, it will simply wither and lie forgotten in the archive of the internet.
But before we get to what it could do, we need to talk about what it can not do.
One of the most beloved traditions in tech is to present a grandiose vision of how one's project will solve the most pressing issue of the day. We will need the next paragraph to excuse ourselves from this rampant techno-optimism.
At the time of writing, 2024, the tech space is flooded by an all encompassing and blinding obsession — AI. Despite a start that over promised and under delivered, AI will likely be as revolutionary as steam in the 19th century and as useful as oil in the 20th. A new wave of automation will affect an unprecedented number of people, disrupting the already precarious balance of power between labor and capital. Unfortunately, as Nobel laureate Daron Acemoglu observed, automation does nothing for labor, it only helps capital [1]. It is great for those who own the machines-that-never-sleep, not so much for those who compete with them. No purely technical solutions, including SOL75, can address such complex societal challenges. Our response must be collective action, public debate and ultimately, government regulation. For automation to be a net positive it must be properly regulated by a functioning democracy.
At the individual scale though, where choices are more a matter of personal freedom than public interest, we should be free to explore automation and to figure out what it can do for us.
If machines, by default, help capital, can we have more people be part of capital? If the immediate benefits go to those who, having direct control over the machines, figure out how to produce useful goods using less resources, why not give individual machines to more people?
Of course, this is not suggesting that everyone should have their own Giga factory capable of producing millions of cars per year, nor does it suggest an increase in production as the solution to all problems. But if automation was in the hands of the individual, it could be used to produce the things that one needs for him or herself, while trying to optimize one's most valuable resource, time.
I see 3D printers as a mesmerizing step in this direction. At the cost of a single machine, one can produce endless different products. Unlike conventional manufacturing, the machine is not specialized. It is also very cheap and, most notably, the process is fully automated. No set up involved, no manual operation required. A modern 3D printer is a one-step automated production machine, available to anyone. It can not compete with the production output of specialized processes (like injection molding), but it was never meant to. In most cases, you only need one or two copies of a design. You fix whatever you need to fix, then move on to the next problem and the next design.
SOL75 wants to be another step in this direction.
Just as a 3D printer enables manufacturing without a factory floor of equipment, SOL75 can get you to a functioning design, without the need for a team of engineers. Like a 3D printer automates manufacturing, SOL75 automates design.
Consider the process of designing a robot. One needs to figure out the right requirements, come up with a suitable geometry, assess the dynamics of the joints, calculate the torque needed for each motor, analyze the material stresses and stiffness, plan the assembly procedure, and so on... Then, once the first prototype is done, if the requirements turned out to be wrong, this needs to be done all over again. The process requires a lot of different specialized knowledge, which usually requires different people. With SOL75, one can work at the requirement level for all the disciplines on which s/he is not an expert.
So what would it look like if anyone had the ability to create personal robots to automate whatever task they needed done?
Robots designed by individuals, to meet their own needs, could mirror the cultural values of the designer/owner in ways current commercial products can not. Having a target market of just one customer is usually not good for business, but in this case, it doesn't matter. The only person you are trying to cater to is yourself.
For example, people from different culinary cultures might come up with different cooking robots. Maybe one for making a specific style of pizza, one for sushi in the style of the Shiga prefecture, or one for baking Saker tarte. The need to design a product for the widest possible market would become irrelevant, allowing for greater freedom of exploration and expression.
Combining the roles of producer and user in the same entity could also solve some of the conflicting requirements that come with each role. For example, as a consumer you would like your (2D) printer to support any type of ink, giving you more options and cheaper alternatives. However, the company that designs the printer earns most of its revenues from selling ink, so it will do anything allowed by law to ensure that only proprietary cartridges can be used. If you are buying a printer, you might spend a lot of time searching for one that works with off-brand inks, while if you are making your own, you might design it to use a bic pen instead.
Personally, I would explore automated food production. Can we build robots that tend to a vegetable garden in complete autonomy? Can I get year-round food production at the cost of some kilograms of plastic and a bunch of electronics?
But none of this is real — not yet, anyway. It is nebulous and distant, and it will not happen overnight. And that's actually a good thing. Moving fast leaves no time to think about what you are breaking, and there are so many questions we need to figure out first. Like what are the dangers of individual automation? Would cheaper robots increase plastic and e-waste, like fast fashion did for clothes? Would it make 3D printed weapons practical? Would it supercharge the existing issue with DIY drone warfare?
I leave you with a lot of questions, not many answers, and a web app that might be useful someday, if enough people commit to it. Admittedly, this is a rather open ending for a very long post. This is not the end of the discussion though. Do you like what we do? Support us and get in touch. Do you think we are wrong? Reach out and tell us why.
For a fascinating discussion on the effects of automation on the economy, listen to this podcast from the center for humane technology podcast.