Beyond the Knee of the Curve

Earlier this year I asked my daughter for permission to convert one of her pieces of art, “Balloon Man,” into an NFT. She refused, told me she thought NFT was bullshit, and that I should make my own art if that’s what I wanted to do. The thing is, to the extent I have any artistic talent at all, it is with words, not pictures, so I gave up on the idea. But today I prompted MidJourney to imagine a woman holding a gold balloon in a blue, empire-waisted dress, “Balloon Woman.” Generative art, whatever its merits aesthetically (or not) offers a strange alchemy: to cross from one domain in the arts to another. Martin Mull once said that writing about music is like dancing about architecture, but what if he was wrong? What if you could do both those things?

Whether we are talking about ChatGPT or MidJourney or other related models that have recently exploded in popularity, inevitably it raises concerns about their impact on the job market. Artists will go extinct. Everyone will be unemployed. We will have to put in place a Universal Basic Income (UBI) to deal with widespread joblessness. FDR recognized that a social safety net to catch people ravaged by capitalism’s creative destruction was the only way to save it from revolution, but he probably could not imagine a world where human work is no longer required at all. Even replacing shovels with spoons will not solve that scale of problem in the labor market. Sometimes crypto in combination with UBI is positioned as part of the answer here, but much as I think Web3 will play a huge part in transforming the economy in the coming decade, I actually don’t think this is why or how it will happen.

Still, these changes will come at a cost. When you disaggregate vision from technique, when the ideas matter more than the doing, you’re better off if you are an architect and worse off if you are a bricklayer: it is very likely that these technologies will favor the winners and make the wealthy even wealthier. Without a much stronger educational system and a conscious push to ensure the benefits of these technologies are spread widely, the Gini coefficients everywhere will spike higher. Furthermore, I believe we may have just passed the knee of the exponential curve on these technological changes. Past this point the human mind struggles to keep pace: we don’t reason about exponential rates of change as readily as linear ones, as pointed out in The Second Machine Age. At the time of that book’s writing, the authors warned that the knee of the curve lay just ahead. I wonder if as of the fall of 2022 the knee’s bend now lies just behind us. We may not be able to adjust quickly enough to these changes, even if they are ultimately positive.

The Second Machine Age talked about such technologies as a way to augment rather than replace human intelligence. Being able to write about music or dance about architecture is just one example: that my artwork on MidJourney is unimaginative and poor has nothing to do with the tool, but with my own artistic vision. It does not matter if that’s expressed in words or paint; MidJourney did not make me a visual artist, and I expect there will be a generation of artists who find creative ways to use these tools to amplify themselves, just like programmers using GitHub Copilot or a question to ChatGPT will boost their productivity. But something curious happened there: someone who has no visual skill was able to create something using a skill he does have, with words. This suggests that another consequence of these technologies is to open up certain domains to people who lacked access entirely. They may not create brilliant art, but they can do something previously impossible for them. So there’s a chance that such changes make the people already talented in this domain better, and make people without that capability able to start. If this is true, the story of work is not over; the lump of labor will divide and divide again, as it always has.

The question then is not can it be stopped (no) or should it be stopped (no) but what do we need to do for this particular transition to be as positive as possible. This is where Web3 does come into the picture in my opinion, at least in part, though as I alluded to above there is a massive educational component as well. DeFi offers a chance to change the economy, not just finance, and there are a number of challenges ahead — not just with AI/ML, but the energy transition and climate change — that require economic change. I write a lot about risk management, and this is maybe an example of social risk management. There is definitely a way these transitions could go very badly for the world if not approached thoughtfully. So while there are a ton of new possibilities being opened up right now, they are not guaranteed to be good ones. This is our choice: do we use this chance to remake the economy to be more equitable, grow wealth broadly, and ensure everyone can participate in the change and not get left behind, or do we just assume it will all work out fine and keep on building?

“You may not be interested in war, but war is interested in you.” —Leon Trotsky

Technology lacks moral valence; at most it raises moral questions when it enables things that were previously not possible; the ethics of genetic engineering is a good example of an innovation introducing new questions about what’s right and wrong. We get a say in the kind of world these innovations produce: none of them, to borrow Vitalik Buterin’s phrasing, is inevitably dystopian, just possibly so. And as Dr. Oppenheimer realized, looking across the desert at the devastating results of the Manhattan Project, often it is the creators who are first forced to think about the consequences of their creations. This is, in my experience, something many technologists are not comfortable with or even equipped to think about, because it’s well beyond the STEM student comfort zone. But being ill-prepared and unwilling to confront these questions does not make them go away. As a very difficult 2022 comes to a close, it’s worth spending some quiet moments in the days ahead thinking not just about what we build or fix, but what we hope to achieve when the work is done. For our part at Cloudwall, we hope to find a way to support these changes in the economy by giving the investors funding the change a better understanding the risks, but the world does not lack for problems. What’s yours?

Happy holidays, and we hope a fairer and more prosperous 2023 lies ahead for us all.