Industrialization/Computerization Analogies
I have two motivations in this post:
thanks to GPT-4, lots of folks looking at automation via AI seriously for the first time and are unfamiliar with much of the thinking and writing about transformative AI from the past 20-30 years
I think the standard analogy of AI to steam power is slightly off, and I want to make at least the outline of the case for it, even if I don’t have time to invest in making the full case
Thus I’m going to highlight what I think are important analogies between industrialization and computerization and what I think that can tell us to expect from transformative AI in some ways. I want to be clear this is not very careful work so there’s lots of caveats and exceptions and such. My aim is to 80⁄20 the analogy and the explanation.
steam engine : industrialization :: computer : computerization
the computer is roughly the equivalent of the steam engine for the computer age (hence the name)
like the steam engine, initially it had limited applications and minimal impact on the economy
then, after a few decades, improvements in design enabled the automation of key work (for the steam engine, textiles, for computers, data processing)
over the next few decades, design improvements enabled more areas of the economy to be automated, making individual workers more productive and making us all richer
then a big new technology transformed the world
assembly line/mass production :: transformative AI
the assembly line and the creation of the modern factory transformed the world and made us all a lot richer
prior to the assembly line, we can think of steam engines mostly automating jobs that used to be done by hand
those jobs were still mostly done under the old paradigm, but in an a more automated way
the genius of the assembly line and the mass production it enabled was to reframe how stuff got made in ways that enabled a higher degree of productivity
transformative AI, of which GPT-4 is the first instance (sort of like Ford’s Highland Park), is computers having the sort of impact as the assembly line did, dramatically reconfiguring work rather than simply automating old ways of doing it
implications for jobs and productivity
we should expect transformative AI to make us all a lot richer quickly, much like mass production did
not everyone did/will like this, because richer mostly comes by making previously expensive things cheap
old ways of working will be outcompeted rapidly
but people won’t be out of a job, just out of old jobs done in old ways, to be replaced by new, more productive jobs done in new ways
these jobs may be less fulfilling in important ways, just as factory work was less fulfilling than craftwork, but they will also make us all richer and enabled greater leisure and luxury
change will also come faster than last time due to increasing economic doubling speeds, so people will have less time to settle in to ways of doing things before more big changes come
this might actually be a good thing, because when it takes 15-30 years to change an industry that’s long enough for people to make a career doing things old ways—just in time to get screwed as they reach retirement age
I’ve not said anything here about risks, though note that the assembly line enabled the mass production of weapons that were way more deadly than what had previously been available. We should expect similar from transformative AI.
(Final note: I think there’s some irony that I didn’t use GPT-4 to help me write this as a real essay with references and nice paragraphs. Let’s chalk this up to me still learning to use LLMs effectively in my work and using one sometimes imposing enough mental overhead to figure out how to do what I want that I prefer to do the lazy thing and not use it, though I expect future iterations of the technology to enable me to be this lazy and easily produce good results.)
How much analogous will that be? Beyond (maybe) certain ‘general’ largely-manual jobs that would be too expensive to replace with robots initially, it seems to me that there’s no activity that humans would have in which they’ll produce a higher value than even HLAI. And human-caring, initially at least.
At some point AI becomes powerful enough that it’s no longer economical to employ humans. That’s important, but it’s also something like the next phase after the phase we’re entering with AI.
The phase we’re entering now is one where AI will automate and transform work in ways that make humans more productive. It’s a bit unclear how long this period will last. My guess is between 15 and 30 years, because it’ll take about that long for us to grow the economy enough to be able to afford to build AI powerful enough to fully replace humans. This is an often overlooked concern: we don’t build AI just because we can, but because it’s economical to. We’ve seen similar patterns during industrialization: we don’t build a factory or automate something until it becomes cheaper than just paying a human to do it in a bespoke way.
So eventually, yes, AI eats everything, but there’s likely enough years before that when we’ll have to live through a world thoroughly transformed by AIs working with humans as productivity multipliers.