Machine Learning and Course of History
The Truths That ML Will Unveil
In the Sundial Theory of History, I argue that history progresses by civilizations discovering the widespread lies of previous regimes, and creating a new regime out of a combination of newfound truths and newfound lies. I also argue that truths believed uniquely by your era – your regime in time – should be treated with suspicion.
Technology has the power to unwind temporal lies. It can make observing these falsehoods so cheap and ignoring their consequences so costly that it causes a cascade of people going from always lying to always telling the truth.
The practical applications of AI excite me. But the social reorganization it will cause is in some cases even more exciting. One unique characteristic of our age is the vast organizational distance from CEO to leaves (employees with no subordinates). This heavily distorts both downwards and upwards communication. This change was both technological and cultural. First came the need for greater organizational bureaucracies in state and enterprise, to manage large supply chains, customer relations, and other coordination. This created a culture of programmable procedure which furthered the need for organizational bureaucracies to an unhealthy degree, particularly in the public sector.
Bureaucratization was part inevitable and part choice. Clearing it through AI will be the same.
In the original sundial article, I picked my candidate for the greatest lie of our age:
The greatest lie of modernity is that empathy is a feeling act. True empathy, the implicit emotional bond given by stories in the past, is something we now only vaguely approximate as ‘noblesse oblige’.
Current AI companions, such as Character.AI or Replika arbitrage this fact, but do not solve it. They provide a cheaper, purer form of the fake empathy people think they want. However, the first step to solving a problem is admitting it exists. Companion AIs will make this undeniable. Generally, this is how the sundial progresses. Technological regime change makes a once formidable belief of an age ‘unbelievable’, in the same sense Nietzsche says that God has become ‘unbelievable’ — everyone’s daily lives constantly contradicts that belief.
My article “Speculation is the Medium of our Age” extends this theme to media about AI. We’re in a transitionary state when thinking about technology, much like the early postmodern era. Thinkers such as Foucault say the deterioration of norms and the unbelievability of modernism, but did not see the logic of new media structures. I see the current process as the reverse. The rules of new media, of AI, of technological reorganization, have been written and are observable; seeing them written down is anathema to the present man in the same way that postmodern philosophy was sacrilegious to the 60s conservative.
The sundial is not peaceful. Each age has a set of norms and beliefs which cannot be interpreted by other ages as anything other than tyranny or anarchy. So why am I optimistic? There is little distance between dissatisfaction with our age and acceptance of the next.
Socionics offers a helpful dichotomy between extroverted ethics (tend to action) and introverted ethics (intentions over action.) also notes we’re in an era dominated by introverted ethics, but would assume cycles rather than linear progression.
Maybe I've missed something, but I'm not really sure what this piece is trying to say.
Like you don't really seem to explain *how* AI is going to reduce the distance between CEOs and employees, or between this age and the next.
Although I do agree with one of your core statements - that every age has norms that seem barbaric or tyrannical when looked at from another age. We can never get everything right all at once. In fact I don't even think "getting it all right at once" is a coherent concept, now that I think about it.