Abhay Venkatesh

The Anti-Search Engine
Technical Resources
The New Atlantis

Against Autopilot World

I recently came across a progressive critique of capitalism which argued that “capitalism forgets that life is social”. Capitalists would dismiss this argument out of hand as self-contradictory non-sense, since clearly capitalism can do cafes, bars, clubs or literal social networks. But rather than quickly jumping to conclusions, let us steelman this argument. What if we rephrase this progressive critique as “today’s capitalism forgets that human life requires agency”?

Zuck is not on Autopilot

There has been an acute transfer of agency from humans to processes in commercial activity. This is reflected in the excessively large size of the financial sector in the Western world. The finance sector is dominated not by substantive ideas but by processes, such as diversification of portfolios or pursuing abstract trading strategies. Another sector deeply affected by this problem of process domination, or “legalism” if you will, is the government sector. The government sector is staffed mostly by lawyer types who are focused on following the laid-down rules, as opposed to say, engineer types who are focused on building concrete and specific things of value. The military industrial complex also suffers from this problem, in which large amounts of money gets spent on cost-plus contracts regardless of outcomes. We could describe such a world as an “autopilot” world, which operates based on some (written or unwritten) algorithm mostly in absence of human agency. Such an autopilot world has lead to recent catastrophes such as that of the F-35 or the Boeing 737 max. To make a reductionist argument, one could think of large parts of the economy (and maybe even the world in general) to be set on such an unthinking autopilot.

We can contrast the worlds of finance or government (or defense, or even healthcare or education) to the world of startups, which is the opposite of a process-driven world. In startups, most of the value gets created by substantive ideas executed using agency. For instance, an idea for creating a social network requires bold new ideas, whereas a step-by-step process to balance a portfolio of stocks simply requires you to follow the instructions. People notice this stark contrast, and for that reason, they pay outsized attention to startup founders such as Elon or Zuck. In practice, this fixation on startup founders turns into some sort of scapegoating mechanism. In the case of Zuck, his scapegoating is an un-personing of him by calling him robotic, or a lizard person on the internet. But we should realize that people like Zuckerberg are actively trying to do substantively new things with real agency to add large amounts of value to the world. People might think Zuck is a lizard person, but he’s very human in the sense that he’s trying to have real agency with his life. We are genuinely lacking this aspect of our humanity in 2019 and we should try have agency of our own instead of trying to pull down those who are succeeding at it. Perhaps the insecurity about automation (even though unemployment has never been lower) is really about a fear of having our own lives on autopilot. Zuck is not on autopilot.

Agency and the End of History

One could argue that there is no use for agency if we lived at the end of history. A world where we abandon all agency would increasingly become a narrow and fake world because an abandonment of agency represents a retreat from the real world. We can observe many aspects of this fake world in 2019, where we consume fake foods, have fake social lives on TV, and perform great fake feats in video games. At the end of history, after the abandonment of agency, we might even descend into becoming “bug-people”, who act not on our agency, but by instinct or reflexes.

But we are not at the end of history. Therefore, we need to find a way to restore agency in our world. If we do not restore agency, then it will become more likely that the future will be worse than we expect it to be. For the belief in the autopilot world to be rational, one also has to believe that all problems have been taken care of, and that there are no more problems remaining to be solved. This is clearly not true. There are a whole host of problems from depression, to loneliness, to mass incarceration, to opioid addiction, to malthusian food crises, to the threat of authoritarianism that should remind us that we do not live in a world where all problems have been solved.

If we are able to stomach the fact that agency is desirable in 2019, then the remaining problem is that agency appears to be an extreme luxury that only people like Elon Musk can afford. Since people have existed in a world that exhibits so little agency for so long, they have internalized the lack of agency, and now see entrepreneurs such as Elon as a negative role model. That is, many people believe that they could never do what Elon has done because it is too hard. It would seem like a life of agency is a kind of aristocratic life, not available to most.

The Pervasiveness of the Space for Agency

However, in reality, there is always space for agency, even in the most restrictive scenarios, and this agency is accessible to everyone. To try and understand why, we can begin by imagining what it might mean to be “sovereign”, a person with maximal theoretical agency. One could define the “sovereign” as the “entity who gets to decide on the exceptional cases”. The exceptional case can be thought of as any situation that is not accounted for by the rules or procedures of the organization in question. Hence, attaining agency is equivalent to attaining sovereingty. Can anyone be a sovereign, at least in a narrow manner under specific circumstances?

In any given state, there is the problem of outlining the mechanistic procedures that will operate that regime in the form of rules and regulations. This problem is unsolvable without the introduction of judges, arbiters, and lawyers. To convince oneself of this, one simply has to either attempt to write down a set of rules (or code) that would replace every lawyer and judge, or enumerate all outcomes of any organization. This has never been possible, and likely never will be unless we are able to invent AGI. Since there can be no complete description of the rules for any regime that works without human judgement, in the end, the rule must be of humans and not of processes. This would suggest within any sort of polity, there should be space for agency (or equivalently, sovereignty). Even in a prison-like scenario, there might be more courses of action than initially visible. Perhaps there is an absolute agency vacuum if and only if there is an intense belief in the impossibility of agency. That might even be the world we live in today.

Startups will Make Capitalism More Human

If we live in a world where there is an intense belief in the impossibility of agency, and simultaneously, if this world of ours is on autopilot, that is, actively not making plans, then paradoxically we might be in a world where there never has been more space for agency. In this world, there might be many exceptional moments happening all the time that get passed by. The exceptional moments, which can even be thought of as miracles of sorts, are the moments for power-law outcomes: the moments for black swans.

Given that this world of ours also coincides with an exceptionally prosperous and less violent world, agency has never been more accessible to the medians of the world. The best place to take advantage of this opening for agency is in the world of startups. It also happens to be the case that we live in a world of low interest rates and easy money, meaning that venture funds cannot find enough startups to fund fast enough. If more agency in our world means more of humanity, then startups could make our commercial world more human. Turn off autopilot.

Thanks to Brian Timar, and Bonnie Kavoussi for reviewing drafts of this and offering their great feedback