There is a scene toward the climax of Captain Marvel in which Captain Marvel gives a riling speech about what it means to be human and the unique capability of humans to persevere all while she’s fighting the avatar of an authoritarian artificial intelligence that rules the Kree civilization. It’s all very touching and meant to stir feelings of triumph and agency in the – assumed – human viewers. Putting aside the fact that Carol Danvers aka Captain Marvel is arguably not entirely biologically human AND that it’s kind of a jerk move to give a species-based speech as she is working to stop a species-based genocide, I couldn’t help but root for Captain Marvel. “Hell yeah,” I said at the theater screen, “I’m a dumb human, too, and that means something and is important!” It truly was a blessed day for humanists everywhere.
Yuval Noah Harari Enters the Chat
Unfortunately for me, and the rest of human society as it currently exists in 2019, this humanist life is in danger, girl. Yuval Noah Harari – Oxford grad, lecturer at the Hebrew University of Jerusalem, pop star of academia – thinks modern society is moving away from Captain Marvel’s deliciously humanist pep talk and toward the Kree’s vision of surrendering authority to algorithms. He explores this plausible future in depth in his book, Homo Deus.
Before Harari takes us down the road of the inevitability of the Supreme Intelligence becoming our world religion – he calls this concept “dataism” but genuinely they’re the same concepts – Harari brings us through sources of societal authority throughout the past and present. The evolution of societal authority goes from the divine to the human to the machine. To me, this seems like a pretty well-presented argument. Humans once sought guidance and judgment from scripture and those who would interpret it: classic Divine Command Theory. From divine authority, humans moved to the source of societal authority that we still hold dear today, humanism. Within humanist concepts of authority, nothing is higher than human emotions. Things are not good because God demands they are, things are good because they make someone feel good. This is where Harari has his diva moment. This is where Harari swan dives off the massive platform of history that he just built up into a melodramatic pool of generalization, technological determinism and sociotechnical blindness. This is where Harari says that homo sapiens as they currently exist will be gone in as soon as 50 years because algorithms will know us better than our feelings can. We will give up the authority that we place in our emotions, and humanity will no longer be that which gives the universe reason. No, the only thing to give us truth in the future is data. The proposition is simple: algorithmic knowledge is beginning to surpass human experience-based knowledge, and thus, people are beginning to view data-driven insights as the ultimate authority. This relinquishment of authority will naturally lead to the worshipping of algorithmic intelligence. Harari calls this new religion Dataism.
We will give up the authority that we place in our emotions, and humanity will no longer be that which gives the universe reason.
My biggest issue with Harari’s claim is that his concept of dataism almost perfectly parallels with the Supreme Intelligence from Captain Marvel. My issue with this is not that he might have copied Marvel (which…. 👀), but rather the absurdity that an alien civilization thousands of lightyears from earth and humanity would have the same path of technological advancement that would lead them to one outcome: divinely intelligent data-driven intelligence. The Chilean government in the 1970s and the French government in the early 1980s had similar text-based computer networking technology. Chile chose to try to create a cybernetic system of total economic efficiency, while the French tried to cheaply phase out phone books and ended up with an anonymous system to explore sexuality. This is all to say civilizations on the same planet given similar technologies created radically different technological outcomes, suggesting that perhaps humanist desires and feelings are projected into the systems they build.
Additionally, Harari indulges in a bit of sociotechnical blindness when discussing the future of humans and technology. Deborah G. Johnson and Mario Verdicchio describe sociotechnical blindness as blindness that obscures “all of the human actors involved and all of the decisions necessary to make AI systems,” and “allows AI researchers to believe that AI systems got to be the way they are without human intervention.“ There’s an assumption underlying most of what Harari posits about the future that artificial intelligence will continue to advance at the same rate of advancement that it is right now. Putting aside that this is such a cliché of artificial intelligence that there are charts on Wikipedia’s Progress in artificial intelligence page comparing prominent AI researchers predictions of advancement with the almost comically different realities caused by changing societal interests and means, this assumption of linear progress with use of data and artificial intelligence completely rejects the human actors who build, train and maintain these systems. With the avalanche of scandal that social media sites like Facebook have been dealing with over data privacy and surveillance, one might imagine a plausible future in which the humanist-driven market decides that uploading experiences into the greater data flows of social media is no longer in demand, and the growth of collective big data hits an impasse of societal proportions.
So, how does this all play out? For humans, who have yet to experience Harari’s predictions, we can’t be sure, but we do have Captain Marvel to look to as a guide. But wait, no, imagine if Captain Marvel ended with one of those corny scenes where Carol Danvers wakes up and confusingly realizes that her fantastic adventure was nothing more than a dream. Maybe, before the camera fades to black, we’d see some ominous shadow dance across Captain Marvel’s laptop screen, suggesting that maybe there IS some supreme artificial intelligence out there. You’d be disappointed, huh? Well, that’s exactly what Harari does to close out Homo Deus. He literally says, “We cannot really predict the future, because technology is not deterministic… all the scenarios outlined in this book should be understood as possibilities rather than outcomes.” We love a bait and switch! I really got so worked up about picking apart his prophecies only for him to conclude his prophecies with an explicit “Not a Prophecy” guarantee, huh?