“This invention will produce forgetfulness in those who learn it, because they will not exercise their memory: they will trust in the external, not in themselves.” These words are not from a neuroscientist talking about artificial intelligence, nor from a politician regulating social networks. They are from Thamus, king of Egypt, who 2,300 years ago, in Plato’s ‘Phaedrus’, argued that any technology that helps remember ends up weakening. He was talking, of course, about writing. But, curiously, the arguments are so current and relevant that they could have been stated today: banning social mediaFor example. And this is the interesting thing. What was Plato’s argument? The quote, as I say, is from the end of the Phaedrus. There appears the call ‘myth of Theuth and Thamus‘: the god Theuth presents writing as a fantastic technology that would improve memory and Thamus, in contrast, responds that what will improve is forgetting. Although it is usually brought up in the context of classical disputes about whether writing is good or bad, the truth is that the good Plato’s argument is a little more subtle: what he is interested in confronting is rather the difference between internalized and practical knowledge, on the one hand; and the knowledge that, even though it is easily available (thanks to writing), has not left a mark on the subject. That is, Plato does not contest writing. He was, rather, describing a pattern: each cognitive technology reconfigures the skills we practice and those we don’t (and therefore let atrophy). ‘Cognitive offloading’. That is the ‘word’ that, from certain areas of cognitive science, is used to download mental work. They can be using notes, to-do lists, calendars, GPSs or search engines… it doesn’t matter, the phenomenon is very similar to what Plato commented on. The available evidence tells us that, in effect, there is a trade-off: Using these systems improves immediate performance (as Theuth argued), but can reduce deep learning (as de Thamus argued). And it makes sense. When we know that something will be accessible, the tendency is memorize its content less and dedicate those resources to memorizing where to find it. In other words, it changes what we do with those resources we have to try to make their use as efficient as possible. In fact, in the same way we have to recognize that this has problems (especially with content that is fundamentally important), but it also has benefits. This ‘resource release’ allows us, for example, learn new things. PlatoGPT. The issue is always very similar: new technologies trigger moral panic in society and then, with hindsight, we see if they were right or wrong. That is to say, we have been in a very long war between early-adopters and late-adopters for 2,300 years. Now it’s up to artificial intelligence and Plato’s reflections are good. Above all, because they help us see AI as something that goes beyond “a tool”: it is a complete system of incentives that pushes us to improve certain skills and atrophy others. The key is whether those skills that we atrophy are necessary for something else. “Put doors on the field.” A few years ago, the philosopher Antonio Diéguez visited us and explained that the idea that technologists repeat so much that “you can’t put doors on the field” was somewhat problematic. Of course you can. It has a cost, it is true; But there is also a risk of being uncritical with all the technologies that knock on our door. We have learned it the hard way in recent years. We live in strange times when nobody knows anything about the social impact of new technologies. But what is clear is that this should not confuse us and make us believe that we cannot know anything about it. Yes we can, yes we can. It’s more. As Plato said, it is our obligation to know. Image | Raphael / Robin Worral In Xataka | Why being a teenager has always been shit and in the age of social media even more
