Dr. Martin Hash Podcast

We are in another era of snake oil salesmen, most notable is the artificial intelligence craze. The hype has reached historic proportions of claims of it being an existential threat, approaching the heights of climate alarmism. The nature of the danger, we’re told, is that A.I.s will soon become smarter than humans, then take over the world. However, the dystopian, science fiction sounding claim is highly improbable because A.I.s have no intelligence, they simply repeat the most common thing everyone says. For example, people assume A.I.s can write but A.I.s can only do tropes, meaning they can only parrot back what already exists, with variations on the theme; they have no intuition nor creativity. Writing A.I.s have no chance of reaching the level of an engaging author because the quality of the data they’re built from is average at best. Even if only good writers are sampled, assuming there’s enough available, one of the elements of good fiction is originality, and by its very nature, A.I.s cannot be original, only derivative. Similarly, A.I.s can’t do research so their non-fiction writing would just be a rehash of existing work. On the plus side, art-generation A.I.s, perform well because the images they learn from are photographs, and talented artists; and most-importantly, subjective. A.I.s can do rote tasks, like driving cars, sorting parts, or packing boxes. They might even be able to do yard work or harvest crops. It’s unclear if A.I.s can serve in the role of doctors because doctors have entered the realm of social workers, where person-to-person contact is the product, again the exact opposite of an A.I. Maybe A.I. “intelligence” can rise to the level of being a companion to the lonely because most people communicate in drivel, just like A.I.s.