The Future Role of Humans

My perspective on ai has shifted quite a bit over these past few years. The trajectory can be summarized as follows.

2024: I thought the LLM’s were cool, but to replace highly analytic work like physics, I thought there needed to be new models which may take many more years, maybe a decade.

2025: I realized the LLM’s really were becoming smart enough to do complex and analytic work, but I didn’t believe there would be such dire economic implications, in part because I thought there may be some sort of limit on how intelligent the ai could get.

2026: I realized there was no limit, at least not any time soon. I had to seriously consider the economic implications.

Naturally, over these past few months, I’ve spent quite a bit of time thinking about how ai is going to change everything. It’s made me reevaluate my physics work. It seems it’s only a matter of time when ai is better at every form of physics explanation than me (maybe that time has already come). This is the some of the most meaningful work I had ever experienced, and naturally it was a little scary to think ai could strip it all away from me, that the demand that existed for the kinds of explanations I produced may suddenly dry up in favor of ai generated explanations.

Everywhere I looked, it seemed career trajectories I’ve considered may not exist because of ai. Of course, when we look back in the future, I’m sure there will be many things people these days have gotten wrong about the impact of ai, and maybe it’s all overblown. But it seems an unstable position to be banking on everything being overblown. I spend a lot of my time these days trying to prepare for a future in which ai surpasses a human in every way possible, and there is essentially no need for a human to do most forms of knowledge work (the current rate of ai growth makes me feel this may be rather soon).

Naturally, then, I try to look far into the future, thinking about the post-ai-revolution world, in order to determine what will still matter when ai can do everything.

Previously, work in many ways was dehumanizing. It stripped humans of their individual qualities, evaluating them based on certain, impersonal objective metrics, such as lines of code per hour and dollars generated per year. Ai handling all such work could potentially have the effect of restoring humanity to each person. This is not guaranteed, but I see no reason why this could not be achieved.

It is important now more than ever to embrace one’s individual and eccentric qualities. Such is the aim of this website. I seek to hone and share all my such idiosyncrasies. If this website can benefit even just one person, one real human out there in the world, then I will consider that a job well done for life.

Maybe everyone can become an artist.

Right now, at least, society only really values the art of those at the very top, many of whom are dead. This is not so encouraging, as it suggests only a few very talented people can be artists. However, this could potentially change. I think about friendship. Almost everyone is able to have friends, even if they may not be at the top of the “best friends in the world” list. In a similar way, maybe the human connection to the artist will become important enough that everyone may have a place in the new world. Art may become like friendship. The enjoyer of the art may want to have a special relationship with the artist, a relationship only possible locally (i.e. not winner take all one artist producing all the art).

Why do I think art will still survive? After all, ai can make beautiful novels and paintings (maybe they’re only mediocre now, but I see no reason why they cannot become truly beautiful in the future). I think about what has survived the test of time. Humanity has gone through numerous technological revolutions. The only works produced by ancient humans which we still care about are their writings and their art. I think that will never change. No one uses a Mesopotamian wheel. No one uses their farming equipment, or the materials they built their housing with (at least, not most of the advanced world). I don’t see this changing anytime soon. Even though writing stories was one of the first things LLM’s were able to do, I actually see writers as one of the last to be replaced by ai (as in, maybe never). I was reading 100 years of solitude, and I thought about how unhappy I was to be reading the translated version, almost frustrated with the translator for getting in between my heart-to-heart conversation with Gabriel García Márquez. Writing is much more than having beautiful words, it’s about creating a link between the author and the reader. I hope to create such a connection in writing my articles and reflections, to convey if nothing else my own personal ideas about physics and the world. Though I can never hear what you, the reader, say, I still very much feel this strong connection.