March 2023

ChatGPT has taken the world by storm. It has demonstrated to the public for the first time that Large Language Models are able to hold a human-like assistant conversation. It can conceivably pass the Turing test most of the time. And there is no apparent end in sight for the improvement of LLMs.

GPTs will be in everything in 5 years. Their ubiquitous understanding of textual input allows for all kinds of different feedback loop setups where GPT is able to interact with external APIs in complex ways. You will have a GPT in your new Tesla. You will have a GPT in your phone. Your computer. Probably in your microwave in 10 years. At this point it is almost certain that we will live in the natural-language-controlled science fiction within the decade. Thanks to Alexas paving the way, the technology exists, and it is cheap.

That’s not to say that GPTs are anything like human intelligence. It still lacks and will forever lack the ability to process its environment in the way humans do. GPTs are very much like the ship computer in Star Trek, and very much unlike the android Data. Make no mistake; pure GPT ability is hard-capped.

There is one very interesting thing about the nature of GPT though. We as humans have always prized our intelligence as our defining factor. That is what separates us from animals. Now, it seems, we have solved that intelligence in some manner. The part which we believed to be the hard one is done.

It is becoming more and more clear that we don’t have language because we’re superior to the rest of the animal kingdom; we are superior because we have language. Now we have the power to bestow that ability to anything we build. And we will. As we reach the limit of what natural language alone can achieve, we begin turning our focus to complementary systems.

It seems only a matter of time until the Second Renaissance. I, for myself, am both thrilled and scared. We’ve reached escape velocity now. Welcome to the sci-fi movie that is the 2020s. Bid your final farewells to the Old World.


Thanks to Benjamin, Hessu, Jari & Itamar for reading different drafts of this essay.