Robots have been performing repetitive factory tasks for decades, with superhuman precision. Large language models are new, and definitely less precise, but more flexible. So what happens when these technologies combine? What would this machine look like? What are the real world implications?
Table of contents
Open Table of contents
The Architecture of the Synthetic Man
Let’s talk technicals: If we’re headed towards this uncertain future, we should at least be able to dissect what these systems would look like so we can steer them towards good. So let’s talk technicals.
Large language models as cognition: LLMs could provide the mental layer for these robotic systems. Reasoning, planning, communication, and proper context could create a synthetic brain. But like humans, a brain in a vaccuum doesn’t do much. So there must be a nervous system equivalent.
Sensors as the distributed nervous system: Cameras, layers of thin, skin like sensor webs, microphones, a fluid based balance system, gas sensors. If concerted together, it is feasible these things could combine into something more powerful.
Diving into the LLM Based Brain
When Phineas Gage got a railroad spike through his brain: Let me tell you a quick story.. Phineas Gage, was a railroad construction foreman in the 1800s. One day, he had a really bad day at work. A iron railroad spike blasted through his head, and obliterated the frontal lobe of his brain.
.gif){kind=link}
After the accident, Gage’s personality was said to be completely different, but he survived and lived for quite some time after the accident. This case became a big piece of evidence in proving the localization of the brain, where certain areas of the brain perform certain functions. This is the difference between your lizard brain and frontal lobe.
LLMs as a part of the robotic brain: Given humans have a brain with localized functions, I am guessing robots will have a similar architecture. Just as humans have Broca’s area that controls speech, robots will have an LLM embed. This will be a part of the brain, rather than the whole of the brain. So by my estimation there are more developments needed to really achieve a synthetic human.
A New Technological Era
What the future could hold: Star Wars. C3P0. The Terminator. A robot that folds your laundry. People falling in love with robots. Clone wars. Robots curing cancer. Discovering a new endless source of energy. The possibilities are endless. Endlessly awe inspiring, terrifying, helpful, descrutive, and creative.
Boom or doom or both?: Sam Altman, CEO of OpenAI once said “A.I. will probably most likely lead to the end of the world, but in the meantime, there’ll be great companies.”. He also said “People are using it to create amazing things. If we could see what each of us can do 10 or 20 years in the future, it would astonish us today.”
So AI curing cancer is on the table. AI washing your dishes perfectly is on the table. AI eliminating car crashes is on the table. AI creating a deadly pathogen is on the table. AI militia are on the table. Anything and everything is possible. Murphy’s Law. Yhprum’s Law. Murphy’s and Yhprum’s Law.
The Tower of Babel & AI Parallels
Whether you’re religious or not, the Tower of Babel is one of the oldest parables known to man, and it has striking parallels to the AI race. The story begins with the whole world having one language, which I think is functionally true today, with modern LLM technology.
Then God saw this, and since the people were so ambitious and nothing would be impossible for them, he made it so they couldn’t speak to one another, and so they stopped building the tower. God stopped them, presumably because it would lead to evil. The reasoning is open to interpretation, of course. But the parallels are interesting. Is AI our modern day Tower of Babel?