By Santiago Hunt
Cover image by 0fjd125gk87.
I’ve been playing a game for the last 2 years called Slay the Spire. It’s a casual card game, which has the beauty that it can be enjoyed in both long and short intervals. During my time playing it, I started following on Twitter a guy called Jorbs (@JoINrbs), who is not only one of the world’s best StS players but also a professional online poker player and strategy game streamer. I would have expected videogames to be one of the last corners reached by the “ChatGPT” craze. So imagine my surprise when I read this:
Nowadays, my Twitter feed has been co-opted by Chat GPT, or more accurately LLM (Large Language Model) comments. They go from amazement, to skepticism, to technophobia, to “We’re all going to die”. The amount of generative AI startups that have spawned out of nowhere is proof to this.
Now, before we go on, one important thing: This is not a “CHAT GPT WILL TAKE ALL OF OUR JOBS” post. I don’t think that will be the case, but more importantly, that is not the point I want to cover. This is not a post about machines. It’s about humans, and the way humans relate to other humans.
For those of us who remember the Pre-Internet world, there was something truly magical about the world wide web of the early days. You could instantly connect with people all over the world. People who shared common interests, passions or just strangers you met in random chat rooms. People.
Most people love to bash on social media, but I’m a firm believer that it has been a net positive for the world. Why? Because it has allowed to extend and expand our connections with other people, be it friends, family or people we would have otherwise never met nor connected with.
The common thread is that both Web 1.0 and 2.0 have expanded and empowered the Human-Human connection. Machines were a medium for that connectivity, but not an end in themselves.
This is changing.
The God in the machine
We have always acquired knowledge from other Humans. Throughout history, we have learned from other humans through direct copying (mimicking) or through oral and written transmission of knowledge. Even today, the algorithms that rule our information consumption are nothing more than sorting and ranking protocols. The knowledge we are ultimately consuming is human generated. We learn and we grow with each other, and thanks to each other.
But this changes with Chat GPT: there is now an AI overlay in the knowledge we access. For the first time in human history, we are retrieving a corpus of knowledge at scale that has an alien, artificial coating to it.
Arguably though, that’s not how LLMs like GPT work. A simple explanation of how they truly function, best exemplified by Yann Le Cunn (Meta’ chief AI officer).
It is clear that LLMs are great at generation, but less so at answering questions. To quote David Deutsch “…in many ways, AI is the exact opposite of AGI…”. As an example, I asked ChatGPT to assist me with card choices while playing Slay the Spire. The decisions it made were lousy at best, if not outright bad. Social media is full of erratic and blatantly mistaken takes from LLMs (Ie. Hallucinations).
But the point is not about the accuracy or reasoning capability of LLMs. It’s about how we are inserting an artificial layer into how humans exchange knowledge with other humans. This layer may potentially become a gatekeeper in our knowledge sharing. Taken into extremes, it is not a waypoint, but an endpoint.
One side of that coin is the sheer awe and magic that these experiences provide (be it ChatGPT, Bard, or others). Their brute force makes people automatically wander about the future of work. However, the flip side is a sense of emptiness. Lack of attribution. Limited feedback. No connection.
Several works of fiction explore the impact of AI on society. The Matrix, 2001 Space Odyssey, Her, Terminator, Hyperion…the list is a long one. These stories typically convey similar learnings and warnings.
But I think there’s a far better story to help us glimpse what the “God in the machine” can mean for humanity. Curiously, it is the first story of all. The Book of Genesis.
The Tree of Knowledge
In Judeochristian tradition, the book of Genesis tells the story about how God created the World, and how Adam and Eve defied God by trying the forbidden fruit from the Tree of Knowledge of Good and Evil.
I find the Book of Genesis to be fascinating. It is widely mocked and criticized because of how childish it seems, and by how glaringly wrong it is when compared to Darwinian evolution theories. I disagree. Not only do I think the Book of Genesis is a fantastic allegory of Darwinian evolution (this probably requires a post on its own), but more importantly, it is far from childish. It is premonitory.
Does this mean that if we eat from the fruit of the “AI knowledge tree” we’ll be damned and we’ll die? Not exactly.
The Book of Genesis story is, among other things, one of coming of age. Innocence, once lost, cannot be brought back. Once humanity “ate” from the Tree of Knowledge, it set itself apart from the animals and plants with which it shared the Garden of Eden. Knowledge made humanity a different type of being. Knowledge evolved humanity.
Now, once again and for the first time in, who knows how long, we are at a pivotal moment. A moment where we are modifying our relationship with knowledge. And because of this, a moment that changes how we Humans…connect with other Humans.
Today, connection with machines is becoming not only a means, but can now also be an end in itself. That is a fundamental change in how humanity lives. A change in how we connect with each other.
We have chosen to, once again, follow the serpent. Eat from the Tree of Knowledge. And innocence, once lost, cannot be brought back.
“Then the eyes of both of them were opened, and they realized they were naked.” Genesis 3:7.
Click here to read more thoughts on AI: Skills for the 21st century. .