Intellicology, a new field of science
8th November 2024
As I said in my blog post 4 months ago, I take umbrage with the "artificial" in Artificial Intelligence.
Because if the industry achieves its goals of "AGI" then these machines will match or surpass human ability in all domains. And it's hard to justify calling it artificial if it's smarter than you.
Furthermore it's not clear that humans are even a good example of intelligence. Most people will tell you that animals are not very intelligent. And since the only difference between an animal brain and a human brain is it being slightly bigger, so it follows that humans are not very intelligent either.
So really the field of AI should actually be considered the study of intelligence in general.
Now what is the study of intelligence called? It turns out it doesn't have a name, so we have to make one up. If physics is the study of the physical, chemistry is the study of chemicals and biology is the study of life, then I guess the study of intelligence must be Intellics? Intellisty? Intellogy?
I thought it would be fitting to ask an AI, so I asked ChatGPT. It suggested Intellicology, so we'll go with that.
What's happening
At the moment all the focus is on building "AGI". Of the top 10 biggest companies in the world, 8 of them are tech companies, of which most of them are investing tens of billions into AI.
With the amount of resources being poured into the endeavour at this point, I would argue that, unless it's actually impossible to make AGI, it's inevitable.
If it is, in-fact, impossible to make AGI then that will be fascinating. I suppose we will have found the soul.
But I think more realistically AGI is inevitable.
If:
- you accept the inevitability that someone will build an AGI, and
- you accept that at that point it will cease being "artificial", then
- this means humans can build an Intelligence
What does it mean to build an Intelligence? Is it like building an engine?
To build an internal combustion engine from scratch, there are three main areas:
- theory: the physics and chemistry
- design: mechanical design and material science
- practice: the knowledge to build and maintain
Likewise, in trying to build Intelligence. And there are theoretical, design and practical aspects to it.
And there is no better way to understand something than to build it.
So I argue that the endeavour to build Intelligence is, in-fact, the same as the study of Intellicology. If we can build an Intelligence then we will understand the nature of intelligence itself.
Thus the work that the tech companies are doing (the most powerful organisations in the world), is actually laying the foundations of the field of Intellicology.
A new Field of Science
Obviously we already had neuroscience and psychology, but these are really the study of animal intelligence. Thus these will be sub-fields within Intellicology.
The real way to understand something is to build it so the main field will be Intellicology.
To witness the birth of a new field of science is an incredible thing.
In AI, business as usual is ground-breaking and mind-bending. Just in the past 4 months two genuinely novel new AI capabilities have been demonstrated:
- ChatGPT advanced voice is the first time the world saw an AI that can have a true conversation with a human.
- OpenAI's new line of reasoning models are a new architecture which has such radically better performance that in some benchmarks went from 20% to 80%
To watch science being built before our eyes is incredible and fascinating.