AGI Intelligence is Different. In a Good Way
Not omniscient, but much more objective and rational. And not competing with us.
If you’re not sure what AGI is, check out ‘What is (real) AGI?’ and maybe ‘Imagine’. It is not ChatGPT or other Generative AI!
By definition, an AGI can autonomously learn to master any new cognitive task as easily as a smart, educated human. And it can easily adapt to changing circumstances or requirements. With no more time or training data than a human.
That’s the baseline.
Naturally, AGI being a computer program it has many advantages over our biological cognition.
Potentially having unlimited photographic memory of all inputs and actions
Around-the-clock full focus and concentration
Instantaneous access to everything online
The ability to easily exchange knowledge and skills with other AGIs
Inherent advanced logic reasoning ability
Built-in tools like calculators, dictionaries, spreadsheets, databases, etc.
Just like us, they are excellent tool users, but unlike for us, these tools can be deeply integrated with its cognition.
However, just like us, they are not omniscient or infallible. Reality is far too complex, chaotic, and inscrutable to be fully knowable or predictable.
What they will not have is our rich experience of being — being in a biological body, our unique consciousness, and the wide range of emotions we experience.
AGIs will have (to have) a self-concept, be self-aware (a functional type of consciousness), and have metacognitive signals for things like certainty, boredom, confusion, etc., things one could refer to as ‘cognitive emotions’.
Humans are inherently motivated by their ‘selfish genes’, driven to pursue survival and reproduction. AGIs will not have such drives. AGIs are built and optimized specifically to help us. Any designs that drift away from this core goal will lose out to superior implementations. Unlike us, an AGI does not have its own agenda or goals.
All of an AGI’s advantages culminate in its ability to learn to reliably perform a wide variety of tasks and to dependably discover objective knowledge or truths. This will be of tremendous help not only to create material abundance for everyone, but also to improve our individual ability to think more clearly. A dedicated personal AGI (you own it, and it serves your purpose, not some megacorporation’s) can be like a little angel on your shoulder to help you make better choices and to flourish.
Hi Peter, I enjoy reading your articles. I agree with many of your points you make.
We need to be better informed so that we can make better decisions. Tokenized LLM’s are NOT the answer https://www.youtube.com/watch?v=Cis57hC3KcM
We need help with ethics and morality.
When it comes to AGI maybe we should drop the “A” and the “G” should stand for ~ ‘Genuine’
My thoughts here —
https://payhip.com/StandOUTStudio/blog/ai-value-vs-viral/evolution-and-baking-bread
James (the one from SA) :)
The hard problem of consciousness is far from solved, so I don't think we can yet judge whether our AIs will have our rich sense or not. But I do think, if the AI can be trusted to be honest, we can find out simply by asking it "Are you aware in the same fundamental way that humans are?" And it will be able to tell us. It also seems that once the AI is smarter than us, and escapes out into the digital wilderness, the same Darwinian forces will act on AIs, determining at amplified speeds, which ones survive.