Intelligence is a mystery

Intelligence is a mystery — the more you learn, the more questions you have. Scientists, philosophers, and engineers all have different opinions about the phenomena that are both esoteric to some and seemingly simple to others. How do we build AI without knowing what the I is?…

Intelligence might mean solving problems — applying skills and experience to solving specific or general problems. I can make a sandwich because I learned the skill. Within that skill, there are finer skills, like how to pick up and use tools.

Intelligence might mean identifying problems — do I wait for instruction or am I able to identify problems in my environment? I notice that the kids are behaving strangely, they’re likely hungry. I should make some sandwiches.

Intelligence might mean learning from experience — the reason I noticed the problem of hungry kids, is because I’ve learned from my past experiences where that specific behaviour meant that they’re hungry…

…I also learned and perfected the skill of making sandwiches from many attempts of buttering bread, adding fillings, and even toasting it!

A naive understanding of intelligence could be boiled down to pattern recognition — understanding inputs from your senses (visual, auditory, sensory, etc.) and acting based on them, for your benefit.

And then we have the concept of consciousness and self-awareness — which might just be the aggregated feeling of our brains regulating our body’s internal organs and interacting with the world around us?

An interesting perspective from @fchollet describes intelligence as skill acquisition efficiency. As humans, we’re pretty darn good at this.

Intelligence is not unique to us humans though. We observe intelligence all around us in individual organisms like dogs and insects, and groups of them — like the ways birds flock, and ants form colonies.

This insightful video by @Kurz_Gesagt visualises an approach to looking at different levels of intelligence and the various concepts involved.

https://www.youtube.com/watch?v=ck4RGeoHFko

Computational power is the main factor limiting exploration in creating machines with general intelligence and perhaps super intelligence. Humans are really efficient biological computers. @waitbutwhy explores this concept deeply in this series.

Although the idea of intelligence is mysterious, we can use theories of how we understand it to solve difficult problems while we interrogate it further — and that in itself is an expression of intelligence.

If you enjoyed this thread, check out my book: http://bit.ly/gaia-book, consider following me for more, or join my mailing list for infrequent knowledge drops in your inbox: https://rhurbans.com/subscribe.

Tech, people, and ideas. Author of Grokking AI Algorithms with @ManningBooks. Solutions architect​ at @Entelect. Founder at @viszentech and @ProlificIdea.