69 years ago Alan Turing came up with the idea of the Turing test to tell when an A.I. Has intelligence indistinguishable from our own.
This morning I began to think about, if enough computational power was available and the code was written, what differences would there be between machines and humans.
I was first thinking that machines would purely be a logical driven entity and we wouldn’t be. Although we are logical driven aren’t we? Yes, fine your logic may be slightly different than mine but, to yourself, your decisions and choices are driven by your logical understanding of a given situation. If you knew enough facts about someone you can almost predict their decisions due to them being logically driven.
So OK, our logical systems could be indistinguishable, but what about our ability to learn? We learn in a variety of ways and as computer science evolves A.I. Is learning in more and more ways also. Pattern recognition, yup we both have it. Image recognition yup again machine learning and humans both are capable. Predicting the outcomes of a given situation which you have never been in before? Hmm this one is a little trickier. It boils down to past experiences and being able to process these memories in a similar sort of way to pattern recognition. So given enough time it’s easy to see that machines will have the ability to learn in just as many ways as we can.
What is the difference then? Emotions?! Well no because our emotions are based on stimuli going through a logical process to determine how we should react. This is why sometimes we experience emotions despite having knowledge to the contrary, such as jealousy. So if we were on a level playing field with A.I. And machines had just as much processing power as we do then I believe it’s plausible they would exhibit emotion dependant upon their experience and objectives.
But what can the difference be then? I know!!! Imperfection! It obviously has to be our flaws. After all humans make mistakes. It’s one way that we learn. Well I spent a few moments thinking about this and if machines had as much storage as our brains and processed as much information as quick as we do, well there would no doubt be read/write errors. Even now, if you download a few gigs then there is a possibility the file may be corrupt because of a read / write error. So in essence machines could well have flaws too. Even if you think “nah there could be error checking to make sure it’s all alright”. Well get two identical computers off the same production line. They will act slightly differently due to nano scale differences in the hardware. Flaws so small that it doesn’t stop the hardware running be affects it’s behaviour. This is why two identical computers can develop unique behaviours.
So A.I. can have flaws too but there has to be a difference doesn’t there? There is a pretty basic one which is biological vs manufactured hardware. But that still doesn’t answer whether there would be behavioural differences between humans and machines.
Now I’m thinking the flaw lays with myself. I think we have all got in the mindset where there would be a difference. The kind of narcissistic attitude why we once believed the galaxy revolved around earth.
Given the right programming I don’t believe there would be a difference. I believe in the future it would be a matter of creating the right objectives to create the correct motivations which would impact the behaviour of the machines.
I know there isn’t really much of a conclusion here it’s just a quick thought I had on this sunny Sunday.