Since you just use the term “evolution”, it’s hard not to say yes; especially when you use the qualifier “somehow” which can cover a lot off ground. I hesitate because I have a bit of a peeve concerning the loose usage of the term evolution.
So, on one hand we have the theory of evolution through natural selection. Darwin himself was not very fond of that name. “Evolution” comes from the Latin “evolutio” meaning “unrolling”. This suggests that the process has an aim or pattern or progression but Darwin believed that speciation proceeded by random drift toward whatever “worked” in a particular environment: if simplicity improved survivability then organisms would become simpler, if complexity improved survivability then organisms would become more complex. The pillars of the theory of evolution by natural selection are: heritability (children resemble their parents), random variability (children are not exact copies of their parents but the differences are random), and survivability/selective pressure (not all offspring will survive and heritable traits have an effect on survival rates). Darwin proposed that these mechanisms together were sufficient to explain how all life could have come from a single common ancestor. This is the evolution that “scientists believe in” and creationists get all excited about.
On the other hand we have the more general sense of evolution meaning development or change in general. There is “Lamarckian evolution” and “directed evolution”. These are all different but get mixed together in common speech.
Thank you for the thoughtful comment suman! You’re right to point out the distinction, “evolution” gets thrown around so casually that the scientific meaning and the everyday meaning often get blurred. In the piece, I was using it in the broader, non-Darwinian sense: systems changing, adapting, reorganizing themselves over time. Not evolution by natural selection, and definitely not the idea of traits being “aimed” toward anything.
When I say “somehow,” it’s because whatever is happening with intelligence right now doesn’t map cleanly onto the Darwin model. There’s no heritability, no reproduction, no selective pressure in the biological sense. What we’re seeing is a different kind of development, more like iterative design and feedback loops than natural selection.
So I agree with your peeve. The scientific meaning of evolution is precise, and it matters. In the article I was talking about change, not Darwinian mechanism. If anything, we probably need a new word for what’s happening with intelligence right now, because it doesn’t fit the old categories very well.
You state "Philosopher Ludwig Wittgenstein ended his Tractatus with the line: “Whereof one cannot speak, thereof one must be silent.” He meant that language reaches its limits long before reality does. Understanding begins where description fails."
That’s a great question, thanks Malcom! Ofcourse no, it can’t eat a peach. But that’s what makes the question interesting. AI can describe the taste of a peach perfectly, model its texture, even simulate the memory of eating one. But it will never taste it. That gap between representation and experience is the line Wittgenstein was pointing to. AI lives entirely on the describable side of that line.
Since you just use the term “evolution”, it’s hard not to say yes; especially when you use the qualifier “somehow” which can cover a lot off ground. I hesitate because I have a bit of a peeve concerning the loose usage of the term evolution.
So, on one hand we have the theory of evolution through natural selection. Darwin himself was not very fond of that name. “Evolution” comes from the Latin “evolutio” meaning “unrolling”. This suggests that the process has an aim or pattern or progression but Darwin believed that speciation proceeded by random drift toward whatever “worked” in a particular environment: if simplicity improved survivability then organisms would become simpler, if complexity improved survivability then organisms would become more complex. The pillars of the theory of evolution by natural selection are: heritability (children resemble their parents), random variability (children are not exact copies of their parents but the differences are random), and survivability/selective pressure (not all offspring will survive and heritable traits have an effect on survival rates). Darwin proposed that these mechanisms together were sufficient to explain how all life could have come from a single common ancestor. This is the evolution that “scientists believe in” and creationists get all excited about.
On the other hand we have the more general sense of evolution meaning development or change in general. There is “Lamarckian evolution” and “directed evolution”. These are all different but get mixed together in common speech.
Thank you for the thoughtful comment suman! You’re right to point out the distinction, “evolution” gets thrown around so casually that the scientific meaning and the everyday meaning often get blurred. In the piece, I was using it in the broader, non-Darwinian sense: systems changing, adapting, reorganizing themselves over time. Not evolution by natural selection, and definitely not the idea of traits being “aimed” toward anything.
When I say “somehow,” it’s because whatever is happening with intelligence right now doesn’t map cleanly onto the Darwin model. There’s no heritability, no reproduction, no selective pressure in the biological sense. What we’re seeing is a different kind of development, more like iterative design and feedback loops than natural selection.
So I agree with your peeve. The scientific meaning of evolution is precise, and it matters. In the article I was talking about change, not Darwinian mechanism. If anything, we probably need a new word for what’s happening with intelligence right now, because it doesn’t fit the old categories very well.
Hbvvcf
Heloo
Its me
Hello abbas!
You state "Philosopher Ludwig Wittgenstein ended his Tractatus with the line: “Whereof one cannot speak, thereof one must be silent.” He meant that language reaches its limits long before reality does. Understanding begins where description fails."
AI reality is composed of 1 and 0.
Can it eat a peach?
Its me
That’s a great question, thanks Malcom! Ofcourse no, it can’t eat a peach. But that’s what makes the question interesting. AI can describe the taste of a peach perfectly, model its texture, even simulate the memory of eating one. But it will never taste it. That gap between representation and experience is the line Wittgenstein was pointing to. AI lives entirely on the describable side of that line.