The idea that human intelligence isn't so much a fundamentally new thing, but a mix of existing things, appeals to me.
I see two features: (1) that we can intercept senses, so a car, bike or horse (or spear, axe or hammer), is neurologically an extension of ourselves; (2). (I think) abstract reasoning has an identical neurological basis to that used by rats, crows and dolphins to solve puzzles, except over abstract concepts instead of over concrete instances.
> a possible test for whether a cognitive enhancement is likely to have downsides, is the ease with which it could have occurred as a natural mutation
Our environment is different today, and traits that would have often killed their possessors 100,000's of years ago could be beneficial today. I'm thinking things like Asperger Syndrome, even Autism (which might have factors that are advantageous in a different combination, like Sickle Cell anemia - one grant malaria immunity; two kills you).
Even the earliest agriculture could have profoundly distorted the fitness function, so we may been evolving very rapidly over our recent genealogical history (rather than stagnating as is often claimed).
Better short-term memory would increase effective intelligence (in my personal experience, anyway), and it seems that evolution could improve this without harm...? Or maybe that's just von neuman bias... it could be that short-term memory is intimately integrated with intelligence - eg. each chunk in working memory is connected to everything else in long-term memory, allowing you to recognize similarities etc. You couldn't afford many buses like that.
"Better short-term memory would increase effective intelligence."
I'm not sure about that. On one psychologically administered IQ test I did really well; supposedly 99.97th percentile of incoming college students. I was 11; this was one of the final straws in my tenure in middle school. Shortly afterward, I took another test, as part of my case to get into college. Apparently, I did pretty well. In most of the sub-tests, anyway.
But then there was a digit recitation test, and a color recitation test, where you had to recite it backwards, forwards, and so on. I must has scored deep in 'idiot.' I could barely do it at all! The psychologist was astonished, and then postulated that I must be really tired, drove me to a cafe to get coffee, and skipped right over the memory tests.
On the other hand, I have a decent memory, even short term, in other areas, so maybe it's not so simple.
I'd love to have the ability to meet a group of people and have their names and faces stored in the equivalent of RAM for the party (once I leave, the names go away - unless I try hard to move them to long-term).
I'm working on developing the ability slowly - using all kinds of memory association techniques but it's definitely not natural.
I've theorized that I don't remember a name until I need to use it. I can game this with some effort by taking the time to artificially "need" to use their name by imagining their face and mentally calling their name.
I find that as soon as I find out something interesting about the person I've just met that the odds of long-term recall improve dramatically.
Context is everything when it comes to recall. I find it takes me at least 5 seconds to remember a remote office co-worker's name if they are visiting our office - by that time, they've already walked past me.
I was in another city at an event and thought I saw someone I knew from my home city. I thought 'what the heck would he be doing here? it can't be him' Later on, I emailed him and he was like 'yeah, i thought it was you. i was there too'
Reminds me of when I was simultaneously marked by the public school system as gifted and in need of special education. Some people have a hard time with the idea that there are different kinds of intelligence.
Eliezer, thank you. Your essays, without fail, introduce me to concepts I had not considered before. They get me thinking, and hopefully, thinking in a different way about the topics.
I really like his writing. He's asking questions that no one else is asking and doesn't pretend to have all (or most) answers, because they're the types of questions that take you to the edge of what your consciousness is capable of thinking.
Sometimes I amuse myself by thinking about how Overcoming Bias reads to your average religious American. It must seem like total nonsense. It's probably impossible to understand without having had a crisis of faith of some sort, at the very least.
I see two features: (1) that we can intercept senses, so a car, bike or horse (or spear, axe or hammer), is neurologically an extension of ourselves; (2). (I think) abstract reasoning has an identical neurological basis to that used by rats, crows and dolphins to solve puzzles, except over abstract concepts instead of over concrete instances.
> a possible test for whether a cognitive enhancement is likely to have downsides, is the ease with which it could have occurred as a natural mutation
Our environment is different today, and traits that would have often killed their possessors 100,000's of years ago could be beneficial today. I'm thinking things like Asperger Syndrome, even Autism (which might have factors that are advantageous in a different combination, like Sickle Cell anemia - one grant malaria immunity; two kills you).
Even the earliest agriculture could have profoundly distorted the fitness function, so we may been evolving very rapidly over our recent genealogical history (rather than stagnating as is often claimed).
Better short-term memory would increase effective intelligence (in my personal experience, anyway), and it seems that evolution could improve this without harm...? Or maybe that's just von neuman bias... it could be that short-term memory is intimately integrated with intelligence - eg. each chunk in working memory is connected to everything else in long-term memory, allowing you to recognize similarities etc. You couldn't afford many buses like that.