[distances]

You may want to be aware that in western groups any talk about national or racial intelligence is usually considered pure racism. I assume it's not like that in India; I've also had a short discussion about this earlier with an Indian colleague when the topic of intelligence somehow came up.

[cko]

If there is some sort of ethnic basis for differences in intelligence (independent of environmental factors such as poverty, culture, etc.) what would be the practical implications of this data?

It seems to me that in the future, economic value will be correlated with the ability to work with increasingly abstract things (which I assume IQ is trying to measure).

Wouldn't the natural tendency to value human worth based on their IQ? Of course not, you might say. But imagine there's some future with a permanent underclass of lower IQ people, subsidized by very high marginal tax rates / UBI. I can see our human nature justify some sort of social Darwinist policies.

Edit: what a weird tangent this thread has taken.

[ben_w]

That’s assuming future values are current values.

If we keep improving AI at current rates, two things can happen at the same time:

1. The cost of running a full-brain simulation puts an upper limit on someone’s IQ-based pay — cost of power vs salary. This is something I would equate to slavery; I don’t like it, yet I expect it to happen.

2. Fully-synthetic AGI with little to no biological duplication (more like AlphaZero then Blue Brain) is probably going to equal the performance of low-IQ humans before high-IQ humans, and if it can do that for less than the cost of 2000 kcal/day then low-IQ humans can only be employed for non-intelligence-based tasks.

But that might just mean we mostly revert to manual labour jobs. I can’t forecast this — I’m not a psychologist, I don’t grok politics, and I am aware that I don’t think the way normal people think.