LessWrong 2.0 Reader

Would you say the same of most other class subjects?

I was homeschooled and then studied math education, so I'm not sure. But my passing impression is (a) yes, it applies to most methods of teaching in schools regardless of subject; but (b) math taught this way is particularly damaging.

I want to emphasize that this is my impression. I'm also not entirely sure why math seems to be more damaging. I have guesses. I just observe that e.g. literature hatred or music phobia aren't nearly as prevalent as math trauma is. Best as I can tell.

I ask because, with the exceptions of reading and persuasive writing, I don't think that any conventional school subject is more applicable to the average person's life than grade-school math.

Yes, people can get through life with an astonishing ignorance of mathematics, but they can get through life with an even more astonishing ignorance of social studies, literature, and the sciences.

Well, sure. But people will also pick up the math they need as they need it for the most part. That's true of most subjects really.

I didn't learn to read in school. I went to kindergarten before being homeschooled, and they were teaching us the alphabet and some basic words, but I could already read books by then. I learned to read because I wanted to read.

There's something very weird in our cultural groundwater around what teaching is. It's like we start with a prescription of subjects and then default to coercion to get students to "know" those subjects. Why? If it's relevant to their lives, we could learn to point out the connection in a way that feels alive to them. If we can't do that, then what makes us so sure that it's relevant for them?

Do you have a different philosophy of education, a different ranking of subjects' importance, or something else?

Yeah I do. I think the most imporant function of widespread education is to make good citizens. Which is to say, children put through an education system need to come out of it better able to engage with the system that runs their civilization, including the education process for the next generation.

In the United States, I think that puts civics as the most important subject. It's really key that citizens understand how their government works, what the checks and balances are, how jury nullification works, what forms of corruption actually do arise even within the current system, etc. Otherwise they don't know how to participate in the government that's supposedly "by the people, for the people". This is vastly more important than learning math they don't naturally pick up in their day-to-day lives.

I think the two things you named are really good though. I wish public education had those as real goals! That'd be nice. I think (1) happens sort of despite the education methods, and (2) happens more through other cultural channels than it does through formal public education. Just my impression.

Oh, it very possibly is the wrongest part of the piece!  I put it in the original workshop draft as I was running out of time and wanted to provoke debate.

A brief gesture at a sketch of the intuition:  imagine a different, crueler world, where there were orders of magnitude more nation-states, but at the start only a few nuclear powers, like in our world, with a 1950s-level tech base.  If the few nuclear powers want to keep control, they'll have to divert huge chunks of their breeder reactors' output to pre-emptively nuking any site in the many many non-nuclear-club states that could be an arms program to prevent breakouts, then any of the nuclear powers would have to wait a fairly long time to assemble an arms stockpile sufficient to launch a Project Orion into space.  

Worth looking at the top ten holdings of these, to make sure you know what you are buying, and that they are sensible allocations:

• SMH - VanEck Semiconductor ETF
  • 22% Nvidia
  • 13% Taiwan Semiconductor Manufacturing
  • 8% Broadcom
  • 5% AMD
• QQQ
  • 9% AAPL
  • 8% NVDA
  • 8% MSFT
  • 5% Broadcom

It might be worth noting that it can be good to prefer voting shares, held directly. For example, GOOG shares have no voting rights to Google, but GOOGL shares do. There are some scenarios where having control, rather than ownership/profit, could be important.

Is it a matter of adding a standard paragraph to NIH grants? Yes. That's what I was thinking of.

NVDA's value is primarily in their architectural IP and CUDA ecosystem. In an AGI scenario, these could potentially be worked around or become obsolete.

This idea was mentioned by Paul Christiano in one of his podcast appearances, iirc.

If you follow standard DEI criteria, I'm commenting on LessWrong; I don't do "standard."😉

More seriously, I apologize. I should have clarified what I meant by diversity. In particular, I mean that diverse groups are spread out in a parsimonious description space.

A pretty detailed example

As a concrete example of one understanding that would match my idea of diversity, consider some very high-dimensional space representing available people who can also do the work measured on as many axes as you can use to characterize them (characteristics of mind, body, experiences, etc.) and reduced by a technique to cause the remaining dimensions to give little mutual information about one another. Define a "diversity-growing procedure" for adding members to be one that chooses new members farthest away from the current subset. The more ways a diversity-growing procedure would choose a particular group, and the fewer exceptions that need to be made to the procedure to end up with that group, the more diverse the group.

Making an instance of this concrete example, imagine that our parsimonious space is 2D and that candidates are at all integer intersections of (0,1,2,3,4) x (0,1,2,3,4). If we choose candidates (2,2),(0,0), and (2,4), how diverse is that group? If you start with (0,0), the farthest away is (4,4), so you'll need to make an exception to add (2,4) (the farthest actually in the group). From (2,4), the farthest away is (0,4); once again, we need an exception to add (2,2). If we start with (2,2), then (0,0) is one of the farthest away, but we need an exception to add (2,4). The sequence (2,4) -> (0,0) -> (2,2) requires one exception. So, we have three ways with five exceptions. (I may have gotten some of this wrong since I did it in my head, but I think this gives the picture.)

This is just one example. Many ways to define diversity match my intuition of spread in a parsimonious description space.

A note on parsimony

Since we're looking for diversity to help us in a particular context, we should choose dimensions that predict differences in that context. For example, a characteristic like "ability to roll your tongue" is probably less predictive of behavior in a research environment than gender, so we might want to down-weight it. However, we don't have good models of what matters yet, so it might be hubris to down-weight characteristics until we know they don't matter for covering research hypothesis space because we've determined the effects by looking at groups actually doing research.

Oops! That was a mistake. I guess we're re-examining these ones tonight (they are pretty good ones). I have a spreadsheet for tracking this sort of thing, I will make some adjustment there to avoid this mistake again.

ASI turns out to take longer than you might think; it doesn’t arrive until 2037 or so. So far, this is the part of the scenario that's gotten the most pushback.

Uhhh, yeah. 10 years between highly profitable and capitlized upon AGI, with lots of hardware and compute put towards it, and geopolitical and economic reasons for racing...

I can't fathom it. I don't see what barrier at near-human-intelligence is holding back futher advancement.

I'm quite confident that if we had the ability to scale up arbitrary portions of a human brain (e.g. math area and it's most closely associated parietal and pre-frontal cortex areas), we'd create a smarter human than had ever before existed basically overnight. Why wouldn't this be the case for a human-equivalent AGI system? Bandwidth bottlenecks? Nearly no returns to further scaling for some arbitrary reason?

Seems like you should prioritize making a post about how this could be a non-trivial possibility, because I just feel confused at the concept.

What would the title be?

If the number of neurons is so important, what about elephants or whales? Perhaps compared to them, humans are morally insignificant.