LessWrong 2.0 Reader

Calculations on Hydroelectric Energy Storage

For those interested in the numbers on pumped hydroelectric storage, we can get more energy by increasing 'head' or the distance that the weight falls, from 6 meters to up to 500 meters for some of the largest projects (and we could in theory go bigger).

Let's pick a more reasonable number like 60 meters:

MASS/house = 15 kWh/house / (9.8 m/s² × 60 m) = 91,836 kg/house = 91 m^3/house

Let's say we have a dam with ~20 meters of water level fluctuation (drawdown). Then that's 5 m^2 per house of surface area.

As a sanity check, Bath County Pumped Storage Station in VA stores about 24000 MWh/ 30 KWh/house = 800,000 houses worth of energy.

800,000 houses * 5 m^2 = 4 km^2

The Bath County reservoir is about 1km^2 so we're in the right range here (the reservoir has a little more drawdown and a way bigger head).

Nice, especially the second half.

To me, this post makes it seem like the natural place to push forward is to try and restrict incoherent policies over lotteries until they're just barely well-behaved - either on the behavior end (continuity etc) or the algorithmic end (compte budgets etc).

I'm going to start by attacking this a little on my own before I even look much at what other people have done.

Some initial observations from the SQL+Python practice this gave me a good excuse to do:

So my current best guess (pending understanding which gear is best for which class/race) is: 

Some theories I need to test:

Wild speculation:

Sure! Here are two of my favorites.

(1) From Leil Lowndes' book:

Don't ask what they do. In the US in my experience, the most common question upon meeting someone is "what do you do?" But the problem with this is that while 65% of Americans are satisfied with their jobs, only 20% of Americans are passionate about their work. From Lowndes:

If you instead ask, "How do you enjoy spending most of your time?" It allows people to mention their job or their hobbies. And homemakers are no longer embarrassed to say, "I'm just a mom" to the question of "what do you do?" 

(2) From Dale Carnegie's book:

Never disagree and say "you're wrong". I am a naturally disagreeable person. Learning about this technique hasn't made me more agreeable, I just express my disagreement differently now. From Carnegie:

Never announce, "I am going to prove so-and-so to you." That's bad. That's tantamount to saying: "I am smarter than you are and am going to make you change your mind."

"We sometimes find ourselves changing our minds without any resistance; but if we are told we are wrong, we resent the imputation and harden our hearts. We are heedless in the formation of our beliefs, but find ourselves filled with a passion for them when anyone proposes to rob us of their companionship. It is obviously not the ideas themselves that are dear to us, but our self-esteem which is threatened."—James Harvey Robinson

I've adopted a more indirect way of challenging people's beliefs. Rather than stating my disagreement, I tend to ask questions (à la the Socratic Method) to get to the root of somebody's belief. Sometimes they'll notice contradictions in their own arguments without me having to point them out.

Some related posts:

• one example [LW(p) · GW(p)] among many of a long runway letting me make more moral choices
• ongoing twitter thread on frying pan agency

Ah. So I guess you should neither ask just questions, nor just always talk about yourself, but rather balance both.

Praising questions is an interesting tip to ease-in more introverted people. This reminds me of another strategy a sociable colleague of mine used on shy people: He made playful jokes about his conversation partner ("teasing"), encouraging them to be bold and "hit back" with something, and then laugh about their joke. Which instantly empowered them. This probably works best when both are men. Making self-deprecating jokes is a safer option, though not as effective.

Thank you for your response!
 

There were 3 situations today where the second case arose:

1. I asked whether GTP was less stable than ATP in AP Bio (I wanted to understand how it was possible that GTP + ADP -> GDP + ATP,  and if this was the case, why ATP was used at all).  My teacher thought I was asking about it from an evolutionary standpoint, and started explaining the evolution of the system.
2. I have a tendency to speak in short bursts of content in which I think I often underestimate inferential distances. (My internal monologue normally consists of vaguely used phrases + a sense of what I mean,  which I think this carries over into my speech).
3. I'm realizing that I find it difficult to phrase sentences correctly. I rewrite almost every sentence I type many times over because I'm not sure about grammar and word arrangement (I spent a minute thinking about whether I should wrap "many times over" in a pair of commas at the start of this sentence).

Here's a new Bookkeeping Theorem, which unifies all of the Bookkeeping Rules mentioned (but mostly not proven) in the post, as well as all possible other Bookkeeping Rules.

If all distributions which factor over Bayes net $G_1$ also factor over Bayes net $G_2$, then all distributions which approximately factor over $G_1$ also approximately factor over $G_2$. Quantitatively:

$D_{KL}\left(P\right[X\left]||{\prod }_{i}P\right[X_i|X_{p{a}^1\left(i\right)}\left]\right)\ge D_{KL}\left(P\right[X\left]||{\prod }_{i}P\right[X_i|X_{p{a}^2\left(i\right)}\left]\right)$

where $p{a}^j\left(i\right)$ indicates parents of variable $i$ in $G_j$.

Proof: Define the distribution $Q\left[X\right]:={\prod }_{i}P\left[X_i|X_{p{a}^1\left(i\right)}\right]$. Since $Q\left[X\right]$ exactly factors over $G_1$, it also exactly factors over $G_2$: $Q\left[X\right]={\prod }_{i}Q\left[X_i|X_{p{a}^2\left(i\right)}\right]$. So

$D_{KL}\left(P\right[X\left]||{\prod }_{i}P\right[X_i|X_{p{a}^1\left(i\right)}\left]\right)=D_{KL}\left(P\right[X\left]||Q\right[X\left]\right)$

$=D_{KL}\left(P\right[X\left]||{\prod }_{i}Q\right[X_i|X_{p{a}^2\left(i\right)}\left]\right)$

Then by the factorization transfer rule (from the post):

$\ge D_{KL}\left(P\right[X\left]||{\prod }_{i}P\right[X_i|X_{p{a}^2\left(i\right)}\left]\right)$

which completes the proof.

Information you should probably remove from the weights

Perhaps it might also be useful to remove information which might reduce the likelihood that 'A TAI which kills all humans might also doom itself' (especially in short timelines/nearcast scenarios).

Curated. This post hit me exactly at a moment where I was trying to become a more "numerate person" (i.e. gain some skills where it was easier to convert problems into 'things with numbers' and then solve the problems with simple math).

I liked the worked examples of the post, and I appreciated the emphasis on why it's important to keep track of units (which I had roughly known, but seeing how it played out in the examples drove the point home more).