LessWrong 2.0 Reader

Re: no coherent “stable” truth value: indeed. But still… if she wonders out loud “what day is it?” at the very moment she says that, it has an answer. An experimenter who overhears her knows the answer. It seems to me that you “resolve” this tension is that the two of them are technically asking a different question, even though they are using the same words. But still… how surprised should she be if she were to learn that today is Monday? It seems that taking your stance to its conclusion, the answer would be “zero surprise: she knew for sure she would wake up on Monday so no need to be surprised it happened”

And even if she were to learn that the coin landed tails, so she knows that this is just one of a total of two awakenings, she should have zero surprise upon learning the day of the week, since she now knows both awakenings must happen. Which seems to violate conservation of expected evidence, except you already said that the there’s no coherent probabilities here for that particular question, so that’s fine too.

This makes sense, but I’m not used to it. For instance, I’m used to these questions having the same answer:

1. P(today is Monday)?
2. P(today is Monday | the sleep lab gets hit by a tornado)

Yet here, the second question is fine (assuming tornadoes are rare enough that we can ignore the chance of two on consecutive days) while the first makes no sense because we can’t even define “today”

It makes sense but it’s very disorienting, like incompleteness theorem level of disorientation or even more

An idea I've been playing with recently:

Suppose you have some "objective world" space $\Omega$. Then in order to talk about subjective questions, you need a reference frame, which we could think of as the members of a fiber of some function $\omega :I\to \Omega$, for some "interpretation space" $I$.

The interpretations themselves might abstract to some "latent space" $\Lambda$ according to a function $\lambda :I\to \Lambda$. Functions of $\Lambda$ would then be "subjective" (depending on the interpretation they arise from), yet still potentially meaningfully constrained, based on $(\lambda ,\omega )$. In particular if some structure in $\Omega$ lifts homomorphically up through $\omega$ and down through $\lambda$, you get exactly the same structure in $\Lambda$. (And these obviously compose nicely since they're just spans, so far.)

The key question is what kind of space/algebra to preserve. I can find lots of structures that work well for particular abstractions, but it seems like the theory would have to be developed separately for each type of structure, as I don't see any overarching one.

Everyone works for money. Only one person, Mr. Purchaser, spends his money and everyone else just saves theirs forever. Suddenly money got a lot more powerful. Mr. Purchaser has literally all the money in the world and has pretty much infinite power, even if he only has $100k. He can make anyone do whatever he wants by paying them a penny, which is now worth about a million dollars[4].

I don't understand this. What use is money that is never spent? Why would Mr. Purchaser's penny induce me to do anything for him?

[4] Not that they'll ever spend the penny on anything, since they're one of the people who never spends any money, but let's pretend they still have motivation to earn it.

This is a pretence too far. The imaginary world you are describing is incoherent.

Money is the slack in the system of trade that saves us from having to exchange only by barter or informal systems of credit — doing each other good turns, in your terminology. In Adam Smith's words, "It is not from the benevolence of the butcher, the brewer, or the baker that we expect our dinner, but from their regard to their own self-interest."

If I imagine a world without any money, but where everyone is somehow able to coordinate and act rationally for the good of all...

If I imagine that, my thoughts run to hive minds in which there are no people as we know them today.

Adam Smith continues: "We address ourselves, not to their humanity but to their self-love, and never talk to them of our own necessities but of their advantages." A sentence that could have been penned by Dale Carnegie.

There are also people who's job it is to be a lot on the telephone and thus are well-reached by telephone even if they are younger.

Thanks for this piece. I admit I have always had a bit of residual aversion to seed oils that I've struggled to shake.

Having said that, as you're pushing so strongly against seed oils in favour of "processing" as a mechanism for poor health, I think I need to push back a bit.

If you want to be healthier, we know ways you can change your diet that will help: Increase your overall diet “quality”. Eat lots of fruits and vegetables. Avoid processed food. Especially avoid processed meats. 

"Avoid processed food" works very well as a heuristic - far better than anything like the "nutrition pyramid", avoiding saturated fats/sugars or calorie counting etc. But it also seems like something that should annoy people who like clear thinking and taxonomies. 

As you note, "processing" includes hundreds of processes, most of which have no plausible mechanism by which they might harm human health. Articles describing the ultra-processed taxonomy often just list a litany of bad-sounding things without an explanation why they're bad e.g. "mechanically separated meat", "chemical modifications" and "industrial techniques". Most of these are either benign when you think about it (we'd all prefer a strong man wearing a vest separating our meat with his bare hands, but come now...), or so vague as to be uninformative.   

If ultra-processed foods are bad because they contain "hydrogenated oil, modified starch, protein isolate, and high-fructose corn syrup" or "various cosmetic additives for flavour enhancement and colour", then it's these products that are bad, not some mysterious processing! 

If it is some technical part of the processing, like "hydrolysis, hydrogenation, extrusion, moulding, or pre-frying" that's bad, surely we should just identify that rather than lumping everything together? 

If it's some emergent outcome of all these processes, like "hyper-palatability" or "energy density", then that's the problem, not the fact of being "processed". If so we should all stop eating strawberries after they hit a certain deliciousness threshold, and avoid literally any edible oil (because all oil is identically energy-dense). 

But, having said that, I still use this heuristic, and I'm pretty glad I trained myself out of preferring highly-processed food when I was less analytical. 

Good question! I would find it plausible that it would have changed, except maybe if the people you'd call would be in their fifties or older.

You're saying governments can't address existential risk, because they only care about what happens within their borders and term limits. And therefore we should entrust existential risk to firms, which only care about their own profit in the next quarter?!

A piece of advice I frequently hear: always make sure you call somebody in the company you're applying for.

Is this still up-to-date advice? Or is messaging someone over LinkedIn or similar more appropriate? Mostly asking because I got the impression that the internet changed the norms to no one doing phone calls anymore.

Good article. 

It's an asymmetry worth pointing out.

It seems related to some concept of "low interest rate phenomenon in ideas". Sometimes in a low interest rate environment, people fund all sorts of stuff, because they want any return and credit is cheap. Later much of this looks bunk. Likewise, much EA behaviour around the plentiful money and status of the FTX era looks profligate by todays standards. In the same way I wonder what ideas are held up by some vague consensus rather than being good ideas.

Feels like there is something off about the following graph. Many people writing critiques care a lot. Émile spends a lot of time on their work for instance. I don't think motivation really catches what's going on.

Epistemic status: generating theories

I theorise it's two different effects in one:

• The voices we hear in the discussion (which links to yours)
• The norms of the communities holding those voices

First, as you say, the voices we hear most are the most confident/motivated, which leaves out a lot of voices, many of whom might talk in a way we'd prefer. Instead we only hear from the fringes, which makes a normal distribution look bimodal.

I wonder if this is more like supply and demand than your "bars" model. Ie it's not about crossing a bar but about supplying criticism that people demand.  And correcting a status market - EA is too high status, let's fix it. 

Secondly, the edges of this normal distribution have different norms. Let's say there are 3 areas:

• one likes steelmanning in disagreements 
• one likes making clear to be on the side of minorities
• one likes being interesting

Let's imagine we are discussing something that has people from all these areas.

The people who like each of these things most strongly perhaps talk more, as in the above example. But not only do they talk more, they talk differently. So now the discussion is polarised in different languages, because the people in the middle are less confident and speak less (this jump feels like weakest step in the argument^[1])

Amount of people with different views (central line is one group of people, who hold all views weakly)

So now we have this:

So I think probably my overall thing about why criticism is poor is something like "criticism looks poor to us because it isn't for us". It is for the people in the same communities by whom it is written. And probably to them our pieces look pretty poor as it is. 

Some questions then:

• How do we respond in language that other groups will understand?
• Should we want to? Torres for instance seems to be a bit of a bully. But I'm not sure that makes their arguments bad. But if I were doing it they would definitely call me out for it.
• Is it worth taking time to really try and write the strongest versions of criticism in language we understand. Or find ways for people to signal confusion

1. ^{^}

   Why should the people in each group talk most confidently? I dunno, but I hear a lot more from Yud, Alman,  Adreeson and Torres than many more moderate voices. Feels like something is going on here. Can anyone suggest it?