MikkW

Karma: 1,226

MikkW 8 Mar 2020 21:05 UTC
3 points
in reply to: pnrjulius’s comment on: Interpersonal Entanglement
I dunno, chess against an AI is pretty compelling

MikkW 16 Mar 2020 6:46 UTC
2 points
on: Is That Your True Rejection?
I think there’s a slight misconception about Aumann’s agreement theorem here: “Common knowledge”, as Aumann defines it (and is leveraged in the proof), doesn’t just mean exchanging beliefs once: common knowledge means 1) knowing how they update after the first pass, then 2) knowing how they updated after knowing about the first pass, 3) knowing about how they updated after knowing about updating about knowing about the first pass, and so on

It’s only at the end of a potentially infinite chain of exchanging beliefs, that two rational agents are guaranteed to have the same beliefs. After a “first exchange”, two rationalists can very well still disagree, just less, and I’d honestly be kind of worried if they perfectly agreed with eachother after a first exchange

MikkW 18 Mar 2020 2:59 UTC
14 points
0
in reply to: Constant2’s comment on: Do We Believe Everything We’re Told?
Sure, you have to catch the baseball, but that doesn’t mean you have to eat the baseball

MikkW 20 Mar 2020 18:10 UTC
6 points
on: Thinking About Filtered Evidence Is (Very!) Hard
I follow most of this, but… uh, what’s PA?

Upon reflection, it seems you’re talking about Peano arithmetic. Would be good if the post spelled that out.

MikkW 20 Mar 2020 18:28 UTC
3 points
on: Thinking About Filtered Evidence Is (Very!) Hard
The set of all (valid) Peano Arithmetic theorems is not computably enumerable- wouldn’t this mean that a listener with a rich hypothesis space doesn’t need to assign a probability to the speaker enumerating them (thereby undermining the argument)?

I guess an important distinction is between computable and computably enumerable- I’m not perfect with CS, but it seems that this proof rests on the PA theorems being computably enumerable, but not computable themselves- is my read on this correct? But it seems to me that one can only enumerate the (valid) theorems by actually computing them

MikkW 21 Mar 2020 4:54 UTC
1 point
in reply to: Said Achmiz’s comment on: Coherent decisions imply consistent utilities
As for completeness, I struggle to see any practical difference between being unwilling to choose between two outcomes, and finding them equally acceptable (which is allowed by completeness).

Or, one can imagine someone relentlessly flopping between two highly (un)desirable outcomes because they are unwilling to settle, and I think it’s obvious what the problem there is.

MikkW 10 May 2020 4:39 UTC
1 point
on: The Mind: Board Game Review
Love the game. Not sure I see the common thread between The Mind and rationality, though?

MikkW 10 May 2020 6:00 UTC
1 point
in reply to: Alexei’s comment on: The Mind: Board Game Review
Gotcha

MikkW 11 May 2020 15:38 UTC
1 point
in reply to: Basil Marte’s comment on: Chronophone Motivations
Note: FPTP doesn’t actually give the mode of the distribution of the population. If we assume a normal distribution centered at 0, with candidates at {-0.25σ, 0, 0.25σ}, even though candidate 0 represents the mode, most voters prefer either of the two candidates towards the tail, so candidate 0 receives the fewest votes, thereby losing

MikkW 23 Jun 2020 3:05 UTC
17 points
on: Hammertime Day 1: Bug Hunt
There have been many moments in life where I’d see something bad happen, or someone who just needed help, but not do anything about it. This would always make me feel somewhat bad about myself.

I added a card to my Anki deck which said: “If you see someone who needs help, help them”. A few months later, I witnessed a car crash (Nobody was hurt, but there was gasoline pouring onto the ground). Part of me said “keep driving, someone else will take care of it, I have things to do”, but I heard in my head: “If you see someone needs help, help them”. And without even taking a moment to debate what I’d do, I hit the brakes, pulled over, and ran to the cars. Thankfully, no-one was hurt, but I called 9-1-1 and made sure everyone involved was okay.

There have been a few other similar situations since then, and that voice that I programmed into my head has always ensured that I did what I knew was right.

MikkW 23 Jun 2020 21:50 UTC
5 points
on: Hammertime Day 2: Yoda Timers
I’ve been finding Yoda timers helpful, but not because they show “look how much you can do in 5 minutes!”—many of my tasks had to do with cleaning things up around the house, and these tasks take at least an hour to actually accomplish.

There was some clutter in my kitchen from more than 2 years ago, including a container of Soylent powder, some syrups that I used to flavour plain Soylent, and bottles of vodka and rum. I only drink ready-to-drink Soylent (I had ordered the powder to try it out, but it’s not as convenient as RTD Soylent), and there’s now enough variety in the flavours of Soylent that I don’t have any need for extra flavours; I also no longer drink alcohol, because the idea of putting poison into my body to relax just doesn’t seem appealing to me these days.

Beyond all the unused items, the area was also just very dusty. I could have easily cleaned the area a long time ago, but it never really registered on my radar as a problem that needed to be fixed. Yesterday, walking through my house as part of the Bug Hunt, the messiness of the kitchen (as well as other parts of my house and car) stood out to me, so I made a note of it as a bug to potentially fix, and once I had chosen that bug for my Yoda timer, even though I knew it would take more than 5 minutes to thoroughly clean the area, the Yoda timer gave me just enough space to create a plan in my mind and rally the motivation to actually tackle this problem that has been there for a long time.

MikkW 7 Jul 2020 2:40 UTC
2 points
on: Classifying games like the Prisoner’s Dilemma
I like the analysis, but W, X, Y, Z is confusing terminology- it’d be easier to read and process if I didn’t have to constantly remind myself which situation each variable corresponds to

Calibrate words, not just probabilities

MikkW18 Jul 2020 5:56 UTC

11 points

3 comments2 min readLW link

MikkW 31 Jul 2020 1:09 UTC
3 points
on: Hammertime Day 6: Mantras
The motto I try to live my life by is “Live a life that would make you happy, if others lived as you live”—it’s like the golden rule, but even better. Obviously, it’s influenced by a lot of thinking about decision theory, both as discussed here on LW, and in my own thinking about how to live life. I do probably have a lot to improve in terms of actually integrating my behavior with this rule.

Some other mantras I like:
- From “How To Win Friends and Influence People” by Dale Carnegie (a very handy guide to not being an asshole): “The only reason you aren’t a snake yourself, is because your parents were not snakes”
- User Dagon on LW once wrote “If you aren’t taking actions contrary to incentives, you aren’t making moral choices”
- The future only exists as expectations based on past experiences
- Even if you are twice as good a medic as the typical medic, if you are ten times as good at firefighting, you should be fighting fires, not doing medicine. (This is Adam Smith restated)
And, I guess I’ll add a quote from a mathematician, just for our host Alkjash: “Sometimes general questions can be cleaner than specific questions, and easier to analyze”—Tim Gowers

MikkW 2 Aug 2020 17:31 UTC
9 points
in reply to: sapphire’s comment on: Tearing down the Chesterton’s Fence principle
Is there a good example that comes to mind of a time when this happened to you?

MikkW 2 Aug 2020 23:18 UTC
5 points
in reply to: Rafael Harth’s comment on: Inner Alignment: Explain like I’m 12 Edition
The post still contains a misplaced mention of MLK shortly after the first mention of Luther:
I’m very confident that this type of person is more common than Martin Luther Kings or Jesuses.

MikkW 7 Aug 2020 0:11 UTC
7 points
in reply to: Raj Thimmiah’s comment on: Hammertime Day 1: Bug Hunt
I think my experience with cards like this has been of a distinct character compared to cards I use for, e.g., languages. Whereas when I’m quizzed on the meaning of a foreign word, I say the answer, then immediately move on to the next card, when I see a card which is a sentence or rule, and where the “correct” answer is already obvious, I pause for a few seconds, and reflect on what that rule means, why I added the card in the first place, and maybe think of some examples of when I’ve successfully applied the rule, or failed to act accordingly.
I think the benefit SRS provides to this framework isn’t that it helps me remember something new that I risk forgetting, but rather, it schedules timely opportunities to refect on principles I chose to live life by, and ensure that I always have the principle in the back of my mind. It elevates a saying from a platitude to something meaningful to me.

MikkW 10 Aug 2020 17:01 UTC
2 points
in reply to: ketchupduck’s comment on: Transportation as a Constraint
This makes me imagine a city-skyscraper, where an inhabitant can do all their work, living, and shopping within the skyscraper, across a limited range of floors. If they only rarely ever need to go to ground level, then that can cut down on how much of the building is dominated by elevators

MikkW 10 Aug 2020 19:25 UTC
2 points
on: Reductionism
Tangential to the main point: one hypothesis for why the artillery gunner thought that “General relativity gives you the wrong answer”, is that maybe he had an experience with a software which could either run “Newtonian mode” or “GR mode”, and the software had to make approximations for the relativistic calculation to be roughly tractable (which might be nonetheless useful for roughly solving problems where relativistic effects matter, but would only reduce accuracy for non-relativistic situations).

Now, the “GR mode” (with approximations) would be a different model from real General Relativity, but could have given the gunner the impression that GR gives substantially different (and worse) answers from Newtonian mechanics in non-relativistic situations

MikkW 10 Aug 2020 20:39 UTC
6 points
on: MikkW’s Shortform
Epistemic: Intend as a (half-baked) serious proposal
I’ve been thinking about ways to signal truth value in speech- in our modern society, we have no way to readily tell when a person is being 100% honest- we have to trust that a communicator is being honest, or otherwise verify for ourselves if what they are saying is true, and if I want to tell a joke, speak ironically, or communicate things which aren’t-literally-the-truth-but-point-to-the-truth, my listeners need to deduce this for themselves from the context in which I say something not-literally-true. This means that almost always, common knowledge of honesty never exists, which significantly slows down positive effects from Aumann’s Agreement Theorem
In language, we speak with different registers. Different registers are different ways of speaking, depending on the context of the speech. The way a salesman speaks to a potential customer, will be distinct from the way he speaks to his pals over a beer—he speaks in different registers in these different situations. But registers can also be used to communicate information about the intentions of the speaker—when a speaker is being ironic, he will intone his voice in a particular way, to signal to his listeners that he shouldn’t be taken 100% literally.
There are two points that come to my mind here: One: establishing a register of communication that is reserved for speaking literally true statements, and Two: expanding the ability to use registers to communicate not-literally-true intent, particularly in text.
On the first point, a large part of the reason why people speaking in a natural register cannot always be assumed to be saying something literally true, is that there is no external incentive to not lie. Well, sometimes there are incentives to not lie, but oftentimes these incentives are weak, and especially in a society built upon free speech, it is hard to—on a large scale—enforce a norm against not lying in natural-register speech. Now my mind imagines a protected register of speech, perhaps copyrighted by some organization (and which includes unique manners of speech which are distinctive enough to be eligible for copyright), which that organization vows to take action against anybody who speaks not-literally-true statements (i.e., which communicate a world model that does not reliably communicate the actual state of the world) in that register; anybody is free (according to a legally enforcable license) to speak whatever literally-true statements they want in that register, but may not speak non-truths in that register, at pain of legal action.
If such a register was created, and was reliably enforced, it would help create a society where people could readily trust strangers saying things that they are not otherwise inclined to believe, given that the statement is spoken in the protected register. I think such a society would look different from current society, and would have benefits compared to current society. I also think a less-strict version of this could be implemented by a single platform (perhaps LessWrong?), replacing legal action with the threat of being suspended for speaking not-literal-truths in a protected register, and I also suspect that it would have a non-zero positive effect. This also has the benefit of being probably cheaper, and in a less unclear legal position related to speech.
I don’t currently have time to get into details on the second point, but I will highlight a few things: Poe’s law states that even the most extreme parody can be readily mistaken for a serious position;; Whereas spoken language can clearly be inflected to indicate ironic intent, or humor, or perhaps even not-literally-true-but-pointing-to-the-truth, the carriers of this inflection are not replicated in written language—therefore, written language, which the internet is largely based upon, lacks the same richness of registers that allows a clear distinction between extreme-but-serious postitions from humor. There are attempts to inflect writing in such a way as to provide this richness, but as far as I know, there is no clear standard that is widely understood that actually accomplishes this. This is worth exploring in the future. Finally, I think it is worthwhile to spend time reflecting on intentionally creating more registers that are explicitly intended to communicate varying levels of seriousness and intent.
What links here?
- How to build common knowledge of rationality and honesty? by MikkW (21 Feb 2021 6:07 UTC; 5 points)
- MikkW's comment on MikkW’s Shortform by MikkW (30 Oct 2020 0:49 UTC; 1 point)

MikkW

Cal­ibrate words, not just probabilities

Calibrate words, not just probabilities