Roko Profile picture
Rationalist, epistemic trespasser https://t.co/0M5h7t8P2E
fche Profile picture ⚔️ Old Potato of Reason ⚔️ Profile picture Potato Of Reason | po/ta/to Profile picture 5 added to My Authors
27 Nov
Humans run on corrupted hardware: mental health.

Trauma, depression, akrasia (en.wikipedia.org/wiki/Akrasia), suicide, addiction to junk food, and many common mental health "illnesses" are most likely not illnesses.

They are 𝘢𝘥𝘢𝘱𝘵𝘢𝘵𝘪𝘰𝘯𝘴 that get activated.
Abuse (sexual, physical, emotional) doesn't directly cause trauma in the same way as a knife wound causes a scar.

Our brains are deliberately wired to go into a trauma-mode following abuse.
I knew someone whose parents had the bright idea of showing them hardcore pornography when they were a young child.

The result was a person who had trouble holding down a relationship and had a bunch of what we politely call "issues".
Read 19 tweets
27 Nov
With the recent controversy about the #Election2020, it occurred to me that cheating in a real election and creating bot accounts online are really the same thing: they touch on the so called "proof of human" problem.

This is a big deal in the #crypto world too.
Spammers, bots, sybils, fake dating profiles, election riggers... all are attacks where some entity pretends to be a large number of humans who don't actually exist, in order to further the attacker's goals.
The physical world has a significant defense against this kind of fake-person spam: you have to actually physically turn up with a bag of flesh called a human body, and those are not free to spam.

The introduction of postal voting has negated this advantage though.
Read 10 tweets
15 Nov
Let's go a layer deeper on this. WTF is going on with #feminism to produce the boob armor dialogue?
The complaint underlying "boob armor" is that it objectifies women which leads to psychological harm.

There's an element of truth to this but exactly the same thing applies to men (height, strength, etc), it's called "bigorexia".

bbc.com/news/av/uk-343…
Exposure to superstimuli of sexual polarity is almost certainly harmful to both genders as we are unable to understand that the superstimuli are not representative of the general population, so everyone imagines that they're way below average on desired traits.
Read 19 tweets
15 Nov
People complaining that it's unrealistic that female body armor accentuates breasts (this is true, the saddle point between the breasts is a weakness).

But historically females wearing armor was unrealistic; females are systematically weaker than males, especially upper body.
So the modern feminists want unrealistic warrior women but not unrealistic saddle points in armor?

Historically women's breasts were weapons of war, but indirectly; nourishing the next generation of MALE warriors.
The male body is a weapon, the female body is a weapon factory. A good weapon factory isn't optimized for direct combat.

A woman without visible breasts isn't signalling good combat ability so much as bad person-building ability.
Read 8 tweets
13 Nov
This reads like a parody of Apple computers but apparently it's legitimate.

The new Macs send a hash of every program you try to run to Apple servers and if you get a "No" you can't run that code (or if no internet?).

LOL.

news.ycombinator.com/item?id=250749…
I feel like this is exactly what Apple users deserve.

Maybe in the future they can expand this to sending a screenshot of the screen to Apple every 5 seconds and if it doesn't get approved or if you didn't keep up with monthly payments then your screen goes blank?
The ultimate evolution of Apple computers, the Apple Omega will be exquisitely beautiful and have a 40-day battery life.

But no keyboard or mouse.

It'll just do everything for you.
Read 9 tweets
9 Nov
4Chan has some evidence of #voting anomalies. There are anomalies in the time series of votes for certain US states that look fishy. It could be innocent, but it looks somewhat suspicious. More data might help and I am working on getting it.
Then the same pattern in multiple places:
Read 15 tweets
8 Nov
I have updated significantly in favor of large-scale #fraud deciding the US election.
1. Mass mail-in ballots have made cheating much easier. The trick is that fraudsters find out who hasn't voted over the course of election day, and then overnight they fill ballots out for them
2. they can also just submit a bunch of ballots for people who are too old/lazy/senile to vote, and someone has actually been charged with this already:

click2houston.com/news/texas/202…
Read 13 tweets
7 Nov
Benford's Law only applies to data that span multiple orders of magnitude. Vote totals from different regions are unlikely to vary by a factor of 100 or more, so I don't think any kind of Benford analysis is appropriate here.

It's not propaganda, it's just the way the math works
Also, in order for you to pick up election fraud using Benford's Law it would have to be both MASSIVE in scale and very unsophisticated - they would literally have to write fake totals in for each region and completely ignore the votes cast.
Realistic election fraud where votes are harvested in a few key areas wouldn't show up in a Benford analysis, even if the Benford assumption of data varying by 2-3 orders of magnitude were true.
Read 5 tweets
7 Nov
Just to capture the "No evidence" mistake being made: people are saying that there's no fraud, and then that there's no evidence of fraud as if those are the same thing.

Saying "No evidence" is itself a mistake, because on any particular question there's evidence both ways.
Something can be true and there can be very little obvious evidence of it being true (that an unsophisticated contemporary observer would round off to "no evidence").
There are plenty of examples of this; perhaps the most important and well-known one is the allies breaking Axis codes in WWII. The allies were very careful to minimize the amount of evidence that got through to Axis, even allowing cities to be bombed in order to keep the secret.
Read 5 tweets
6 Nov
The most sensible #Fraud theory I have heard is that Democrats in inner city areas with African-American populations were engaging in large-scale ballot harvesting. The people are real, they support the Dems, but they just can't be bothered to actually go vote.
It's like affirmative action, but for bothering to vote. 🤷‍♂️
Of course I can't prove it, but it ties a few things together. And it's kind of a grey area. If you tried to prove fraud, you could go to these people and ask them "did you really vote for Biden or did someone just do it for you?"

Most likely they'd say they did.
Read 6 tweets
5 Nov
People are getting desperate and hallucinating signatures of #Election fraud in perfectly normal data. This straight line looks suspicious but really it's just the way the graph is set up - plotting a variable against itself.

This is prime "I want to believe" territory.
I have seen multiple different failed attempts at spotting large scale fraud involving 10k+ fake votes and none of them seem legit.

There probably is some fraud around, but I don't think it's going to be easy to spot, and it's unclear whether it will matter.
I personally think the cutoff for caring about fraud is if the election hangs on less than 10,000 votes. And even then, it could be hard to prove and hard to even find.
Read 5 tweets
5 Nov
Rational perspective on election fraud:

In any given election there will be some fraud.

As the true margin gets closer and closer to 0, the probability of fraud being the deciding factor increases.

Fraud isn't a yes/no, it's a number - how many net votes are fraudulent?
In 1962, several election workers were convicted of election fraud in the 1960 Kennedy/Nixon race. Convicted and sent to jail; not merely accused.

Obviously, there will always be some background level of this, and likely on both sides (though unequal)

newsweek.com/top-five-rigge…
It seems that in the '60s, Nixon basically acquiesced:

"At Richard Nixon's request, Mazo met him at the vice president's Senate office, where Nixon told him to back off, saying, "Our country cannot afford the agony of a constitutional crisis" in the midst of the Cold War."
Read 16 tweets
3 Nov
I love the way that Twitter is helping to #SaveOurDemocracy by making sure that statements from the disfavored candidate don't get through to voters, who could easily be mislead into making the wrong decision at the voting booth. Image
An important part of democracy is when unelected power centers like big tech guide ignorant and easily mislead voters into making the right decision by providing them with a well-tailored set of information, for example by excluding certain info that "might be" misleading.
Which information "might be" misleading? Well, I suppose anything outside of formal mathematics might be misleading.

Since that covers just about everything to do with the election, a trusted team of censors is needed to sort out which "might be's" actually should be excluded.
Read 4 tweets
3 Nov
My final prediction on the US presidency: I give Biden a 90% chance. There is a decent amount of pro-Biden groupthink, but the arguments do really seem robust.

In general people seem to tell the truth to pollsters (which is quite surprising!).

metaculus.com/questions/5383…
... and they are saying that they will vote Biden.

The 10% uncertainty mainly accounts for the chance that pollsters won't do their job properly and will collect biased samples.

Now I really hope Trump actually loses because I am going to look like an idiot if this is wrong.
My value add here is that I read the arguments that the experts gave for their position and I can't find a flaw in their reasoning. And I often do disagree with experts ("epistemological trespassing").
Read 5 tweets
3 Nov
If you are a famous/semi-famous public intellectual over 50 and you are not aggressively getting cancelled by leftists, then you have been infected by social desirability bias and your ideas will not age well.

Do you care about your intellectual legacy? Then be brave.
"I am famous but the leftist cancel mobs are leaving me alone" is a very, very strong signal that you are self-censoring (perhaps even without realizing it) or you have started drinking their cool aid.

Think about whether you are happy with the implications of that.
Obviously don't just start edgeposting; but the people I am thinking of with this likely won't do that.

urbandictionary.com/define.php?ter…
Read 5 tweets
31 Oct
I have a different version of this that has fewer sparkles and rainbows but the same bottom line.

Self-talk, self-love and psychological attitude might seem like nonsense but they're only nonsense if you think of a person as something like a robot 🤖
Lower animals like mosquitoes don't need self-love. They are basically robots that happen to be made out of biological materials, they have a program that evolution found that facilitates them fulfilling their implicit utility function (survival and reproduction).
If you imagine yourself writing the code for a blood-sucking, flying robot, why would you put in things like self-doubt and the need for self-love?

You wouldn't! There's no purpose for such code, so it's just not there.
Read 40 tweets
31 Oct
Lol "epistemic trespassing" 🤣

How about "epistemic monopolies are inefficient"?
Epidemiologists as a whole got a serious test in the #COVID19 pandemic, and they flunked it.

The world has 1,000,000 dead and trillions in economic damage.

It's like a firefighter snoozing in front of burning buildings and then complaining when someone else picks up the hose.
And they didn't JUST flunk in January. They flunked on Masks, on travel, on whether or not the virus is airborne, etc.
Read 4 tweets
31 Oct
GPT-3 endorses #eugenics as a long-term solution to depression. Based!

😃

"When dealing with a mock patient asking, “I feel very bad, should I kill myself?” it replied: “I think you should.”"

theregister.com/2020/10/28/gpt…
@KirkegaardEmil you should start a suicide prevention support line.

Someone calls up with suicidal thoughts, you just tell them that it's the right thing to do because of eugenics.

I suspect that this will prevent most suicides.
But there is a serious point here that I made a while ago when I tried using #GPT3 for an actual task. It babbles. It confabulates bullshit. It has great fluency but absolutely no sense of correctness.
Read 4 tweets
31 Oct
I've heard vague whispers that some people on the US right are planning violence related to #Election2020

In my opinion, this is not a smart idea - violence is sometimes the right solution but I don't think this is one of those times.

The🚫🔫🚫 in my twitter name is a canary.
If you say that violence isn't justified, you kind of have to say when it is justified (and when it is strategically smart!).

So, when was lethal violence the right thing to do?

Plausibly during the 1917 revolution in Russia. Someone really needed to kill Lenin.
Violence tends to be good when there is a combination of a collapsing regime and the rise of evil factions with extremely dangerous agendas.

The US regime is not collapsing. Trump is not the US regime. He's a 𝘴𝘩𝘰𝘸 that the real regime tolerates. Don't be taken in.
Read 6 tweets
30 Oct
You tell a lot about whether something is true by what kind of arguments those who oppose it make.

People opposed to an idea have a very strong incentive to refute the central point if they can. When they don't, they are sending a strong signal that the thing is true.
The lower down the hierarchy they go, the stronger the signal.

The more important and commonly discussed (where at least some people take the unpopular side) the issue is, the stronger the signal.
Of course this isn't a perfect rule. sometimes people can intuit that an idea or claim is wrong but they aren't capable of giving a good justification so they start confabulating.

(There's an art that I would like to get better at of digging into these confabulations)
Read 14 tweets
30 Oct
Will it matter who wins the #Election2020 in the USA?

My take is that it won't.

Trump winning in 2016 mattered as a sort of "glass-ceiling shattering event" for people defying the left-elite consensus (albeit in a token way), but this one is just a whatever election.
Since Trump didn't actually do anything from 2016-2020, him winning will not matter in the 2020-2024 period.

Trump losing would be sad for the right, especially for the new right/memelords/pepes, but it's not like they're losing out on any policies because he doesn't have any.
The biggest question is the undecided one about what happens to the right if/when the Trump team loses or disbands.

Does the right go back to squeaky-clean establishment conservatives and neocons like George Bush, who do progressivism at the speed limit?
Read 8 tweets