Richard_Kennaway

Karma: 8,797

Computer scientist, applied mathematician. Based in the eastern part of England.

Fan of control theory in general and Perceptual Control Theory in particular. Everyone should know about these, whatever subsequent attitude to them they might reach. These, plus consciousness of abstraction dissolve a great many confusions.

I wrote the Insanity Wolf Sanity Test. There it is, work out for yourself what it means.

Change ringer since 2022. It teaches learning and grasping abstract patterns, memory, thinking with your body, thinking on your feet, fixing problems and moving on, always looking to the future and letting both the errors and successes of the past go.

I first found an LLM useful (other than for answering the question “let’s see how well the dog can walk on its hind legs”) in September 2025. As yet they do not form a regular part of anything I do.

Richard_Kennaway 15 Aug 2026 11:43 UTC
2 points
0
on: Why I’m Skeptical of Longtermism
1. The future will be vast in expectation.
There is a tension between prediction and action. (No, these are not the same thing.)

We can straightforwardly predict, on the basis of current technology and simple near-future extrapolations of it, that it is possible for the future to be vast. Francis Bacon successfully did that already. But whether it turns out that way depends on people — us in the present, and our successors indefinitely forward — to bring it about. There is no expectation value that is not conditional upon that.

Richard_Kennaway 10 Aug 2026 11:34 UTC
2 points
0
on: Richard_Kennaway’s Shortform
SMBC on weird philosophy.

Richard_Kennaway 10 Aug 2026 8:59 UTC
5 points
0
in reply to: Karl Krueger’s comment on: Karl Krueger’s Shortform
Looking on the bright side, if there was no security, it didn’t hack the security!

Richard_Kennaway 9 Aug 2026 19:09 UTC
4 points
0
in reply to: Dean Valentine’s comment on: lc’s Shortform

Lie detectors work.

Says company selling lie detectors.

Richard_Kennaway 8 Aug 2026 11:02 UTC
2 points
0
on: Some wild metaphysics that seemingly everyone needs to accept?

So for each of these universes, our probability of it being the “true reality”, and us actually being a freak observer in it is nonzero.

There is a large gap between “nonzero” and “big enough to matter for any decision I will ever make”. We are talking a smallness well below .

Here’s a source of “freak observers” in Tegmark level 4: the number between 0 and 1 whose decimal expansion is the concatenation of all Turing machines applied to all finite initial tapes. Everyone and everything computable is in there, countably many times. How concerned am I? Not at all. What does it even mean to say that “I” am in there? “I” is only a single point in a continuous landscape of entities like me or less like me or nothing like me, with nothing to single one out as “me” without already knowing what “me” is.

The only way to find a book in the Library of Babel is to write it oneself.

Richard_Kennaway 6 Aug 2026 19:18 UTC
9 points
0
on: Richard_Kennaway’s Shortform
“John Cage’s 639-year organ performance reaches another milestone with rare chord change”

May there be people to see out its remaining 614 years.

Richard_Kennaway 6 Aug 2026 10:35 UTC
7 points
1
in reply to: winstonBosan’s comment on: winstonBosan’s Shortform
Is disempowerment a convergent result of having above-human AI at all, with a Butlerian jihad the only way out?

Richard_Kennaway 5 Aug 2026 16:02 UTC
6 points
2
in reply to: anaguma’s comment on: winstonBosan’s Shortform
I’m wondering who Ethan Roland and Zeke Reffe-Hogan are. There are google hits, but none of the candidates look very plausible.

Richard_Kennaway 5 Aug 2026 14:13 UTC
2 points
0
in reply to: sirawit’s comment on: Norms Are Broken Laws
I don’t want a clean kitchen. Why are there other people deciding that I would want a clean kitchen?

No-one is. I am not. I do want my own kitchen to be clean, my whole house to be in good order. So I do the work to get what I want, because there is no-one else, I’m it.

It is up to you to decide what you want. Referring to your own examples, which of these do you want to do?
- Chatter excitedly at the top of your voice on public transport. Punch anyone who asks you to keep it down.
- In a self-clear cafeteria, just leave everything on the table.
- Be rude.
- Diss everyone.
- Treat your promises like pie-crust, made to be broken.
Is it ok if everyone else does the same?

Or maybe you don’t even want to decide what you want, and what you will pay to get it?

But you have said you want laws to enforce customs. So now, who are you to decide such things for others?

There are a lot of things my parents wouldn’t allow me if they I were still in their control. In fact, nowadays my parents allow me

That’s a strange juxtaposition: “if … I were still in their control” and “nowadays my parents allow me”. Which is it? You are independent of them, or they are able to allow (and therefore disallow) you to do this or that?

Richard_Kennaway 5 Aug 2026 14:07 UTC
10 points
0
in reply to: winstonBosan’s comment on: winstonBosan’s Shortform
For those not in the very select group of Gwern’s Twitter followers: “Guardian Angels: LLM Personalization for Productivity and Security”.

Richard_Kennaway 5 Aug 2026 8:46 UTC
5 points
−3
on: The AI Race is Not a Prisoner’s Dilemma
I didn’t find the following game in the article. It could be called “Doomsday”, or “IABIED”.

A, B
Strike, Strike –Bignum, –Bignum
Strike, Refrain: –Bignum, –Bignum
Refrain, Strike: –Bignum, –Bignum
Refrain, Refrain; 0, 0

Richard_Kennaway 5 Aug 2026 8:18 UTC
2 points
0
on: Norms Are Broken Laws

I’m not claiming that everyone belongs to one among the three characters. There may be others, but I’m not considering the other types here, for the reason of simplification. I argue that it is enough to consider only these.

It is not enough to consider only these, because there is a crucial aspect of all of your examples that you have left out, which vitiates the entire posting.

Cleaning the kitchen has a cost: the time and effort put in.
Cleaning the kitchen has a benefit: a clean kitchen.

Not cleaning the kitchen has a cost: a dirty kitchen.
Not cleaning the kitchen has a benefit: saving the time and effort.

You have left out of your description the benefit of cleaning and the cost of not cleaning. But a clean kitchen is the real, physical, visible result for which we clean our kitchen. The cost is effort and it is worth making that effort.

You have analysed the situation as if it were purely a symbolic ritual, undertaken for no reason but complying with the ritual, like not stepping on the cracks between the paving stones. That may be how such things seem to small children who know nothing, but as we grow up we — at least, most of us — come to appreciate that what we originally learned as imposed rules has a real purpose, by making the world a better place for ourselves and everyone else. Some of those rules we may decide do not serve a purpose and can be dropped. Being an adult means you get to decide your own rules. How much do you want a clean and tidy kitchen? The kitchen on its own is a small matter, but the general issue is a large one.

Wanting laws to enforce such things is like wanting your parents to still wipe your nose for you.

Richard_Kennaway 2 Aug 2026 13:54 UTC
14 points
9
on: Richard_Kennaway’s Shortform
Twenty-two concluding sentences to make your essay sound like AI:

It’s not X. It’s Y.
It’s not X. It’s not Y. It’s Z.
X changed. Y didn’t.
X was just the tell. Y was what it revealed.
X was a distraction. Y was the thing to watch.
X was never in play. The play was always about Y.
That wasn’t a symptom. It was the disease.
That wasn’t the disease. It was only a symptom.
It wasn’t rationality. It was legibility.
It wasn’t legibility. It was obfuscation.
It wasn’t obfuscation. It was precisely, literally, true.
It’s not dead. It’s sleeping.
It’s not sleeping. It’s dead.
Not confused. Operating as intended.
Not a failure. Operating as intended.
They never mattered. Only the idea of them mattered.
Everyone knew. Then someone said so.
It couldn’t happen. Then it did.
That’s the way that can be named. It’s not the eternal way.
It’s three. It’s also one.
No commas. Full stops.
Negation. Assertion.

Richard_Kennaway 2 Aug 2026 7:57 UTC
1 point
0
in reply to: Thomas Kwa’s comment on: strawberry calm’s Shortform
Jam tomorrow when there may be no tomorrow.

Richard_Kennaway 31 Jul 2026 9:20 UTC
5 points
0
on: Why Prayer Is Not Answered by Marion Zimmer Bradley
“Are all our prayers answered? Yes, they are; what people who ask that question often don’t realize is that sometimes the answer to our prayer is ‘no’.”

— Christopher Durang, “Sister Mary Ignatius Explains It All For You”

A video adaptation of the relevant scene..

Richard_Kennaway 30 Jul 2026 16:35 UTC
2 points
0
in reply to: sirawit’s comment on: sirawit’s Shortform
Heinlein remarked that “the wisdom of the ages” usually sounds like just being too tired. So in case it sounds like I’m just too tired, here’s some inspiration in the opposite direction, from Ratavia Rock Project.

SCREAM UNTIL DAWN

THIS WORLD WILL HEAR ME

Richard_Kennaway 30 Jul 2026 10:40 UTC
3 points
0
in reply to: sirawit’s comment on: sirawit’s Shortform
The track record of such extremism is poor. I was going to say it would take too much time to search out people’s reports of mental breakdowns, suicides, and other ways of going off the deep end, but conveniently a new one has appeared today. You can also search LW for “Ziz” and “Ziz(z)ians”. These are by no means the only ones.

If you read further in my Insanity Wolf page, I hope you will find things to disagree with, such as strangling all the evil people with each others’ intestines, and maximising happiness by killing everyone. The former was inspired by a remark of Voltaire, and the latter by those utilitarians who believe, literally, that life is suffering and that it would be better for everyone never to have been born.

But if I suppose your assent limited to absolutist positive utilitarianism á la EA and Peter Singer, there is a substantial issue there that has been much discussed by philosophers without any resolution. Is there supererogation?

A supererogatory act is one that it is praiseworthy to do, but which is not required. It is action over and above the call of duty. Is there any such thing?

Peter Singer and some EAs say no. Supererogation is a contradiction in terms. The good that you can do is the good that you must do. At every moment you must be doing the best thing you possibly can. All the obvious objections are folded into the basic injunction. You don’t know the best thing? Then the best thing is to find out. You’ll burn out? Then take some R&R, but only as much as needed to carry on the Work. It’s an impossible standard? Yes, but it’s the right standard, you just don’t want to because it’s hard. But—? Denied. You must make of yourself a slave to the good. “O feet happily fettered on the road to salvation!” While you are sleeping, the world is burning! You cannot ascend the torrent of your sins! Nothing for yourself, everything for the cause! Jam tomorrow? Jam never! Your life must be a desert! You are nothing, the cause is everything! Smile! The torture will continue until you comply. You can’t reason with it, you can’t bargain with it, and it absolutely will not stop, ever, until you are dead.

The crazy train goes straight to hell. Seeing this, I do not get on it at all. And yet it starts out so reasonably, how could anyone disagree?

Suppose, Glaucon, that you are faced with a choice between a good deed and an evil one, which should one choose?

Of course the good one, Socrates, that’s what we mean by the words “good” and “evil”!

Then, Glaucon, suppose a very good deed and a moderately good one?

The better one, of course, Socrates, that is what we mean by “better”!

And therefore, Glaucon, is it not clear that the proper choice must always be the best possible choice that you can make? Is not that what we mean by “best”?

I cannot disagree with that, Socrates.

And therefore, Glaucon, you must agree that you are obliged to follow the Gospel According To Insanity Wolf to its very end?

That is very distasteful to me, Socrates, and I fear that your broad, straight path of reason, easy to follow and difficult to step aside from, has nevertheless led us into error. I am perplexed. But Socrates, I observe that your life is not in such derangement as would result from taking this conclusion seriously. I think you have led me down a false path in order to provoke me to find a better one.

(The ancient sources break off here.)

I have never seen a philosopher make a defence of supererogation that does not come down to declaring the absurdity of the extremity. But pro tem, that absurdity is enough for me.

“If daylight itself needs proof, nothing can be proved.”

Richard_Kennaway 29 Jul 2026 9:12 UTC
4 points
0
in reply to: sirawit’s comment on: sirawit’s Shortform
I can’t make much sense of your comment, but here’s some background to Insanity Wolf.

Insanity Wolf is a meme of long standing. I didn’t invent it, but I created my Insanity Wolf page to highlight what seemed to me some dysfunctional thought patterns that recurred on LessWrong and the rationalsphere generally. For example, the first section of that page is about Effective Altruism and totalising Singerian ethics, but there is much, much more. Roko’s Basilisk is in there, and the idea that you do not exist, and universal arguments, and straightforward lunacy confidently screamed. I don’t make up these memes. The ideas are in the water supply, they come to me and I write them down. I have more than 1000 more that I haven’t got round to adding, but maybe nowadays I could vibe-code something. Perhaps an Insanity Wolf quote of the day for people who need to regularly wallow in shit to remind themselves why it’s a bad idea to wallow in shit.

HORROR! HORROR! HORROR!
DON’T THINK ABOUT THE HORROR!

I don’t know what your last sentence means, but if you want a “morally good wolf”, go see Courage Wolf. Courage Wolf is wise, wise but stern, very stern. Eliezer approves! Here’s my concentrated dose of Courage Wolf.

Also related, Eliezer’s Cognitive Trope Therapy, which informed the content warning I wrote for the Insanity Wolf page.

Your suggestion of Yell culture is an invitation for people to be Insanity Wolf.

Richard_Kennaway 29 Jul 2026 7:31 UTC
3 points
0
in reply to: sirawit’s comment on: sirawit’s Shortform
IF THEY DON’T LISTEN
SCREAM LOUDER

Richard_Kennaway 28 Jul 2026 11:39 UTC
2 points
0
on: Inevitable Uncertainty in Probabilistic World Models
Once you have measured all the fish in the pond, what is the uncertainty in your model parameters uncertainty about? Well, the model parameters obv, but at that point what are the model parameters about?

Perhaps they are about the generating process that produced that pond of fish. This is uncertainty that cannot be eliminated by measuring all the fish. But it is also something you did not know about the system.

If you knew the generating process before you started, then the measurements are telling you about the sort of population that that generating process leads to. And again you did not know everything about the system. You will never know “everything” about a fishpond.

If you know you’re going to measure all the fish, do that and then eyeball the distribution. Maybe it doesn’t look normal at all, and if you want to approximate it with just a few parameters you should fit something else. What family of distributions to select is itself a parameter, and starting by eyeballing the pond may give you a bad value.

There’s also a separate question about what you are doing, in Bayesian terms, when you “eyeball” a thing and come up with a “prior”. Looks like a posterior to me, but then what was your original prior before you were born? Solomonoff?

I cannot say how any of this relates to the generating process that produced Wentworth’s apothegm.