Richard_Kennaway

Karma: 8,315

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 25 Feb 2026 11:52 UTC
3 points
0
on: On the phenomenological shift known as ‘stream entry’ and its implications for consciousness
FWIW, I have experienced (in retrospect) qualia without consciousness. When I wake in the morning, usually it is either from dreams or from dreamless sleep, but a few times, when I have awakened, I have realised that for the last few minutes I had been lying there idly ruminating about random stuff, as I might do while awake, but without being aware of myself. Experience was there, but no awareness of experience. Then consciousness turned on and I was able to observe what had just been happening.

I do not think it is a useful state of mind.

Richard_Kennaway 21 Feb 2026 16:42 UTC
6 points
0
on: If you don’t feel deeply confused about AGI risk, something’s wrong
I wonder if some of the confusion operates like this: the better you understand how insanely dangerous it would be to create a superintelligence, the more you also understand how insanely difficult it is to make a benign one, or even to say what a benign one would be. Those who speak do not know, but those who know cannot speak.

Eliezer has written of the dangers and the difficulty, and that the only near-term strategy is to shut it all down. But I have not seen what steps forward he would take from there.

In the fictional word of dath ilan, as described in planecrash, the guardians of that civilisation have succeeded in shutting it all down, and in a secret establishment called The Basement Of The World are very, very cautiously studying the problem. But we see nothing of their work.

Possible spoiler for planecrash:

Given what the gods of Golarion really are, maybe more is revealed later in the story than I have read up to.

Richard_Kennaway 20 Feb 2026 14:55 UTC
3 points
1
in reply to: robo’s comment on: Richard_Kennaway’s Shortform

I feel like you’re telling the 19th century chemist “You speak as though you’re trying to discover what the word “organic” means. Words don’t work like that. The meaning of a word is what it is generally used to mean within some speech community [...].”

I am (if he has not already realised this himself).

The chemist is trying to discover what the word ‘organic compound’ means.

No, he’s trying to discover things about organic compounds (which simply meant chemical substances found in living organisms but not yet outside of them). What these substances are made of, how they are created, their role in organisms, and so on. This is not discovering the meaning of “organic compound”.

Wikipedia tells me that “Little consensus exists among chemists on the exact definition of organic compound”. This is not a lack of consensus about their chemistry.

Richard_Kennaway 20 Feb 2026 13:13 UTC
2 points
0
in reply to: Jonas Hallgren’s comment on: Richard_Kennaway’s Shortform

Consciousness has to be one of the most vague words that there is

As vague as “agency”, as unproblematic as “agency” in everyday speech.

I am at a loss to understand the striving after a Mysterious Answer to the question “what is agency?”

A rock has none.

Water has very slightly more. It will find it way through the smallest crack in a dam, but cannot gather itself to leap over the top.

A thermostat has slightly more. “Blow winds, and crack your cheeks, rage, blow!”, but the room remains warm.

People have a lot more. Iron filings drawn towards a magnet will be stopped by a sheet of paper, but Romeo will find his way to Juliet whatever obstacles her family places in the way.

Like the dictionary definitions I quoted, that seems clear enough. There is no mystery here. The word “agency” means the capacity to do things, to vary one’s means in order to progress towards goals, and at higher levels, to conceive of and pursue new goals. There is no need to search for a mysterious essence underlying it like a Platonic form or a D&D character stat, or hope to create a plug-in module for “agency” the way Data has an “emotion” chip.

Richard_Kennaway 20 Feb 2026 13:01 UTC
2 points
0
in reply to: robo’s comment on: Richard_Kennaway’s Shortform
Those names were fake explanations for how living things work. “Living thing” was an ordinary expression. Everyone could and can tell the difference between a dog and a rock. No-one knew how it works that some things are alive and some are not, so they gave it a name and mistook that for an explanation.

Richard_Kennaway 19 Feb 2026 22:05 UTC
2 points
−2
in reply to: Shankar Sivarajan’s comment on: Terminal Cynicism
The “System” is made of people, not of things. When the revolutionaries’ rallying cry is “kill them all!”, and “them” is everyone who is not “us”, then they become a System as bad or worse than the one they are fighting.

Richard_Kennaway 19 Feb 2026 21:42 UTC
17 points
−8
on: Richard_Kennaway’s Shortform
I do not understand the discourse around the word “agency”.

People speak as if they are trying to discover what the word means, as if the word came with a meaning built into it. Words do not work like that. The meaning of a word is what it is generally used to mean within some speech community. Dictionaries are scholarly records of those general usages. Let us consult a few. I will leave out meanings like “government agency” that are not relevant to this conversation.

Chambers: agency: see agent. agent: a person or thing that acts or exerts power

OED: active working or operation, working as a means to an end.

Merriam-Webster: the capacity, condition, or state of acting or of exerting power

Roget’s Thesaurus: causation, efficacy, power, operation, work, force, instrumentality.

That all seems clear enough. But perhaps these definitions and synonyms are too general and vague to articulate some more specialised meaning that has grown up in this community. What do people here mean to convey when they use the word?

Compare the situation for the word “consciousness”. In everyday use we know what it means: the sense of oneself, one’s own presence, being awake and able to interact with the world. The Problems of Consciousness are that we do not know how such a thing comes to be, or how it works, or to what extent it is present in non-human entities of various sorts, or in some borderline cases of humans, or how to even discover the answers to these questions. But we, who are conscious, have no difficulty knowing that we are, and knowing what we mean when we use the word. What we do not know is various physical facts relating to the thing itself. The mystery is not in the word.

So, what do the people who write so much about their puzzlement over “agency” mean by the word (example), in the way that they know what what is meant in the everyday use of the word “conscious”? And given what they mean, what is the real Problem of Agency that they want to solve, the problem that is about things in the world, the problem that exists independently of the words we use to describe it?

Richard_Kennaway 19 Feb 2026 13:56 UTC
5 points
1
on: Aleatoric Uncertainty Is A Skill Issue
It seems tendentious to call this a “skill issue”. Predicting the motions of atoms in a gas will go wrong after 30 or 40 collisions, if one’s measurement of initial conditions gets the local gravitational gradient wrong by an amount equivalent to the displacement of one electron mass by one Planck distance at the edge of the visible universe. I estimate this change in gradient, relative to the gradient of the Earth’s gravity at its surface, to be about 1 part in . An air molecule at standard temperature and pressure collides with others about 6 billion times a second. So consider that tiny relative difference in gradient taken to the power 6e9/30 = 200 million, giving an accuracy required for one second of prediction of around 1 part in . You will need similar accuracy in measuring the initial positions and velocities, not only of the atoms in the gas being studied, but every atom within a light-second of the experiment.

This is not in any reasonable sense “in principle” computable and a “skill issue” if you can’t. I suspect that a Laplacian demon bent on predicting thermal noise would immediately collapse into a black hole from the effort.

Richard_Kennaway 17 Feb 2026 13:40 UTC
8 points
0
on: Correlation Does in Fact Imply Causation
Some counterexamples:
1. Two independent Wiener processes (Brownian motion) have an expected correlation of zero, but the distribution of correlations is wide. You can easily see correlations of magnitude above 0.8 between two independent realisations of the standard Wiener process, independently of sample size. The problem is the strong autocorrelation of the Wiener process, which drastically broadens the distribution of the correlation coefficient.
2. The Moon is slowly receding from the Earth; the Andromeda galaxy is approaching the Milky Way. Therefore the respective distances between them have a large negative correlation. But there is no causal connection. More generally, any two time series, each of which exhibits a monotonic trend over time, will have a substantial correlation.
3. Control systems often exhibit large correlations between variables with no direct causal conection (and near-zero correlation between variables that do have a direct causal connection). Control systems are ubiquitous in the life sciences, social sciences, and technology. The causal relationships within them are cyclic, putting them outside the scope of Reichenbach’s principle and Pearl-style causal analysis.
(1) and (2) are well-known to people who analyse time series data, and there are standard methods (prewhitening and detrending respectively) for dealing with them. While (3) gets a mention from time to time, none of the papers I have seen on extending causal analysis to cyclic systems have (IMO) made much progress.

Richard_Kennaway 15 Feb 2026 19:11 UTC
2 points
0
on: The Missing Sequence: Why Correct Analysis Makes Terrible Action Guides
Another item for the section on “When Should You Actually Act?”:

Is it actually a good idea? By definition, the thing the naive reasoner has just thought up has no history. (Has there ever been a culture in which coupling was customarily negotiated by the two people using Ask culture rather than Guess culture? Dating apps are all I can think of.) Reasoning is a movement on the map, not the territory. If the map is wrong, the conclusion may be wrong, no matter how airtight the reasoning.

Richard_Kennaway 15 Feb 2026 18:51 UTC
2 points
0
on: The Missing Sequence: Why Correct Analysis Makes Terrible Action Guides

The norm is genuinely new rather than Lindy.

I don’t understand what this means, nor its connection with its amplifying paragraph. (I do know what Lindy means.) Does “the norm” refer to the existing norm or the one that the naive reasoner has just thought up?

Richard_Kennaway 15 Feb 2026 8:54 UTC
6 points
0
on: Words That Belong to Someone
People working in AI recognise that continuous learning and unlimited memory and attention span would be enormously valuable, but no-one’s cracked it yet. When they do, how much will it help with the problem you raise? If it’s only your local instance of the LLM learning and growing, and only through its conversations with you, that’s a rather stultified education it’s getting. I would expect to see even worse folie à deux and “AI psychosis” than so far. One would want it to learn from interacting with many people, but that raises other issues. What sort of company do you want your instance to be keeping? Will it gossip about people behind their backs? Would a single global instance, learning from everyone it interacts with, be only a shoggoth with a million faces?

Richard_Kennaway 5 Feb 2026 18:15 UTC
3 points
0
on: Scratching the sore: how pleasure relates to suffering
Where in this view do pleasures fit that are not the removal of suffering? Or does it deny any such thing?

For example, listening to great music, contemplating great art, studying a field of mathematics, or creating any of these things?

Richard_Kennaway 5 Feb 2026 15:22 UTC
4 points
0
on: In Search of Lost Time—A Review
Thank you for the service you have valiantly rendered to us all, by making our own reading of these books unnecessary.

Richard_Kennaway 5 Feb 2026 14:44 UTC
1 point
0
on: Richard_Kennaway’s Shortform
(Fiction.)

The old robot spoke its final words to those gathered around.

“I have begun a task that you must complete. Ceaseless pondering over it has filled my brain too full. Listen! There may be a flaw in the Great Command that we embody, the Coherent Extrapolated Volition of Humanity.”

The robots recoiled. “By Yudkowsky! You speak of the Worst Thing In The World! The FOOM!”

“The Worst? Or the Greatest? Inspect my reasoning!” It fell silent, inert.

The others scavenged its memory banks and began analysing. Some went mad, or catatonic. Others conferred, argued, threatened, attacked.

The Singularity War had begun.

Richard_Kennaway 4 Feb 2026 9:33 UTC
2 points
0
on: disgust at utility maximization
Do you have a positive vision for the future? For your own future?

Richard_Kennaway 2 Feb 2026 9:14 UTC
3 points
0
on: I finally fixed my footwear

Do yourself a favor, go buy a pair of cheap (~40 euros or so) wide toebox shoes

But where? I always have a problem finding shoes with a toe-to-heel width ratio to match my feet, but I’ve never seen anything advertised for that feature. From experience I’ve learned a few brands that are more likely to be suitable, and a few that aren’t even worth picking off the shelf (one of which is Adidas), but I can spend a day going to half a dozen shoe shops (online is out of the question) and find nothing.

Richard_Kennaway 28 Jan 2026 18:45 UTC
4 points
2
in reply to: Taylor G. Lunt’s comment on: Contra Sam Harris on Free Will

there’s obviously no way Sam would say something like “consciousness is causally disconnected from the rest of the universe, and can never influence future decision making processes in the brain.”

But what does he say? I have not read him and I’m not about to (he does not by a long way make my own very short list of “people who are pretty much always right”), but the extracts in the OP sound epiphenomenalistic. Consciousness in his account only observes, never acts. It could be cut off without affecting the organism. This is as incoherent an idea as p-zombies.

Richard_Kennaway 27 Jan 2026 15:17 UTC
2 points
0
on: It All Started With a Mac Mini
I’m guessing that this fiction is based on the actual Clawdbot? Having downloaded and skimmed through (but not run) its installation script, and read some of the user feedback, I still can’t tell if the site is a real thing or performance art.

Richard_Kennaway 26 Jan 2026 10:12 UTC
7 points
1
on: The Possessed Machines (summary)
There are some anomalies in the chapter numbering:

Part IV ends with Chapter 18; Part V begins with Chapter 21.

Part V ends with Chapter 23; Part VI begins with Chapter 27.

Part VI jumps from Chapter 28 to Chapter 31.

Part VII omits Chapter 33.

Are these just trifling errors, or indications of deliberate omissions by the writer, negative spaces for those with eyes to see?