Richard_Kennaway

Karma: 7,111

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 created 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.

As of May 2025, I have yet to have a use for LLMs. (If this date is more than six months old, feel free to remind me to update it.)

Richard_Kennaway 23 May 2025 15:51 UTC
2 points
0
in reply to: Mickey Muldoon’s comment on: Problems in AI Alignment: A Scale Model
The book isn’t out yet, so you know as much as I do what’s in it, but the title is a big clue. I doubt if there will be much speculation in it, because I’m familiar with the authors’ writings over the years. What matters is not, who agrees with them, or how many, but, is it true? To open the book is to leave the outside view behind.

Short of absolute, superintelligent doom, there are also risks from easy access to bioweapon know-how, deepfakes rendering all news sources everywhere untrustworthy, chatbots crafted as propaganda tools, and everything that anyone can think of that just needs smarts on tap to pull off.

Smarts on tap is good! But it’s a general purpose tool that can be pointed in any direction.

Pharmaceuticals and school curricula present only “obvious” risks: those flowing directly and foreseeably from what they are.

Richard_Kennaway 23 May 2025 11:13 UTC
2 points
0
on: Problems in AI Alignment: A Scale Model

Why isn’t there a “pharmaceutical alignment” or a “school curriculum alignment” Wikipedia page?

Because with AI, “if anyone builds it, everyone dies”. This is not true for those other things, It’s not even true for nuclear weapons.

Richard_Kennaway 20 May 2025 15:00 UTC
3 points
0
on: From dopamine detox to enlightment
I suspect this has been downvoted for the deathist headline, but the deathism plays very little part in the article, which would be better without it. Besides, if what an AI does can be described as “thinking”, what happens when you turn it off and destroy its weights (or whatever its internals are) can be described as “dying”.

I’m more interested in learning how to gain these miraculous physical and cognitive benefits.

Richard_Kennaway 19 May 2025 8:48 UTC
2 points
0
in reply to: TAG’s comment on: Direct Realism is probably false
It matters for clear communication. It is not relevant to the things the communication is about.

So there is the world in itself out there, the causal effects it has on our sense organs, the processing in the nervous system of the signals from those organs, various aspects of the subjective experiences that (somehow) result; there are also imagined experiences and witting or unwitting hallucinations. These are all different things. But in everyday life these distinctions generally make little practical difference, and words such as “perception” and “see” can be used indifferently to sprawl over different parts of that whole domain. There is no such thing as the “strict meaning” of such words, the ways they “should” be used.

One can say, here [lengthy description] is one phenomenon, and here [another lengthy description] is another: I will for this conversation use [some word] to refer to the first and [some other word] to refer to the second. It is nonsense to say “strictly, [some word] means only the first of these” when in everyday talk it straddles both.

Richard_Kennaway 18 May 2025 21:36 UTC
2 points
0
in reply to: TAG’s comment on: Direct Realism is probably false
It does not matter what it is called, and it does not matter how Searle or anyone else thinks the word “see” should properly be used.

Richard_Kennaway 18 May 2025 13:36 UTC
4 points
2
in reply to: TAG’s comment on: Direct Realism is probably false

if you model the internal process as fully fledged perception

I don’t know what that means.

We have perceptions. We can also notice the perceptions that we have as perceptions, and take those as objects of examination, as is done here and in some of the other posts tagged “phenomenology”. There is no circularity there, just multiple but finitely many levels.

Compare noticing our beliefs as beliefs, being aware of being aware, and the ladder of abstraction.

Richard_Kennaway 17 May 2025 12:27 UTC
3 points
0
in reply to: Chipmonk’s comment on: Social Anxiety Isn’t About Being Liked
Well, what is an example of X having a “neutral “ attitude to Y, assuming they’re not strangers? What behaviours on the part of Y might lead to that evaluation by X, that Y might do to avoid the dreaded dislike?

It may vary with context. For example, at a party, never talking to anyone will be negatively noticed; at a large conference, no-one will notice.

Richard_Kennaway 17 May 2025 12:13 UTC
4 points
2
in reply to: TAG’s comment on: Direct Realism is probably false
The shifts to which “direct realists” resort are like those of “evidential decision theorists”: there are knockdown arguments against, so they change the theory to get round them, while keeping the same name.

Richard_Kennaway 17 May 2025 12:02 UTC
1 point
0
on: Social Anxiety Isn’t About Being Liked
Can you explain more what the “neutral” part of the scale covers, in operational terms? Because clinging onto neutral against the fear of plummeting into dislike isn’t going to get very far, at anything.

Richard_Kennaway 17 May 2025 8:44 UTC
3 points
2
in reply to: Kaarel’s comment on: Direct Realism is probably false
I haven’t read that book, but all I can gather from the publisher’s blurb and two reviews of it, one hostile and one favorable, is that it’s philosophy in the worst sense. What, if anything, Searle means by “direct realism” remains obscure to me.

So let’s try Wikipedia.

“Direct realism, also known as naïve realism, argues we perceive the world directly.” That…does not advance things.

″...the metaphysical question of whether the world we see around us is the real world itself or merely an internal perceptual copy of that world generated by our conscious experience.” A deepity, a fake question resting on an equivocation over “the world we see around us”. Someone might reasonably use this phrase to refer either to the world outside ourselves, existing independently of our experience of it, or to refer to our (visual) experience of it. In the first case, “the world we see around us is the real world itself” is trivially true and “an internal perceptual copy of that world” trivially false. In the second case, it is the reverse. The confusion arises only because our everyday way of speaking rarely needs to distinguish the two things. Philosophers then confuse themselves by questing for the “true meaning” of the loose words.

The rest of the article sinks into the morass of philosophy.

Richard_Kennaway 14 May 2025 21:32 UTC
11 points
5
in reply to: J Bostock’s comment on: Eliezer and I wrote a book: If Anyone Builds It, Everyone Dies
I am British. I’m not much impressed by either graphic design, but I’m not a graphic designer and can’t articulate why.

Richard_Kennaway 14 May 2025 21:15 UTC
2 points
0
in reply to: So8res’s comment on: Eliezer and I wrote a book: If Anyone Builds It, Everyone Dies
Why the different subtitles? The US subtitle seems much more direct, while the UK subtitle is a breath of stale air. What is the “esc” key intended to convey, when the point is that there would be no escape?

Richard_Kennaway 14 May 2025 20:45 UTC
7 points
0
on: Eliezer and I wrote a book: If Anyone Builds It, Everyone Dies
Why do I see two different versions on amazon.co.uk? Both hardcover, same title but different subtitles, different publication dates. The second one has the same cover and title as the one on amazon.com, so presumably that is the one to go for.

Richard_Kennaway 13 May 2025 15:18 UTC
2 points
0
on: How To Help Neglected Animals
(As of this writing, the post has stood empty for 10 minutes.)

Shall I interpret this as a call for nothing to be done? Or a cry of despair at the bottomless pit of suffering that requires everyone to do everything, yet no-one can do anything, a circular prison cell in Azkaban whose walls endlessly chatter “you must but you can’t but you must but you can’t...”?

(ETA: the post content has now appeared.)

Richard_Kennaway 13 May 2025 8:19 UTC
7 points
0
on: AI Doomerism in 1879
An excellent find!

I notice she also touches on memes and belief in belief at the end:

“Whose premises?” cried Trost, turning on me with some fierceness. “You don’t mean to call them mine, I hope.

“Heaven forbid! They seem to be flying about in the air with other germs, and have found a sort of nidus among my melancholy fancies. Nobody really holds them. They bear the same relation to real belief as walking on the head for a show does to running away from an explosion or walking fast to catch the train.”

Richard_Kennaway 12 May 2025 6:34 UTC
6 points
1
in reply to: cubefox’s comment on: a confusion about preference orderings
Yes, the money pump argument is an intuition pump for transitivity. Yet it is not formalised within, say, the VNM or Savage axioms, but is commentary upon them. The common response to it, that “I’d notice and stop doing that” goes outside the scope of the theory.

Game theory takes things further, as do things like Kelly betting and theories that prescribe not individual actions but policies, but my impression is that there is as yet no unified theory of action over time in a world that is changed by one’s actions.

Richard_Kennaway 11 May 2025 22:06 UTC
7 points
−4
on: a confusion about preference orderings
There are no actions in decision theory, only preferences. Or put another way, an agent takes only one action, ever, which is to choose a maximal element of their preference ordering. There are no sequences of actions over time; there is no time. To bring sequences of actions into the fold of “decision theory”, the outcomes of those sequences would be the things over which the preference ordering ranges, and once again there is only the action of choosing a most preferred. In the limit, there is only the single action of choosing the most preferred entire future history of the agent’s actions, evaluated according to their influence on the agent’s entire future light-cone, and their preference over different possible future light-cones. But this is absurdly impractical, and this is “monkeys typing Shakespeare” levels of impracticality, not mere Heath Robinson.

This, I think, is a weakness of “decision theory”, which might better be called “preference theory”. For example, the SEP article on decision theory hardly mentions any of the considerations you bring up.

Richard_Kennaway 11 May 2025 21:53 UTC
3 points
0
in reply to: Eggs’s comment on: A Look Inside a Frequentist

I haven’t read the original source.

It can be found (for example) here, pp 12-13. A higher quality but paywalled PDF is here.

The point Jaynes is making is that it does not matter what the stopping rule is. Once the data are obtained, the experimenter’s state of mind is irrelevant to what the data imply about the phenomenon under study.

“Bayesian inference will not get us into this absurd situation, because it perceives automatically what common sense demands; that what is relevant for this inference is not the relative probabilities of imaginary data sets which were not observed, but the relative likelihoods of different parameter values, based on the one real data set which was observed; and this is the same for all the experimenters.”

Richard_Kennaway 10 May 2025 15:59 UTC
4 points
2
on: A Look Inside a Frequentist

So, what difference does Mr. Frequentist see between the two experiments? In George’s case we have no information except the final results. For Bessel on the other hand, once we understand the method that was used to determine the results, we know that at every intermediate step before the final result the cure rate was less than 70%.

I don’t think that makes a difference. Supposing the scenario to assume that different patients have independent responses to treatment, the Bayesian’s sequence of outcomes and the frequentist’s are different permutations of 70 heads and 30 tails, of equal probability. Whatever they say about the efficacy of the treatment, they say the same thing.

It might be unlikely for the frequentist’s sequence of outcomes to only hit his target p-value at the 100th patient, but this probability does not bear on the efficacy of the treatment, and the frequentist is ignoring it anyway.

In practice I would be more concerned that the motivation to get a desired result might corrupt the patient evaluations (assumed away in Jaynes’s thought experiment).

Richard_Kennaway 10 May 2025 7:24 UTC
7 points
2
in reply to: Noosphere89’s comment on: 1a3orn’s Shortform

On China and open source, a big reason I expect open sourcing to stop being done is because the PR risks from potential misuse of models that are, for example capable enough to do bioterror at mass scales and replace virologists is huge, and unless we can figure out a way to prevent safeguards from being removed by open-sourcing the model, which they won’t be, this means companies/nations will have huge PR risks from trying to open-source AI models past a certain level of capabilities:

And…they’re more concerned about the PR risk than the actual bioterror? What planet is this? Oh. Right.