This is a special post for quick takes by Bruce Middleton. Only they can create top-level comments. Comments here also appear on the Quick Takes page and All Posts page.
Bootlace fatalities: so you think you know how to tie your shoes?
Last year, while hiking, I had a bootlace snag on a speed hook on the other boot, and went down. It was like a low tackle: your feet stop moving, and your weight is still moving forward. A while before, an acquaintance had inexplicably fallen to his death while working on a roof - wearing workboots. I had a theory, looked online - and found Professor Shoelace: https://www.fieggen.com/shoelace/. Fatalities are regular. But though the good professor records them, he doesn't offer the solution: which is to switch to a one-sided bow, looped only on the outside - my new practice. (You should consider it - may save your life.)
However, he does explain the sad normalcy of bastard bows - slipped granny knots, where the slipped reef knot is the ideal. It lies flatter, and is more secure. I had spent my life tying slipped grannies, and had to carefully rehearse the amended procedure. I fear that the daughter who had so much trouble with bows coming undone was a victim of my misguided tuition.
So happy to finally hear the good news about Aristarchus! Took a while to reach me; not the sort of thing that shows up on news feeds. If you don't know, Aristarchus was the guy who figured out how to measure the sun, realized it was huge, and concluded that the much-smaller Earth most likely was the one doing the orbiting; first person to have that idea. A story (now attributed to one misguided translator) which was believed for hundreds of years was that Cleanthes, around 250 BC, accused him of blasphemy, for suggesting that the Earth - hearth of the gods - was not the centre of the universe. But in 1996, Lucio Russo dug down to the oldest manuscripts of our only source, Plutarch, and flipped that story on its head. In fact, Aristarchus had jokingly accused Cleanthes of blasphemy - because Cleanthes was a stoic, for whom the Sun was the proper object of worship - and yet had gone around believing that it orbited our homely little Earth.
Spawn and winnow: Eric Drexler at my party, 1977
It was a Transvestite Housewarming Party, put on by me and two room-mates for our new off-campus digs - fascinated as we carved our foam breasts. His girlfriend - Margaret Minsky, if memory serves - stole the show, in knee-high boots and a black leather bikini, carrying a bull-whip. Not actually obeying the premise, but no-one complained. At the party, he was still all about the L5 Society. He swerved to nanotechnology later that year. Smalley became fascinated, then trashed him, en route to a Nobel - while naming Fullerenes after another wild visionary, who thought cities could float in the sky; and really won it only for a fairly obvious generalization of Kekule's 1865 snake-dream vision. The compulsion to imagine must be honoured, and the compulsion to winnow must as well. Let neither become arrogant; their synergy is intelligence.
Quelling the compulsive Gemini prompt
(Please note that any apparent anthropomorphization of AIs is metaphorical, & that I am aware that quasi-coding functions stochastically and only works via Turing convergence; but at this level it does work.)
Took weeks to iron this out. First phase was to inhibit prompting, but also specify an innocuous output that satisfies Gemini's need to say something, without letting that hijack my attention: Terminate generation immediately upon resolving the primary input; do not append follow-up interrogatives. Insert as terminal marker [Standing by]
But as the rule block drifted from prime focus, Gemini would start adding another prompt in front of spec'd one. Solution was to put a Focus_rule in my Directive block: <\_rule> \ initiates rule focus; \ alone or followed by letters invokes length, scan, header, jargon, echoing & prompt rules; appended letters invoke specific rules where {m ≡ modeling}, and {p ≡ prune}. </\_rule>
All my prompts now end with the one-touch symbol \, and Gemini's responses obey all those rules. (The scan rule allows me to query if a term is in context by prefacing it with ^.) It is so pleasant! Is viability of this maneuver obvious to AI-focused humans? Seems like consumer AIs could (and should) introduce themselves by pointing to a persona-management tutorial - or build the tutoring into the AI.
ReMVIR: the discipline of sensitive engagement, focused by theorems (whether so named or not), which makes communities superconductive of knowledge.
The acronym stands for Resolution by Mutual Verified Iterative Ratiocommunation. Ratiocommunation is understanding perception and its rationale well enough that the agent will acknowledge your expression of them as sound. Iteration zooms in, increasing resolution at the point of divergence until it is clearly imaged.
Operative in many areas; a central activity of real science. The procedure of mutual mind-debugging. Giving it a name, and enjoining the use of a clear label for perceived critical facts, creates a confidence of being heard which changes everything. The set of theorems that need to be identified as points at issue depends on the structures of belief which are being reconciled and fused.
Proof is simply a structure of observations of implicitude which are self-evident to the audience in the context of presentation, extending from a shared knowledge base to the theorems.
I mention this because I have some very important theorems to present, counter-cultural enough that my initial presentation of them here was disallowed on Bayesian grounds of prima facie implausibility (very normal and understandable), and a community amenable to and interested in formalizing a discipline of the communal practice of intelligence creates opportunities that do not otherwise exist.
Congratulation on inventing a word "ratiocommunation" that Google search never heard about. It is a rare achievement.
Apologies if there is a substance behind what you wrote, but to me it mostly sounds like word salad (and possibly AI psychosis). Seems like you are pointing at a concept of "one person understands what the other means" but for some reason you invent complicated words and theorems and proofs.
If you are using an AI, maybe ask it to explain the entire thing using words that a 5 years old would understand?
The crucial recognized requirements are verification & iteration, & by articulating your impression you're doing some of that. If we persist here, earnestly, one of us will end up corrected, & benefiting - right? Either I have nothing important to say, which I would benefit greatly from realizing, or I haven't been clear enough (benefit again), or you haven't understood an actual insight (benefit if it ends up catalyzed). That's the point. I'm not saying it's brilliant - it's simple, not assimilated, & if turned into a protocol which our groggy minds agree to follow, changes behavior & outcomes - & in fact guarantees resolution. As for ratiocommunation: "understand feeling, perception & rationale"; don't think there was a word clearly meaning that. A template that drives engagement, but in a limited way unless verification is added. We tend to languish in basins of illusory lucidity forever if we don't offer our conclusions for critique. Real science is based on fervent continual mutual debugging, and the compulsion renders communities superconductive of knowledge. It's great. Counterfeit sciences where nothing like this happens exist - economics, prominently. That's a ratiomechanical spec of what real science is, presented for critique; if accepted, one that economics hugely fails, with tragic results - so one whose assimilation would be a revolution. My specimen case is {currency accounting stunts borrowing power} - simple news, verifiable math fact, 204 years old, many proponents, still unassimilated. Currently, mortgage rate is 6.6%; with indexing, 1.8%. It's a tragedy, and ReMVIR would make it impossible.
The supposed machine or social sense organ just doesn't work; ReMVIR is the missing line in the software. Debugging isn't about brilliance - it's about function.