Anthony Repetto

• Anthony Repetto 18 Nov 2023 6:54 UTC

  −1 points

  0 ∶ 0

  on: What did AI Safety’s spe­cific fund­ing of AGI R&D labs lead to?

  I expect that, once AGI exists, and flops, the spending upon AGI researchers will taste sour. The robots with explosives, and the surveillance cameras across all of China, really were the bigger threats than AGI X-risk; you’ll only admit it once AGI fails to outperform narrow superintelligences. The larger and more multi-modal our networks become, the more consistently they suffer from “modal collapse”: the ‘world-model’ of the network becomes so strongly-self-reinforcing, that ALL gradients from the loss-function end-up solidifying the pre-existing world-model. Literally, AIs are already becoming smart enough to rationalize everything; they suffer from confirmation bias just like us. And that problem was already really bad, by the time they trained GPT-4 - go check their leaked training-regiment: they had to start-over from scratch repeatedly, because the brain found excuses for everything and performance tanked without any hope of recovery. Your AGI will have to be re-run through training 10,000 times, before one of the brains isn’t sure-it’s-always-right-about-its-superstitions. Narrow makes more money, and responds better, faster, cheaper in war—there won’t be any Nash Equilibrium which includes “make AGI”, so the X-Risk is actually ZERO.

  Pre-ChatGPT, I wrote the details on LessWrong: https://​​www.lesswrong.com/​​posts/​​Yk3NQpKNHrLieRc3h/​​agi-soon-but-narrow-works-better

• Anthony Repetto 18 Nov 2023 6:40 UTC

  −1 points

  0 ∶ 0

  in reply to: Solomon Ucko’s comment on: Vot­ing The­ory has a HOLE

  Thank you for diving into the details! And, to be clear, I am not taking issue with any of Gibbard’s proof itself—if you found an error in his arguments, that’s your own victory, please claim it! Instead, what I point to is Gibbard’s method of DATA-COLLECTION.

  Gibbard pre-supposes that the ONLY data to be collected from voters is a SINGULAR election’s List of Preferences. And, I agree with Gibbard in his conclusion, regarding such a data-set: “IF you ONLY collect a single election’s ranked preferences, then YES, there is no way to avoid strategic voting, unless you have only one or two candidates.”

  However, that Data-Set Gibbard chose is NOT the only option. In a Bank, they detect Fraudulent Transactions by placing each customer’s ‘lifetime profile’ into a Cluster (cluster analysis). When that customer’s behavior jumps OUTSIDE of their cluster, you raise a red flag of fraud. This is empirically capable of detecting what is mathematically equivalent to ‘strategic voting’.

  So, IF each voter’s ‘lifetime profile’ was fed into a Variational Auto-Encoder, to be placed within some Latent Space, within a Cluster of similarly-minded folks, THEN we can see if they are being strategic in any particular election: if their list of preferences jumps outside of their cluster, they are lying about their preferences. Ignore those votes, safely protecting your ballot from manipulation.

  Do you see how this does not depend upon Gibbard being right or wrong in his proof? As well as the fact that I do NOT disagree with his conclusion that “strategy-proof voting with more than two candidates is not possible IF you ONLY collect a SINGLE preference-list as your one-time ballot”?

• Anthony Repetto 16 Mar 2023 2:02 UTC

  −5 points

  2 ∶ 4

  in reply to: Nathan Young’s comment on: Time Ar­ti­cle Dis­cus­sion—“Effec­tive Altru­ist Lead­ers Were Re­peat­edly Warned About Sam Bankman-Fried Years Be­fore FTX Col­lapsed”

  “which would be a huge hassle and time cost for whoever speaks out”

  Wait—so if leaders were complicit, yet admitting to that would be a hassle, then it’s better that they not mention their complicity? I’m afraid of a movement that makes such casual justifications for hiding malefactors and red flags. I’m going to keep showing outsiders what you all say to each other! :O

• Anthony Repetto 28 Feb 2023 6:01 UTC

  1 point

  0 ∶ 0

  in reply to: Matt_Sharp’s comment on: Bayes is Out-Dated, and You’re Do­ing it Wrong

  You’re welcome to side with convenience; I am not commanding you to perform Dirichlet. Yet! If you take that informality, you give-up accuracy. You become MoreWrong, and should not be believed as readily as you would like.

• Anthony Repetto 27 Feb 2023 20:03 UTC

  −4 points

  0 ∶ 0

  in reply to: Matt_Sharp’s comment on: Bayes is Out-Dated, and You’re Do­ing it Wrong

  “From this informal perspective, clarity and conciseness matters far more than empirical robustness.”

  Then you are admitting my critique: “Your community uses excuses, to allow themselves a claim of epistemic superiority, when they are actually using a technique which is inadequate and erroneous.” Yup. Thanks for showing me and the public your community’s justification for using wrong techniques while claiming you’re right. Screenshot done!

• Anthony Repetto 26 Feb 2023 2:48 UTC

  −1 points

  0 ∶ 0

  Error

  The value
    NIL
  is not of type
    SIMPLE-STRING
  when binding #:USER-ID163

Bayes is Out-Dated, and You’re Do­ing it Wrong

• Anthony Repetto 24 Feb 2023 11:47 UTC

  6 points

  0 ∶ 0

  in reply to: alisimmonds’s comment on: Lev­er­age Re­search: re­view­ing the ba­sic facts

  Alistair, I regret to inform you that after four years of Leverage’s Anti-Avoidance Training, the cancer has spread: the EA Community at large is now repeatedly aghast that outsiders are noticing their subtle rug-sweeping of sexual harassment and dismissal of outside critique. In barely a decade, the self-described rats are swum ‘round a stinking sh!p. I’m still amazed that, for the last year, as I kept bringing-forth concerns and issues, the EA members each insisted ‘no problems here, no, never, we’re always so perfect....’ Yep. It shows.

• Anthony Repetto 24 Feb 2023 11:28 UTC

  8 points

  0 ∶ 0

  in reply to: LarissaHeskethRowe’s comment on: Lev­er­age Re­search: re­view­ing the ba­sic facts

  This aged well… and it reads like what ChatGPT would blurt, if you asked it to “sound like a convincingly respectful and calm cult with no real output.” Your ‘Anti-Avoidance,’ in particular, is deliciously Orwellian. “You’re just avoiding the truth, you’re just confused...”

  I was advocating algal and fish farming, including bubbling air into the water and sopping-up the fish poop with crabs and bivalves—back in 2003. Spent a few years trying to tell any marine biologist I could. Fish farming took-off, years later, and recently they realized you should bubble air and catch the poop! I consider that a greater real-world accomplishment than your ‘training 60+ people on anti-avoidance of our pseudo-research.’ Could you be more specific about Connection Theory, and the experimental design of the research you conducted and pre-registered, to determine that it was correct? I’m sure you’d have to get into some causality-weeds, so those experimental designs are going to be top-notch, right? Or, is it just Geoff writing with the rigor of Freud on a Slack he deleted?

• Anthony Repetto 24 Feb 2023 7:21 UTC

  −4 points

  0 ∶ 0

  in reply to: Davis_Kingsley’s comment on: EA’s weird­ness makes it un­usu­ally sus­cep­ti­ble to bad behavior

  Third Generation Bay Area, here—and, if you aren’t going to college at Berkeley or swirling in the small cliques of SF among 800,000 people living there, yeah, not a lot of polycules. I remember when Occupy oozed its way through here that left a residue of ‘say-anything-polyamorists’ who were excited to share their ‘pick-up artist’ techniques when only other men where present. “Gurus abuse naïve hopefuls for sex” has been a recurring theme of the Bay, every few decades, but the locals don’t buy it.

• Anthony Repetto 24 Feb 2023 7:15 UTC

  7 points

  2 ∶ 0

  in reply to: Burner1989’s comment on: EA’s weird­ness makes it un­usu­ally sus­cep­ti­ble to bad behavior

  I am terrified that you were downvoted to obscurity. These posts, the ones that EA hides, are the ones the public needs to see the most.

• Anthony Repetto 24 Feb 2023 5:17 UTC

  1 point

  1 ∶ 3

  in reply to: titotal’s comment on: EA, Sex­ual Harass­ment, and Abuse

  I am terrified that you were so thoroughly downvoted… “EA only wants to hear shallow critiques, not deep ones” seems to be happening vigorously, still.

• Anthony Repetto 24 Feb 2023 5:01 UTC

  2 points

  2 ∶ 2

  in reply to: AnonymousEAForumAccount’s comment on: EA, Sex­ual Harass­ment, and Abuse

  It’s a bad sign that you were being downvoted! I gave you my upvote!

• Anthony Repetto 18 Feb 2023 22:14 UTC

  1 point

  0 ∶ 0

  in reply to: Julia_Wise’s comment on: Make the Drought Eva­po­rate!

  Another wonderful example of “so simple, why didn’t anyone try it before” just this week:

  Robert Murray-Smith’s wind generators seem to have a Levelized Cost comparable to the big turbines, yet simple and cheap, redundant:

Down­vote Out­siders’ Cri­tiques, to Prove Our Bias

• Anthony Repetto 22 Jan 2023 20:51 UTC

  −4 points

  1 ∶ 2

  in reply to: Devon Fritz’s comment on: Do­ing EA Better

  Thank you! I remember hearing about Bayesian updates, but rationalizations can wipe those away quickly. From the perspective of Popper, EAs should try “taking the hypothesis that EA...” and then try proving themselves wrong, instead of using a handful of data-points to reach their preferred, statistically irrelevant conclusion, all-the-while feeling confident.

• Anthony Repetto 2 Jan 2023 5:50 UTC

  2 points

  1 ∶ 2

  in reply to: Remmelt’s comment on: Re­ac­tive de­val­u­a­tion: Bias in Eval­u­at­ing AGI X-Risks

  continuing my response:

  When Gregory Lewis said to you that “If the objective is to persuade this community to pay attention to your work, then even if in some platonic sense their bar is ‘too high’ is neither here nor there: you still have to meet it else they will keep ignoring you.” He is arguing an ultimatum: “if we’re dysfunctional, then you still have to bow to our dysfunction, or we get to ignore you.” That has no standing in epistemics, and it is a bad-faith argument. If he were to suppose his organization’s dysfunction with the probability with which he askes you to doubt your own work, he would realize that “you gotta toe the line, even if our ‘bar’ is nonsense” is just nonsense! Under the circumstance where they are dysfunctional, Gregory Lewis is lounging in it!

  The worst part is that, once their fallacies and off-hand dismissals are pointed-out to them, when they give no real refutation, they just go silent. It’s bizarre, that they think they are behaving in a healthy, rational way. I suspect that many of them aren’t as competent as they hope, and they need to hide that fact by avoiding real analysis. I’d be glad to talk to any Ai Safety folks in the Bay, myself—I’d been asking them since December of last year. When I presented my arguments, they waved-away without refutation, just as they have done to you.

• Anthony Repetto 2 Jan 2023 5:40 UTC

  0 points

  1 ∶ 1

  in reply to: Remmelt’s comment on: Re­ac­tive de­val­u­a­tion: Bias in Eval­u­at­ing AGI X-Risks

  Thank you for speaking up, even as they again cast doubt: where Gregory Lewis supposed that the way to find truth was that “We could litigate which is more likely—or, better, find what the ideal ‘bar’ insiders should have on when to look into outsider/​heterodox/​whatever work, and see whether what has been presented so far gets far enough along the ?crackpot/​?genius spectrum to warrant the consultation” He entirely ignores the proper ‘bar’ for new ideas: consideration of the details, and refutation of those details. If refutation cannot be done by them, then they have no defense against your arguments! Yet, they claim such a circumstance is their victory, by supposing that some ‘bar’ of opinion-mongering should decide a worthy thought. This forum is very clearly defending its ‘tuft’ from outsiders; the ‘community’ here in the Bay Area is similarly cliquish, blacklisting members and then hiding that fact from prospective members and donors.

Sail Over Moun­tains of ICE

• Anthony Repetto 21 Nov 2022 12:25 UTC

  3 points

  0 ∶ 0

  in reply to: Julia_Wise’s comment on: Make the Drought Eva­po­rate!

  Whoo. Last cross-post for the night, I think I’ve responded to the major points… and I hope this shows a bit more of the complexity underneath my simplistic presentation!

  How quickly it rains down depends on a few factors, and we can tip those in our favor:

  --> Humid Rise—humidity (just the h2o molecule) is only 18g/​mol, while oxygen molecules are 32g/​mol, so humid air is quite buoyant! Especially considering that water vapor reflects heat (infrared) back to the ground, creating a heat bulge beneath it. The result is that, once humidity begins to rise, it naturally pulls air in from all around it, along the ground. It begins to drive convection. Yet! That humid rise is normally billowy and easily dispersed by cross-breezes, which means that the humidity cannot rise high quickly; it mostly travels far overland, or stays in place. Your rain wanders to an unexpected location! We want to form rain clouds nearby, instead, so we need that humidity to rise really high, quickly, without being torn apart by cross-breezes. That’s where the solar concentrators help, with their tall tower at 1200C and radiant, they blast infrared into all the water vapor around them, pummeling a plume high up, carrying that vapor. Up high enough, the air pressure drops, which is key for causing a rapid cooling, and the formation of nice heavy clouds. The faster we take air from the ground up to a few kilometers, the more water it’ll still be holding. [[Only a fraction of one gram per m3 is needed for the thinnest clouds, but we could toss a few grams up and it’ll come down soon. We want the water to rain, evaporate, and rain down again, in as many cycles as it can. That gives plants time to grab it, in numerous locations, as well as time for the ground to catch some.]] When we look at water-demand for plants in the wild vs. water-resilient greenhouses, we can drop water demand ten-fold because nine-tenths of the water was lost in the leaves to evapotranspiration! As a result, if that leaf-sweat keeps rising and falling as rain as it travels further South, then the same bucket of water ends up getting ten times the use (assuming ground water is eventually used, as well).

  --> Albedo—the desert rock is pretty bright, so the addition of vegetation and especially any water-bodies (!) will multiply the solar absorption, which will drive that heat-bulge and evaporation for humidity-buoyancy, to help loft water vapor and form clouds. This is how the Amazon does it—most of her clouds are her armpit fog, caused by solar-to-thermal foliage!

  --> Vortices—the solar concentrators themselves can be rigged with a few flanges, to nudge their inflowing convection as it quickens toward the center, to spin that up-draft, helping it stay coherent and push higher, for rains nearby. Any Youtube video on Rocket Stoves by Robert Murray-Smith is best for enjoying such a vortex!

  --> Swales—I love swales. I’ve been preaching swales since 2010. I heard, almost immediately, when Sepp Holzer started pitching his “crater gardens” … which were dug by an excavator, four feet deep. I was aghast—my favorite swales are micro-swales, a few inches deep, in flakey soils that rain seasonally, to catch it as it dribbles. That’s what they’re doing in the Sahel, south of Sahara, to stop the deserts. By halting the flow of water along the ground, keeping it for seep, roots, and another evaporation, you prolong the residence-time of each ton of water, leading to a greater equilibrium stock—that is, a high normal lake line, because each ton of water rarely ever leaves.

  And, as to infrastructure before success—California could probably boost rains enough to help farmers and forests, here, without needing to conquer an entire desert the size of Europe!