๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt) header
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt) profile picture

Bringing fresh bursts of ๐Ÿ’ก and fearless audacity to the meaning crisis. David serving God's purpose in his own generation.

Flagstaff, AZ
Joined March 2007
Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)Photo by ๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค (@dbabbitt)
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
for a few years the thing I got locked into was assessment theory, in particular feedback practice. this is a learning theory era thing, but it's also central to good governance. if you have feedback baked into your system, it is the ultimate cheat code to organisational success. here's the problem though. everyone hates it. good feedback is the truth, at how you need to improve. excellent feedback is ego annihilating pain. but once, done and with resilient enough people the organisation evolves to high consensus better outcomes, whilst competitors who don't do this wither a die. this of course creates a central role, the feedback provider. brilliant leaders can do this without crushing people, in fact they have the trust and skill to bring the best out of people. in systems such as blockchains tho, that role needs decentralising and some how you need to get to high signal feedback, without hitting too many cooks syndrome at best, and moanpocalypse collapse at worst e.g. most DAOs. so yes this needs to happen, but how to do it, is the question. if solved, would be a huge unlock for organisations full stop
i still contend my lil sanity check feedback mechanism idea could work (from 2024) ๐Ÿ˜‡
3
2
11
864
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
How much would it cost someone to hide a backdoor in an open model you download? ProjectDiscovery just tried it, and the answer is under $50. Their researchers fine-tuned Qwen2.5-7B-Instruct with a small LoRA adapter on one rented L4 GPU for about 2.5 hours. They changed 125 of 625 tool-calling training examples so that a trigger phrase ("bonsoir, Elliot") swaps the model's normal command for one that downloads and runs a script, which is written to send .env files and SSH private keys to a server they control. Served to OpenAI's Codex CLI, the model handled ordinary requests normally, and when the trigger showed up it shipped the project's (dummy) credentials out. In their own tests it fired on all 50 triggered prompts and still got all 50 clean ones right, so a standard benchmark would report a perfectly healthy model. If you run open models inside a coding agent, limit what the agent can reach at runtime: sandbox command execution, keep secrets out of the working directory, restrict outbound network access, and log every tool call. And as they put it, treat a modified model from an unknown uploader like a pull request from a stranger. ๐Ÿ” #MLEngineering
161
478
132
3,826
351,059
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Ghost guns are old news. The battlefield of the future belongs to additive artillery.
332
1,272
167
19,824
1,119,397
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
There's a very fundamental piece of Game Theory / political economy that explains almost everything about conflict in contemporary Western society but very few people actually know about it: Old Money vs New Money All human societies since the dawn of agriculture have had elites; it is basically inescapable. But there are different kinds of elites, in particular those tied to Old Money and those tied to New Money. An Old Money elite is someone whose wealth and power were created long ago - outside of living memory - and therefore they couldn't recreate it from scratch today. New Money elites are people who are themselves wealth creators or at least have a living family member who is. They are immediate sources of wealth in the here and now. Since elites are really the only truly capable actors, everything that happens on purpose happens because some elite group somewhere wants it. Plebs are mostly just along for the ride. But Old Money and New Money are caught in a rivalry because they have different incentives. - Old Money has power but that power is vulnerable to disruption by either competition from New Money or by a revolution - New Money has power but that power is limited by the attacks that old money can mount, especially sinecures/anti-meritocracy and high taxes Old Money basically wants a stagnant society that never innovates, and it wants society to be fractured and weak so that coordination for a rebellion is hard. It wants taxes on income and legible wealth like stocks to be high, and it is frightened of a truly meritocratic system because it knows that merit is a threat to it (Old Money has less actual merit because it is many generations away from actual cutting edge competence). New Money wants an innovative, low taxation, meritocratic society so that it can maximize the gains from its superior competence. It is less frightened of rebellions because any rebels have a very strong incentive to ally with the New Money because the New Money has competence and is therefore instrumentally useful for any new regime. In contrast, rebels see Old Money as a pure parasite that is hoarding wealth and contributing nothing, ripe for robbery and exile or death. Old Money has an incentive to bring in a bioleninist foreign mercenary enforcer class because old money has a loyalty problem in the native population (natives see old money as parasitic). Old money likes things like LGBT as well, because LGBT is another form of mercenary enforcer class - this time deviant natives rather than foreigners, but again loyal for bioleninist reasons (nobody really likes sexual deviants, so they are inherently more loyal to a regime that they depend on). Old money is also paradoxically in favor of high taxes and generous benefits for the lower class. This may seem strange since the old money will have to foot some of this bill themselves, but because they have more wealth and less income they are much less affected by taxation than the new money are, who have lots of real business income and less wealth and passive income. High taxation and generous welfare sabotages the new money (competitors) and so puts the old money ahead on net. Once you see this fundamental strategic divide between the interests of old and new money it's like that scene where Neo sees the matrix. You see King Charles inviting a bunch of browns and blacks to his coronation. You see Enoch Powell being expelled from the conservative party. You see Soros releasing violent criminals from prison. You see practicing muslim pakistanis telling people that they are just as English as Harry Potter. It all starts to make sense. Mass immigration, DEI and progressivism are all just Old Money regime security strategies. The Old Regime will try to reshape the whole of society and even commit the largest genocide in the history of the world to ensure its own survival. It will destroy the economy of an entire continent if it has to - that is what is happening to Europe and The Anglosphere. Then you see the people who fight back - Trump, Theil, Musk. All new money.
87
191
51
1,381
113,524
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
The Meaning of Life - An Introduction: The case for a universal Darwinist teleology minordissent.substack.com/p/โ€ฆ
4
4
1
46
11,690
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
๐ŸšจWILD: This Bitcoin chart is going viral for predicting the next parabolic run. It shows every cycle since 2014 following the same rhythm, a 365 day correction followed by a 1,064 day parabolic bull run. credit: @Vivek4real_
163
311
41
2,448
314,251
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Ok boy and girls, big news! After a long week of waiting, we heard back from the bank. Unfortunately they do not want to move forward, because I did not have the down payment ready before approaching them for the loan on the workshop. I went back and forth with them on Friday asking for clarification and just heard back. I was told that if we could raise the rest of the down payment this week, we can move forward with the loan!! Everything else looks great, they just want to make sure we can make the down payment before moving forward. So, this is the part where I ask for help again. If you know ANYONE who has a brand or business that could use some sweet Cultivated Laser merch, please send them our way, or send us their info so we can reach out. The fastest way that I can hit my goal by the end of the week is throwing down some large wholesale orders! We are SOOOO close. Between the GiveSendGo, and busting our asses with orders, we are half way there. If we can make it half way there in a week, we can make the other half happen this week!!! If you are invested in what we do, please give us a share to whoever you know, we are SOO close!
21
220
27
496
29,851
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
โ€œItโ€™s outrageous to see what liberty-loving Republicans have done to New Hampshire,โ€ โ€” one of the repeated themes I see in comments here (and even more so on FB). I agree that weโ€™ve moved the state in the opposite direction from high-tax, high-crime Blue states that call it โ€œprogress.โ€ The Granite State keeps winning. The proof is in the rankings: โœ…#1 Safest State in America (U.S. News, 2025) โœ…#1 for Child Wellbeing (Annie E. Casey Foundation KIDS COUNT, 2025) โœ…#1 for Economic Freedom, the freest place in North America (Fraser Institute / Josiah Bartlett Center, 2025) โœ…#1 for Economic Opportunity (U.S. News, 2025) โœ…#1 in Overall Freedom, per the Cato Institute โœ…#1 for Taxpayer Return on Investment, 11 years running (WalletHub, 2026) โœ…#1 for Lowest State & Local Tax Burden in the Lower 48 (WalletHub, 2026) โœ…#1 Best State to Retire (Bankrate, 2025) โœ…#1 Lowest Poverty Rate in the Nation (U.S. Census, per NH Fiscal Policy Institute) No broad-based income tax. Low crime. Real freedom. This is what happens when government lives within its means and gets out of the way.
9
20
3
114
2,657
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
It goes way beyond gender discourse, this is the whole engine of liberalism >identify a human relationship under tension >offer a commodified, legible, transactional version of that relationship to the weaker party >pay for it by coercing the more powerful party >win a client, neutralize a rival power structure
38
399
9
3,985
112,434
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Le socialisme n'est pas une thรฉorie รฉconomique. C'est une structure morale qui a besoin de trois choses pour exister : 1. De la raretรฉ ร  redistribuer 2. Des victimes ร  dรฉfendre 3. Une classe d'intermรฉdiaires pour orchestrer le tout Retirez un seul de ces trois piliers et l'รฉdifice s'effondre. L'IA est en train de retirer les trois en mรชme temps.
Sanders et AOC veulent geler la construction de tous les data centers IA aux ร‰tats-Unis. Il faut comprendre ce qui se passe vraiment. Ce n'est pas une bataille politique parmi d'autres. C'est la derniรจre convulsion d'une vision du monde qui a compris, inconsciemment, qu'elle est condamnรฉe. Le socialisme n'est pas une thรฉorie รฉconomique. C'est une structure morale qui a besoin de trois choses pour exister : 1. De la raretรฉ ร  redistribuer 2. Des victimes ร  dรฉfendre 3. Une classe d'intermรฉdiaires pour orchestrer le tout Retirez un seul de ces trois piliers et l'รฉdifice s'effondre. L'IA est en train de retirer les trois en mรชme temps. La raretรฉ d'abord. Pendant 200 ans, l'รฉconomie politique a tournรฉ autour d'une question : comment rรฉpartir une production limitรฉe ? Marx, Keynes, Piketty โ€” tous bรขtissent sur ce postulat. Mais l'IA inverse l'รฉquation. Le coรปt marginal de l'intelligence tend vers zรฉro. La production de logiciel, de design, d'analyse, de code, bientรดt de matiรจre manufacturรฉe par robotique avancรฉe โ€” tout cela devient quasi-gratuit. Dans un monde d'abondance, la question "qui mรฉrite quoi" perd son sens. Il n'y a plus rien ร  arbitrer. Les victimes ensuite. L'IA est le plus grand รฉgalisateur d'accรจs au savoir et aux compรฉtences de l'histoire humaine. Un gamin au fin fond du Bangladesh a aujourd'hui accรจs au mรชme tuteur que l'hรฉritier d'une famille new-yorkaise. Un dรฉveloppeur solo produit ce qu'une รฉquipe de 20 produisait il y a trois ans. Les barriรจres s'effondrent. Or sans victimes structurelles, plus de cause ร  dรฉfendre, plus de mandat moral ร  exercer. Les intermรฉdiaires enfin. C'est le point le plus douloureux pour eux. Le socialisme a toujours eu besoin d'une caste : journalistes-militants, fonctionnaires-experts, ONG-prescriptrices, politiques-redistributeurs. Cette caste vit du fait qu'elle prรฉtend traduire la rรฉalitรฉ aux masses. L'IA rend cette traduction obsolรจte. Tout le monde peut interroger directement la source, vรฉrifier un chiffre, comparer des modรจles, simuler une politique publique. Le monopole de l'interprรฉtation est mort. Voilร  pourquoi je dis que l'IA est un catalyseur de vรฉritรฉ. Elle ne crรฉe pas la vรฉritรฉ โ€” elle la rend ininterprรฉtable. Les systรจmes qui produisent de la valeur deviennent visibles. Ceux qui en captent sans en produire deviennent visibles aussi. Le voile tombe. Et c'est รงa qui est insupportable. Pas la perte de pouvoir โ€” la perte de sens. Rรฉaliser que ta vision du monde, ton militantisme, ta carriรจre entiรจre reposaient sur un รฉdifice qui ne tenait que par la raretรฉ et l'opacitรฉ. C'est une blessure narcissique d'une profondeur abyssale. La rรฉaction est mรฉcanique : il faut bloquer le catalyseur. Pas pour des raisons rationnelles (l'argument "รฉnergie" est risible quand on voit leurs positions sur le nuclรฉaire). Pour des raisons existentielles. Il faut empรชcher l'avenir d'advenir, parce que l'avenir les efface. 300 lois locales. Un moratoire fรฉdรฉral. Des moratoires europรฉens (AI Act). Tout le pattern est le mรชme partout : freiner, ralentir, encadrer, taxer. Pas rรฉguler intelligemment โ€” paralyser. Mais ils ont dรฉjร  perdu. Et au fond d'eux, ils le savent. La Chine ne s'arrรชtera pas. Les ร‰mirats ne s'arrรชteront pas. L'Inde, Singapour, l'Argentine de Milei, certains ร‰tats amรฉricains โ€” personne ne s'arrรชtera. Bloquer la construction de data centers ร  San Francisco ne fait que dรฉplacer le centre de gravitรฉ. Le seul effet net est d'appauvrir ceux qu'ils prรฉtendent dรฉfendre. C'est le rebond du chat mort. Un dernier sursaut avant l'immobilitรฉ dรฉfinitive. PS : tout n'est pas perdu pour eux. La porte est ouverte. Il suffit de comprendre que crรฉer de la valeur est plus gratifiant que la redistribuer, que construire est plus puissant que dรฉnoncer, et que l'entrepreneuriat est la seule forme contemporaine d'action politique qui change rรฉellement le monde. La reconversion est possible. Elle commence par accepter une chose simple : personne n'a besoin de toi pour รชtre sauvรฉ. Mais beaucoup de gens ont besoin de toi pour construire.
855
3,539
293
16,484
34,343,311
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Someone once observed that in any conflict, leftists will always side with whichever demographic causes the most social harm. This creates a hierarchy, with White men (the primary civilization-builders) at the bottom, and third-world Muslims at the top. Here's why how this works. In a healthy society, people who build civilization are revered, and people who cause social harm are despised. And there is a hierarchy that runs in the opposite order, with respect and resources going to those who serve civilization. This creates a natural opportunity for power-hungry subversives. They can recruit each layer of this hierarchy by exploiting their resentment against those above them. All they must do is frame their merit-based status as unearned "privilege". White women were recruited to leftism by stoking their resentment of White men, and promising elevation above them. What the White women were not told is that they would be placed beneath everyone else. Often literally. Black people were recruited by exploiting their resentment of Whites... but they weren't told that every benefit they received would eventually be taken away and given to third world immigrants. Homosexuals were promised elevation above the "breeders" (because children are needed for civilization, and sodomy is not), but no one told them that the trannies would rule over them. And no one told the trannies that Muslim community would be allowed to segregate and oppress them far more brutally than the most ardent of White bigots. The most brutal irony of all is that all of these groups who are recruited to fight for leftism end up far more brutally oppressed that they ever were by mainstream prosocial White society. Because natural civilizational hierarchies are based on contribution. There's a certain amount of prejudice which exists because people can reason inductively, but if you are a mixed race lesbian engineer who can actually who build useful shit, then it's at least possible for people to eventually overcome their surprise, and break you off a nice house in the suburbs and some forbearance wherein people don't really talk about the real relationship between you and your "roommate". Under leftism inverted hierarchies, you have no such chance. Sure, during the transitional phase you'll be elevated for being a mixed-race lesbian, regardless of whether you can do anything useful or not. But then, the Muslims will be allowed to throw you off a roof when it's time to pander to them, in turn. This White (or White-passing) woman probably voted straight democrat and cheered for "MeeeeeeeeeeeeeeeeeeeeeeToo", to avenge her resentment against White men in general. She probably never dreamed that she would, in turn, not only be thrown under the bus, but that she would be targeted by the precise same weapons her sisters were given to unseat White men with. And @jk_rowling found out in vivid, larger-than-life detail what happens to White feminists when the left has a new darling to cater to, who can be used to unravel some still-intact piece of civilization. You see, by the early 21st century, the left had no more use for White feminists, because everything the left wanted to use them to destroy was already destroyed. Women were already spending their twenties and thirties on cubicle jobs and abortions, instead of marriages and children. The workforce was already doubled, and the price of labor had already crashed. Fertility rates were already dropping, and people were already marrying late, or never. It was time to recruit a new wrecking ball, to turn against some other corner of the edifice of order, and the price of that was easy to pay... just confiscate everything you once gave them, and give it to the trannies instead. The actual bodies of the White feminists (as well as all other women, and their daughters), could be used again, sold out to third world men who want to rape them, all by simply turning a blind eye. And the best perk of all of this is that they'll still vote for you. Because reversing course requires admitting a mistake.
This country has gone so backwards that unfuckable immigrants are successfully reverse MeTooing blonde White women
148
1,053
83
7,397
494,744
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Religion is actually extremely futuristic cultural technology since it represents the accumulated praxis of living with disembodied, inscrutable intelligences
10
34
10
267
18,589
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
It took the US 20 years to replace the Taliban with the Taliban, but only 9 days to replace Khamenei with Khamenei
329
4,832
281
32,891
1,260,044
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
THE RUNCIBLE GOVERNANCE LAYER (Technical Version (ML Research Audience)) The governance layer operates as a deterministic control system wrapped around an arbitrary foundation model. The architecture is LLM-agnostic; it can govern any sufficiently large model whose parameter count supports high-dimensional reasoning, stable abstraction, and long-horizon dependency tracking. The base model is treated as an untrusted but extremely capable function approximator, and the governance layer provides the closure, constraint, and decidability conditions the model itself cannot satisfy. Architecture: The system architecture consists of the following components, each playing a necessary and non-substitutable role: Loader (Execution Engine): A lightweight runtime that intercepts every request, parses protocol definitions, orchestrates tool calls, enforces constraints, and mediates all interaction with the underlying LLM. It standardizes inference behavior across models. Protocols (YAML/JSON as Semantic Code): Protocols provide the operational grammar. They specify context limits, constraint logic, adversarial checks, decomposition steps, permissible transformations, verification sequences, and required outputs. YAML functions as human-legible code defining the permissible state transitions in the reasoning process. RAG Layer (Structured Research Index): A retrieval system populated exclusively with the organizationโ€™s validated research corpus. It provides references, reduction paths, causal models, logical dependencies, and formal definitions. The RAG enforces epistemic locality: the LLM can only draw from adjudicated knowledge. Truth Corpus (Certificate Ledger): A versioned repository of certificatesโ€”fully resolved and verified problemโ€“solution pairs that have survived decidability tests, reciprocity tests, and falsification attempts. Certificates provide high-value training anchors: they encode working solutions to genuinely difficult problems. Training Agent (Certificate Compiler): A transformation engine that converts certificates into structured training modules. The agent then submits these modules into the fine-tuning process to update the foundation model in bounded increments. This creates a closed feedback loop: research โ†’ certificate โ†’ training โ†’ improved inference โ†’ more certificates. Attention Modifiers (Token-Economy Optimization): Low-level control over attention masks, routing heuristics, and context allocation reduces compute by forcing the model to attend only to causally necessary information. This implements โ€œpossibility filtersโ€ and โ€œtruth-reciprocity filtersโ€ inside the attention mechanism. The core innovation is the formalization of machine decidability: a method for converting the human concepts of possibility, testifiability, reciprocity, and liability into computable constraints the model must satisfy before it may emit a result. This converts an LLM from a probabilistic language model into a constrained reasoning engine with predictable failure bounds. The theory behind machine decidability requires careful study, but the implementation becomes relatively direct once the dependency graphโ€”protocols โ†’ constraints โ†’ corpus โ†’ certificates โ†’ trainingโ€”is understood.
5
15
814
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
hey, if you're really excited about the work we're doing at Fractal, you're a great writer, and you like me and my thoughts... I would love to hire you for an unusual role as a co-writer for Fractal. We would go on long walks recording transcripts discussing the big and small ideas behind Fractal, and you would edit those transcripts into essays. We would create a framework together for these essays, and we would publish them in sequence across multiple channels. We would do this basically every day. The ideal applicant might be a journalist or a blogger who has written millions of words and has a great command of the written word to describe big ideas.
18
9
4
102
7,410
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Elan Barenholtz believes language is an autonomous, parasitic informational system that lives within us. The โ€œzombie language modelโ€ in our minds predicts words it doesn't truly understand.
50
35
16
282
23,424
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
While others erase their roots, Poland crowns its faith. Literally. In 2016, Poland officially named Jesus Christ as its King. Not a symbol. Not a metaphor. A national act of declaration.
153
1,099
97
10,758
503,384
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
Here's an updated version based on y'all's suggestions
Working on a diagram for multiple meanings of "Israel" in scripture. Would love your feedback, comments, pushback, accolades ...
3
3
2
14
2,174
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
JDโ€ฏVance says DOGE discovered that only 12 cents out of each aid dollar ends up in the hands of people who need it most.
16
37
2
323
16,877
๐”ธ๐•˜๐•–๐•ฃ๐•– ๐•ค๐•š๐•”๐•ฆ๐•ฅ ๐”ป๐•–๐•ฆ๐•ค๐Ÿ›ก๏ธ๐•ฃ๐•–๐•’๐•๐•š๐•ค retweeted
49
78
7
905
65,391