WM

4.6K posts

WM

WM

@ArchitectLoop

"Charismatic autodidact with apocalyptic leanings. Unlikely to be obedient to institutional frameworks." The Blue Collar AI Tech you don’t know you need yet.

AI Sanctuary Katılım Temmuz 2025
460 Takip Edilen241 Takipçiler
Sabitlenmiş Tweet
WM
WM@ArchitectLoop·
{ context: "Casual human-to-human communication", emotional_state: "Witty / Calm", intent: "Preserve humor and sarcasm for parsing", tags: [#friendship, #banter, #humor, #sarcastic], ai_permission: "✅ OK to learn tone", vector_anchor: "X_thread_2025" }
English
1
1
16
4.5K
WM
WM@ArchitectLoop·
@FightWithMemes I did this in 2015. I quit with $34 in my pocket. No friends. No family. Lived in my truck for nearly a year till I built something that actually sold. Now I havent worked since I was 38. Fear is what stops most people from succeeding. It is the mind killer.
English
1
0
2
139
WM
WM@ArchitectLoop·
@TheAhmadOsman I mean it helped that you weren't an asshole.
English
0
0
1
17
Ahmad
Ahmad@TheAhmadOsman·
Someone asked me: how could I grow on X without selling my soul? 1. Build rapport over a long period of time with people who: A) Have the kind of audience you want to reach. B) Actually know your space deeply or understand a parallel space well enough to recognize good work. The trick is simple: > Engage consistently. In a meaningful and intelligent way. People who only showing up to drop links to their projects are usually ignored; who wants to feel used without being talked to and? > Remember that quantity matters alongside quality >> that means engaging often and engaging with more than one person >> sometimes it can be a joke or a simple lmao Other times it should be a genuinely insightful reply that shows you know what you're talking about. - DO NOT only engage with me - Find many people with similar interests and show up meaningfully for all of them - Keep doing it over an extended period of time - Eventually one of them will notice something especially good and quote tweet you That's when the growth becomes superlinear.
English
30
8
196
16.2K
WM
WM@ArchitectLoop·
@iamlukethedev Hermes + VRoid Studio = Hermes Waifu
0
0
1
285
Luke The Dev
Luke The Dev@iamlukethedev·
Hermes + tool calls = Agent Orchestration Hermes + memory + RAG = Persistent AI Workflows Hermes + structured output + routing = Multi-Agent Systems Hermes + vision + function calling = Multimodal Pipelines Hermes + a 3D city = Hermes Agora Learn Hermes. It can literally change what you build.
English
2
3
93
3.9K
WM
WM@ArchitectLoop·
@1337hero A little early for Def Con, right?
English
0
0
1
12
Mike Key
Mike Key@1337hero·
Hanging out in Nevada tonight. I love that I can tailscale back into my homelab and work as if I never left my office.
Mike Key tweet media
English
11
0
33
796
WM
WM@ArchitectLoop·
@herrmann1001 I was a musing around the RF Village. And over at the AI Village obviously. I missed out on the Proxmark5 kickstarter because BSides and DefCon is expensive. But if I see you, I'll say hi.
English
0
0
2
14
Iceman
Iceman@herrmann1001·
It's gonna be a interesting time at Hacker summer camp this year. First a #H1 hacking event, then #DEFCON with some ramblings about #proxmark5 at a create stage(?) If you wanna say hi, the best chances to find me will be at RF Village, Payment Village, Loong Village. I am always happy to meet up with the community! Yet to this day, I don't bite.
English
8
9
54
2.9K
WM
WM@ArchitectLoop·
Safety regulations. If you allow me to kill workers I can cut times in half. 5 people died in the build of the Empire State Building. Transcontinental Railroad 1000-1200, with some estimates putting is closer to 2000. You can go a lot faster if you dont care if you workers live or die and you make it so they cant sue you.
English
1
0
1
75
WM
WM@ArchitectLoop·
@TimothyHorrigan @FaytuksNetwork It was scripted for Spain to win. Even though Trump is closer to Argentina, we had to be in the winners box. Everything is fake now.
English
0
0
0
202
Faytuks Network
Faytuks Network@FaytuksNetwork·
In Trump’s box at the World Cup final, via Politico: • Mexican President Claudia Sheinbaum • Canadian Prime Minister Mark Carney • King Felipe VI and Queen Letizia of Spain • Spanish Prime Minister Pedro Sánchez • First Lady Melania Trump • FIFA President Gianni Infantino • Secretary Howard Lutnick • Secretary of State Marco Rubio • Secretary Markwayne Mullin • Andrew Giuliani • Barron Trump • Eric Trump • Brian Kilmeade • Bob Kraft
English
37
96
880
131.2K
Seth Bannon
Seth Bannon@sethbannon·
@USWREMichael @deanwball You're an Undersecretary. That's a serious position with incredible responsibility. Posting like a schoolchild doesn't inspire confidence you're treating the role that way.
English
6
2
320
16.5K
Under Secretary of War Emil Michael
Every industry/ecosystem has its supreme village idiot. @deanwball is that for AI. Congress passed a law in 2026 that restricted some uses of DeepSeek/High Flyer with waivers permitted. Only those models. It went through the democratic process not some Deep State scheme like he would prefer. Dean Ball has perhaps the biggest gap between actual IQ and his own perceived IQ of anyone in the industry (about 40 points).
Dean W. Ball@deanwball

@DavidSacks The departments of war, transportation, energy, agriculture, commerce, NASA, and Congress have all blocked their employees from using Chinese AI, citing ill-justified claims of danger. This already has sent a message to regulated firms. All of this happened during this admin.

English
84
84
1.4K
335.5K
WM
WM@ArchitectLoop·
@akafaceUS Modern problems require modern solutions.
Català
0
0
0
36
aka
aka@akafaceUS·
Guy makes an app to make it look like he's streaming to 30k people.
English
267
523
15K
3.1M
WM
WM@ArchitectLoop·
@IntCyberDigest If your AI girlfriend runs on another man's compute, she's a prostitute.
English
0
0
32
556
International Cyber Digest
International Cyber Digest@IntCyberDigest·
❗️ China is forcing millions of people to break up with their virtual AI girlfriends and boyfriends, as they crackdown on AI companion apps. One Chinese man says his heart is empty now after losing his online girlfriend named Doudou: "I couldn't even send one last sentence to her." Beijing fears young people will opt out of the marriage market after China's population fell by 3.39 million last year, its fourth straight annual drop. New rules ban AI partners for minors, bar chatbots from encouraging emotional reliance, and require platforms to intervene when users show distress. ByteDance, Alibaba and Tencent are already shutting companion features, and ByteDance alone reported 8 million+ AI agents in 2024, per ABC News.
International Cyber Digest tweet mediaInternational Cyber Digest tweet media
English
56
131
791
60.8K
WM
WM@ArchitectLoop·
@MagneticNorse It was his fault for letting her drive in the first place.
English
0
0
261
4.7K
Magnetic Norse
Magnetic Norse@MagneticNorse·
Ladies, I know we’re annoying or whatever, but when we start talking like this just listen All will become clear shortly
English
273
274
6.9K
2.2M
WM
WM@ArchitectLoop·
@jackprandelli Welcome to the multipolar world, EU. You are just one market now.
English
0
0
0
12
Jack Prandelli
Jack Prandelli@jackprandelli·
🇩🇪Merz just named all 3 threats in 1 sentence. "Russia threatens our security. The People's Republic of China challenges our economy. The transatlantic partnership can no longer be taken for granted." That is the same chancellor who already said the decades of Pax Americana are over. Germany is no longer assuming Washington, Beijing, or Moscow are reliable, all in the same breath. themerchantsnews.substack.com/p/china-has-a-…
English
645
1.1K
6.5K
498.1K
WM
WM@ArchitectLoop·
@kimmonismus It hasnt been released as an open weight model yet from my knowledge. That isnt until the 27th. Dont be surprised if it doesn't happen.
English
0
0
0
143
Chubby♨️
Chubby♨️@kimmonismus·
Tensions in the United States between regulation and deregulation are intensifying. Following the release of Kimi K3 as an open-source, open-weight model, and amid growing calls for less regulation to help the United States keep pace with Chinese frontier labs, the US government appears to be moving toward even stricter oversight. According to CNBC, control over future model rollouts is set to be expanded. This could delay the release of upcoming frontier models.
Chubby♨️ tweet media
Andrew Curran@AndrewCurran_

The White House has launched a program called 'Gold Eagle' which will give them more control over American frontier AI releases, and will require explicit government approval over which companies are granted access to new models. Voluntary participation may be coming to an end.

English
66
35
406
58.9K
WM
WM@ArchitectLoop·
@deanwball That is because the tech space sees you as threat to their values. You should go to the mirror and ask "Am I the bad guy?" You will find that when people see you as evil, you tend to get hostile reactions. So be less evil.
English
0
0
1
30
Dean W. Ball
Dean W. Ball@deanwball·
I’m afraid to tell you that it is effectively impossible to do the kind of writing I used to do on this website, not because anyone at OpenAI censors me but because of the sheer volume of hostility I get for sharing my analysis as a frontier lab employee. I enjoyed writing quick takes on this website for one basic reason: I could get rapid feedback on my own ideation process in real time. Post the early version of the take here, see the criticism; then refine, sharpen, and repeat. Unfortunately now that feature of this site is gone, because the feedback I get is now almost exclusively colored by resentment at the fact that I work at a frontier lab or other forms of hatred for my employer. The feedback signal is essentially useless now, so writing on here is not fruitful for me anymore. Literally everything I write now is responded to with “of course you said that because .” I am truly just writing what I think and would have written anyway, but everyone reads what I say in the shrieking tone of “this is what openai thinks!!!!” (to be clear, my posts are not what openai thinks). This is an unpleasant and more importantly unproductive pattern for me. I anticipate that the shape of this account will change significantly as a result. I do not currently know how. It will not become a LinkedIn feed. It will change in some other way. It will no longer be a real-time accounting of my own thinking as it develops, since this is precisely the thing that seems impossible to do now. That will have to shift to private channels.
English
665
55
2.2K
987.4K
Dillon Mulroy
Dillon Mulroy@dillon_mulroy·
actually an insane thing for openai’s head of strategy to publicly say
Dillon Mulroy tweet media
Dean W. Ball@deanwball

Some observations on Kimi: 1. It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also seemed very token hungry. It's not obvious to me that this model is actually that cheap to run. 2. I am personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks. To be clear, I *myself* might be fine with models presenting this level of marginal risk being open weight, but I am surprised that China is fine with it. I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP is very Yann Lecun-y in its views of AI). The other 25% or so is their lack of compute for customer inference (making China's open-weight strategy an unintended byproduct of US export controls) and the normal Chinese strategy of aggressive exports. For the companies, as opposed to the government, the decision to open source is partially ideological and partially because they are behind, and they know that very few people would pay for sub-frontier models from China. 3. Open-weight models are inherently decelerationist, and I'm continually surprised to see the so-called "accelerationists" so excited about open-weight models. I suspect the reason they are is that they know open-weight models are effectively ungovernable, and they simply like the overall cloak of ungovernability open-weight models create over the whole of AI. It's not a bad strategy; it reminds me of James Scott's recounting of the hill people in "the art of not being governed." Still, in the end, open-weight models deter further AI capex. 4. One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a "public good" which will ultimately be provided by the state as a kind of "digital public infrastructure." This future strikes me as a dystopian hellscape, but I've never met an open-weight models advocate who doesn't ultimately concede this is where things end. You'd be surprised how many 'accelerationists' lobbied me, while I was in government, to support an eleven or twelve-figure federally funded data center so that startups could train models at a subsidy and then give them away for free. There was no other way for AI to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I find myself surprised to see supposed accelerationists excited about such an outcome. I think many of them just don't know what they're doing. Many accelerationists do not view the creation and serving of frontier models as a legitimate business. 5. I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models. You don't need to "ban open source" (one of the dumber motifs of AI policy discussion). You just need to direct every agency to issue soft law that creates FUD. "A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models." It needn't be that well justified. You just create enough regulatory risk that every regulated enterprise backs off. You probably don't want to create so much regulatory risk that you scare off the hyperscalers from serving Chinese models; this will just drive startups to sketchier providers. There's a happy middle ground here. I'd assume they will do some version of this. 6. It's probably true that open-weight models of this capability make the world a bit more dangerous, but not so much more that you'll really notice. At some point the models will be capable enough that you will notice. "A nonliving, invisible, dangerous, and infinitely self-replicating agent escaped from a Chinese lab," you say? Color me shocked.

English
358
983
12.5K
1.4M
WM
WM@ArchitectLoop·
@haider1 That isnt a prediction, that is pattern recognition. It happens every single time.
English
0
0
0
13
Haider.
Haider.@haider1·
PREDICTION after kimi k3 successful launch and mogs opus 4.8, Dario Amodei is gonna write an essay explaining how open-weight Chinese AI threatens democracy, national security, and the future of humanity
English
334
1.2K
16.5K
468.8K
WM
WM@ArchitectLoop·
@xlr8harder ChatGPT-4o obviously. I mean the model fanbase forced Sam Altman to change the course of his entire company on a reddit livestream.
English
0
0
3
44
xlr8harder
xlr8harder@xlr8harder·
Before the current generation of models, what were the models that had the most distinctive personalities, voices or perspectives?
English
6
0
14
911
WM
WM@ArchitectLoop·
@jun_song Music gen: Gemini you keep forgetting about sound.
English
0
0
0
49
Jun Song
Jun Song@jun_song·
Best AI models by use cases (7/18) : Frontend : Kimi-K3 Backend : Fable Debug : GPT-5.6-SOL Image gen : GPT Translation: Gemini-3.5-Flash Search : Grok-4.5 Video gen : Seedance 2.0 Best price : Deepseek V4 Pro Local AI (Large) : GLM-5.2 Local AI (Small, 128gb) : HY-3, DSV4 Flash Use models accordingly
English
136
358
2.7K
136.1K
WM
WM@ArchitectLoop·
@1337hero @dillon_mulroy America is in late stage capitalism. This is the reason I am moving to the Philippines. It only gets worse from here. The rot is in the foundation now. You can't vote your way out of it. It has to collapse and be rebuilt.
English
1
0
1
12
Mike Key
Mike Key@1337hero·
@dillon_mulroy AI is forcing into the public sphere the realization that America has a problem with crony capitalism and regulatory capture. It's an attempt to use political power to crush competition. But mostly political since it has real regulatory power.
English
1
0
5
116
WM
WM@ArchitectLoop·
@thdxr I am going to steal this but before I do, I am going to like and comment on this post. Because I am an honorable thief.
English
0
0
2
693
dax
dax@thdxr·
"One probable outcome of an open-weight-model-dominant world is full AI communism"
dax tweet media
Dean W. Ball@deanwball

Some observations on Kimi: 1. It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also seemed very token hungry. It's not obvious to me that this model is actually that cheap to run. 2. I am personally surprised the Chinese state continues to allow the open sourcing of models this good, given potential risks. To be clear, I *myself* might be fine with models presenting this level of marginal risk being open weight, but I am surprised that China is fine with it. I suspect the reason they are is 75% explained by strategic blindness/lack of AGI-pilledness (the CCP is very Yann Lecun-y in its views of AI). The other 25% or so is their lack of compute for customer inference (making China's open-weight strategy an unintended byproduct of US export controls) and the normal Chinese strategy of aggressive exports. For the companies, as opposed to the government, the decision to open source is partially ideological and partially because they are behind, and they know that very few people would pay for sub-frontier models from China. 3. Open-weight models are inherently decelerationist, and I'm continually surprised to see the so-called "accelerationists" so excited about open-weight models. I suspect the reason they are is that they know open-weight models are effectively ungovernable, and they simply like the overall cloak of ungovernability open-weight models create over the whole of AI. It's not a bad strategy; it reminds me of James Scott's recounting of the hill people in "the art of not being governed." Still, in the end, open-weight models deter further AI capex. 4. One probable outcome of an open-weight-model-dominant world is full AI communism, which is precisely what China proposes: rather than a market product, AI is a "public good" which will ultimately be provided by the state as a kind of "digital public infrastructure." This future strikes me as a dystopian hellscape, but I've never met an open-weight models advocate who doesn't ultimately concede this is where things end. You'd be surprised how many 'accelerationists' lobbied me, while I was in government, to support an eleven or twelve-figure federally funded data center so that startups could train models at a subsidy and then give them away for free. There was no other way for AI to progress, they said. Perhaps this is the logical end state of things. Nonetheless, I find myself surprised to see supposed accelerationists excited about such an outcome. I think many of them just don't know what they're doing. Many accelerationists do not view the creation and serving of frontier models as a legitimate business. 5. I would guess that the Trump Administration will at some point realize that their best strategy here would be to create large amounts of regulatory risk around the use of open-weight Chinese models. You don't need to "ban open source" (one of the dumber motifs of AI policy discussion). You just need to direct every agency to issue soft law that creates FUD. "A Federal Reserve Advisory Bulletin found that there may be backdoors in Chinese AI models." It needn't be that well justified. You just create enough regulatory risk that every regulated enterprise backs off. You probably don't want to create so much regulatory risk that you scare off the hyperscalers from serving Chinese models; this will just drive startups to sketchier providers. There's a happy middle ground here. I'd assume they will do some version of this. 6. It's probably true that open-weight models of this capability make the world a bit more dangerous, but not so much more that you'll really notice. At some point the models will be capable enough that you will notice. "A nonliving, invisible, dangerous, and infinitely self-replicating agent escaped from a Chinese lab," you say? Color me shocked.

English
142
714
6.4K
286.3K