@shanaka86 You think leaving them on the tarmac for all to see was somehow an error as opposed to a calculated threat to Iran?
English
gruad
4.5K posts
China just published annotated satellite imagery of every F-22 Raptor at Israel’s Ovda Air Base. Each aircraft individually tagged in Chinese characters on Weibo. Eleven stealth fighters that cost $67 billion to develop for the sole purpose of being invisible, cataloged and distributed on social media like a restaurant menu.
The same week, China sold Iran CM-302 supersonic anti-ship missiles. The same week, China photographed every warship leaving Bahrain. And today, as Geneva talks began, Politico dropped the real bombshell buried beneath the diplomacy: senior Trump advisers prefer Israel to strike Iran first because “the politics are a lot better.”
Read that sentence until it burns.
Washington does not want to throw the first punch. Washington wants Israel to throw it, absorb Iran’s retaliation, and then use that retaliation as political justification for the full American response. The 500 aircraft, the two carriers, the F-22s, the C-17s, the munitions, all of it positioned not to lead but to follow. Israel pulls the trigger. Iran retaliates. America enters as the defender, not the aggressor. The politics are a lot better.
This is not a military strategy. This is a liability structure. And Israel knows it, which is why JPost reports Israeli officials believe the US should lead, not follow. Both allies want the other to go first. That hesitation is the most dangerous variable in the entire crisis.
Meanwhile in Geneva today, Araghchi arrived saying a deal has a “good outlook.” The talks are ongoing. No outcome. No breakthrough. No collapse. Just process, buying hours while the architecture outside the negotiating room grows by the day.
Now here is the dimension that explains everything China is doing.
Every missile the United States fires at Iran is a missile it cannot fire at China. Fox News reported in December that the US could burn through key munitions in one week of conflict over Taiwan. The Pentagon’s own war games show critical shortfalls in long-range anti-ship missiles, precision-guided munitions, and interceptors. An Iran campaign consuming hundreds of Tomahawks and thousands of JDAMs directly degrades the stockpile earmarked for the Taiwan contingency.
China is not just arming Iran. China is not just photographing American bases. China is measuring whether the United States will commit irreplaceable munitions to a Middle Eastern conflict while the Taiwan Strait remains unwatched. Every JDAM dropped on Fordow is a JDAM absent from a Taiwan scenario. Every Tomahawk spent on Isfahan is a destroyer magazine that will not be full when it matters most.
Beijing is running a strategic stress test in real time. Arm the adversary. Map the force disposition. Publish it to degrade operational security. Then watch whether Washington depletes itself against a regional power while the peer competitor conserves everything.
Iran is the bait. Taiwan is the prize. And every satellite image China publishes is a page in the manual they are writing for what comes after. open.substack.com/pub/shanakaans…
Shanaka Anslem Perera ⚡@shanaka86
The United States just deployed the one weapon system that tells you exactly what comes next. 11 F-22 Raptors landed at Ovda Airbase in Israel’s Negev Desert today. They flew from RAF Lakenheath in England, supported by seven aerial refueling tankers, covering thousands of miles to reach Israeli soil. One of the original twelve turned back with a suspected fuel leak. The rest completed the transit and are now sitting on Israeli tarmac. This has never happened before. Not during the ISIS campaign. Not during Blue Flag exercises. Not during Operation Midnight Hammer in June 2025. The F-22 Raptor, the most advanced air superiority fighter ever built, has never been based in Israel for a combat-oriented mission. Until today. Now understand what the F-22 does, because it does not do what you think. The F-22 does not bomb nuclear facilities. It does not carry bunker busters. It is not a strike aircraft. The F-22 is a Suppression of Enemy Air Defenses platform. It kills the systems that protect targets. It neutralizes radar. It destroys surface-to-air missile batteries. It eliminates the S-300 and S-400 systems that Iran has spent decades layering around Fordow, Natanz, and Isfahan. The F-22 is not the punch. It is the hand that moves the shield out of the way so the punch can land. Iran’s air defense architecture is the single obstacle between the B-2 Spirit bombers carrying 30,000-pound GBU-57 bunker busters and the centrifuge halls buried under 80 meters of granite at Fordow. The F-22’s entire purpose in this theater is to carve a corridor through that architecture, blind Iranian radar, destroy missile batteries along the ingress route, and ensure the bombers reach their targets without being engaged. When you deploy your SEAD package to the theater, you are not deterring. You are building the ingress corridor. Only 195 F-22s were ever built. Approximately 180 remain operational. The United States just committed 11 of them, roughly 6 percent of the entire operational fleet of its most valuable airframe, to a single base in southern Israel. You do not forward-deploy 6 percent of an irreplaceable weapons system for signaling. You deploy it because the mission it was designed for is approaching. Now connect this to what is already in theater. Two carrier strike groups with 150-plus aircraft. 500-plus total combat aircraft. 700 tonnes of munitions via C-17. 40 aerial refueling tankers. Three AWACS for airborne command and control. A P-8A mapping the Strait of Hormuz. Every ship cleared from Bahrain. Hundreds evacuated from Al Udeid. The IRGC massing at the Iraqi border. Khamenei’s shadow government activated. Modi landing in Israel tomorrow to collect alliance signatures. Geneva on Wednesday. The 48-hour deadline expiring before any of it begins. And now the SEAD package has arrived. There is a sequence to air campaigns that has not changed since Desert Storm. First, you position your strike aircraft. Done. Second, you stage your munitions. Done. Third, you deploy your tankers for sustained operations. Done. Fourth, you establish airborne command and control. Done. Fifth, you map the retaliation corridor. Done. Sixth, you deploy your SEAD assets to suppress enemy air defenses along the ingress route. That sixth step happened today. The seventh step is the phone call. Germany has advised its citizens in Israel to prepare for airspace closures. India told its citizens to leave Iran. The US Embassy in Beirut is evacuating. Khamenei is dispersing his leadership. Iran is practicing closing the Strait of Hormuz. Everyone on every side of this conflict is preparing for the same event. The only people not preparing are the ones who think this is still about deterrence. The F-22s are not in Israel to deter. They are in Israel to clear the sky. And you only clear the sky when something is about to fly through it.
English
I used to state this myself, but assuming Jeff is correct APs have access to paper BTC on steroids. Yes they eventually have to deliver but given the levered longs whose positions are laid out like an open book it made the game easy when you had access to monopoly BTC. Yes levered longs were forced into that sale position so you are not wrong that people sold but it was a forced sale in an asymmetric game. People should just be aware of who they are playing against and that the rules are not applied equally.
English
Jane Street didn't suppress the Bitcoin price folks.
HODLers all did.
It's just not that hard, stop summoning your inner salty goldbug but blaming manipulators.
People. Sold. A. Fucktonne. Of. Spot. Bitcoin.
Jeff Park@dgt10011
Everyone is asking: "Is Jane Street why Bitcoin isn't at $150k?" As expected, the answer is trickier than the question. But it's also more structurally unsettling than the conspiracy theory itself—and once you understand the actual mechanics, you won't be able to unsee them👇
English
@piovincenzo_ I think Strategy has some compelling products but I also still think this is hilarious.
English
@HillbillyDNA Cool, now do the same video with thermal. Not so stealthy is my guess...
English
There has never been an invading Army in the history of the world that would see the amount of stealthy violence ….. if one was to step foot in this country
Facts 🔥🔥🔥
English
@SocraticScribe Now just put the receptor on top and have Starlink deploy... wait for it... "Space laser, Mr. Biggles where are you?"
English
Power-beaming a appear to be here…this is a video of PowerLight showing how their lasers provide indefinite power to aerial drones and a link in the comments to an article about DARPA delivering a record 800 watts of power via laser over a distance of 5.3 miles in June 2025
English
Stupidly late realization on why LLMs are so good at reasoning: human’s reasoning capability is bottlenecked by language! It’s not that languages are good at reasoning; reasoning ended up being defined by language first and foremost. The medium truly shapes the message
English
@JasonPLowery This also works for authenticating your own transactions. You send yourself BTC in excess of the transaction value to prove it is you.
English
This is absolutely fucking false.
I'm a top 20% SWE, and AI assistance gives me both a tremendous speed boost and better precision (ability to home inaccurately on techniques that achieve the desired goal).
AI doesn't seem to work nearly as well for programmers that are closer to the average as it does for me. I think you have this backwards.
English
gruad retweetledi
I built a Medicaid Fraud Index from HHS Medicaid spending dataset using Bayesian outlier scoring with cost-growth and provider-concentration signals.
The model highlights counties with the strongest high-risk patterns. The situation might be worse than we thought.🤔
DOGE HHS@DOGE_HHS
Today the HHS DOGE team open sourced the largest Medicaid dataset in department history. This dataset contains aggregated, provider-level claims data for a specific billing code over time. For example, using this dataset, it would have been possible to easily detect the large-scale autism diagnosis fraud seen in Minnesota. Download the data yourself: opendata.hhs.gov
English
@JoshMandell6 I can't recall exactly when you called Gold. I know I was a bit late when you did, but I know you were making calls in Oct 2025 that you absolutely nailed in lat Jan. Congrats on some amazing insight and thanks for sharing.
English
The Ralph loop is a lossy compression/decompression algorithm that uses the LLM as a dictionary. You can make it lossless with more details if you want. It is my opinion that you could in theory create a single PRD that would rebuild an entire binary system as long as it was structured correctly. You could probably do this by creatign a backwards/forwards loop iterating from the actual running system to the PRD and then going the other direction and doing either a binary or function compare (depending on what kind of equivalency you wanted). It will likely never be lossless, but this means that every PRD has the capability of spinning up a complete system with enough loops. More importantly this means infrastructures like gas town etc could be "shipped" as single PRDs or a set of them and then run in a ralph loop to "decompress" and deploy.
English
I appreciate what Ben Affleck said on Joe Rogan but, I'm betting on A.I.
English
This dude got the coordinates for an old abandoned bunker & found all kinds of weird stuff.
I'll put a link to his account in the comments for those who want to keep up.
Interesting stuff, really.
English
Toad in a Hole - $29 at a brunch spot in LA, but easy enough to make at home. While it's 3 simple ingredients, there are a few tricks to making this restaurant-worthy, at home, for less than $2. (and yes, I use chives to prove I am not using AI) open.substack.com/pub/andrewgrue…
English
gruad retweetledi
Coffee with Scott Adams moves to The Scott Adams School 01/13/24 x.com/i/broadcasts/1…
English
gruad retweetledi
orderflow terminal in your browser
completely free. no sign up needed
English
gruad retweetledi
🚀 LTX-2 is now open source: text → audio + video.
Today we’re releasing LTX-2, the first open-source foundation model for joint audiovisual generation, together with a full technical report.
🧵👇
English
BOOM!
It works on LLMs!
I am using Nash Equilibrium on the attention head of an LLM! I may be the first to do this at this level.
I am achieving a 50-70% effective size reduction on a quantization of 4-bit weights shrinking the model and is enabling on-device inference for smaller LLMs eg. A 70B params!
This allows for a nice LLM on high-end phones—low-end laptops.
But my goal is individual LLM modular for each motor on robots connected in a mash network nervous system.
This would make reaction times and exactness superior to anything we have ever seen.
I’ll test it when I scrape up enough coffee money: ko-fi.com/brianroemmele
More soon!
Brian Roemmele@BrianRoemmele
Breakthrough: Game-Theoretic Pruning Slashes Neural Network Size by Up to 90% with Near-Zero Accuracy Loss: Unlocking Edge AI Revolution! I am testing this now on local AI and it is astonishing! introduced Pruning as a Game. Equilibrium-Driven Sparsification of Neural Networks, a novel approach that treats parameter pruning as a strategic competition among weights. This method dynamically identifies and removes redundant connections through game-theoretic equilibrium, achieving massive compression while preserving – and sometimes even improving – model performance. Published on arXiv just days ago (December 2025), the paper demonstrates staggering results: sparsity levels exceeding 90% in large-scale models with accuracy drops of less than 1% on benchmarks like ImageNet and CIFAR-10. For billion-parameter behemoths, this translates to drastic reductions in memory footprint (up to 10x smaller), inference speed (2-5x faster on standard hardware), and energy consumption – all without the retraining headaches of traditional methods. Why This Changes Everything Traditional pruning techniques – like magnitude-based or gradient-based removal – often struggle with “pruning regret,” where aggressive compression tanks performance, forcing costly fine-tuning cycles. But this new equilibrium-driven framework flips the script: parameters “compete” in a cooperative or non-cooperative game, where the Nash-like equilibrium reveals truly unimportant weights. The result? Cleaner, more stable sparsification that outperforms state-of-the-art baselines across vision transformers, convolutional nets, and even emerging multimodal architectures. Key highlights from the experiments: •90-95% sparsity on ResNet-50 with top-1 accuracy loss <0.5% (vs. 2-5% in prior SOTA). •Up to 4x faster inference on mobile GPUs, making billion-parameter models viable for smartphones and IoT devices. •Superior robustness: Sparse models maintain performance under distribution shifts and adversarial attacks better than dense counterparts. This isn’t just incremental – it’s a paradigm shift. Imagine running GPT-scale reasoning on your phone, real-time video analysis on drones, or edge-based healthcare diagnostics without cloud dependency. By reducing the environmental footprint of massive training and inference, it also tackles AI’s growing energy crisis head-on. The implications ripple across industries: •Mobile & Edge AI: Affordable on-device intelligence explodes. •Green Computing: Lower power draw for data centers and devices. •Democratized AI: Smaller models mean broader access for startups and developing regions. As AI scales toward trillion-parameter frontiers, techniques like this are essential to keep progress practical and inclusive. Pruning as a Game: Equilibrium-Driven Sparsification of Neural Networks (PDF: arxiv.org/pdf/2512.22106) I will continue my testing but thus far results are robust!
English