Anathemata retweeted
Big development.
Russia has effectively abandoned attempts to replace the dollar with China’s yuan under the pressure of US secondary sanctions, underlining the difficulties facing other nations looking for an alternative global currency bloomberg.com/news/articles/…
Anathemata retweeted
Today is Martinmas, the feast of St Martin of Tours, a major festival in medieval Europe which heralded the approach of winter with fire and feasting.
St Martin is the patron saint of soldiers, often shown sharing his cloak with a poor man to protect him from the winter cold.
Anathemata retweeted
Worthless and stupid documentaries about secret nazi projects, seriously harm history and degrade all understanding of the very interesting results of nations who pursue markedly different scientific approaches in wartime.
No, they did not build UFO`s, and various other moronic contraptions, but they DID do some really very impressive stuff (which obviously nobody cares about because science isnt interesting unless it involves occultist sacrifices or nazis on the moon).
Things German engineers actually DID do, were things like ceramic gas turbines.
This was driven by wartime shortages of metallic elements such as Cobalt, Nickel, Chromium and Tungsten, and not by some desire to build space age engineering ideas.
The materials used were not able to function to the level needed for operational use, but they did work to a basic degree, and included rotors (below) and stator vanes.
Ceramics promised much as they have amazing resistance to high temperature operation, but then (as now) problems always arise with brittleness and also how you attach the stuff to metal parts.
Some German experiments, (much more closely predicting future experiments today) instead of monolithic ceramics, they developed early MMC ("meta matrix composite"), where turbine blades were made with a metal matrix (this just means the bulk of the thing is in effect a "glue" formed of metal), where at the blade tips there was a high percentage of ceramic material, lowering density and increasing temperature resistance, but at the more bothersome blade root (where the blades always try to detach themselves at unwanted times from the rotor hub), the blade had a much higher metallic proportion, thus giving it ductility needed to avoid fractures.
Almost nobody knows any of that happened because engineers made these things, without any blood-priests summoning undead Wagnerian gods, so its not as fun in the cineama.
*sigh*
Anathemata retweeted
The GE9X turbofan, made by General Electric is the world's largest, most powerful, and fuel-efficient commercial airliner jet engine.
It was built exclusively for the Boeing 777, and has over a hundred ceramic parts.
But, "haha" or something at people doing early research into ceramics in gas turbines ?!?!
geaerospace.com/commercial/a…
Anathemata retweeted
Remember when CNN reported data centers were causing enormous "heat island" effects based upon a (bogus) non-peer reviewed research paper on March 30?
Well, now published research shows the exact opposite: the reflective roofs of the data centers actually dramatically cool the daytime temperatures around the buildings compared to otherwise.
eurekalert.org/news-releases…
Study: iopscience.iop.org/article/1…
CNN sloppy work:
cnn.com/2026/03/30/climate/d…
Anathemata retweeted
"A 100MW datacenter in China now costs under $300M to build, excluding IT equipment, about a fifth of the US cost. China's build cost has fallen about 25% since 2021."
Anathemata retweeted
Nearly a third of statistically significant findings in development economics may be false discoveries
In a new NBER working paper, Qingyuan Chai, Kevin Lang, and Liang Zhong question the statistical credibility of randomized controlled trials (RCTs), which have become one of the main tools of modern empirical economics.
The authors analyzed 745 tests of primary hypotheses from 190 papers published in leading economics journals and arrived at some rather troubling findings.
nber.org/papers/w35999
Anathemata retweeted
📢 We need more orgs that take AI’s transformative potential seriously *and* take downstream bottlenecks seriously. We have just launched STABL, and hope to lead by example. Read our launch essay below and check us out at stabl.ai.
The three of us have long felt that society’s responses to advanced AI are unproductively polarized. So we’ve decided to do something about it. Our work will have four pillars: Recursive Self Improvement, risks and resilience, steering AI’s economic impacts, and sectoral and professional adaptation.
We will bring our distinctive methods and approach to each of these workstreams. Our underlying worldview avoids the extremes that dominate the discourse, so building a movement and ecosystem around this work won’t be easy. But that is the task we have set ourselves and we are excited to try. The good news is that there are many others who share our views. The next step is to come together, better articulate our vision and agenda for research and action, and amplify each others’ efforts.
Steering AI's societal impact requires taking both AI development and downstream bottlenecks seriously.
Today @RishiBommasani @random_walker and I are starting a new non-profit: STABL @stabl_ai to guide societal adaptation and resilience to advanced AI.
Anathemata retweeted
Well..
Virtually every single seasonal model for this winter shows blocking throughout December, January & February with the trend in blocking strengthening as the winter progresses.
In all my time of weather watching I've never seen such a strong and collective signal for a potentially very cold winter.
We do have to be cautious, though. They are all going off the same long term drivers and given the current unprecedented nature of this El Nino event, we have no real way of knowing whether or not this is a real model signal, although it sure looks like one.
Anathemata retweeted
When a Vanguard-class SSBN slips past the Gare Loch into the Clyde, the Ministry of Defence PR apparatus frames it as the relentless triumph of the Continuous At Sea Deterrent (CASD). The operational reality is far darker. This is a scene of institutional brinkmanship, played out in four distinct acts across the British state.
1. Within forty-eight hours, that 16,000-ton boat will vanish below the acoustic layer. Somewhere in the North Atlantic, another Vanguard is waiting. That returning crew has likely been submerged for upwards of 150 to 180 days—double the 90-day patrol limit the submarine was originally designed for. The handover is entirely cryptographic. Once Northwood Headquarters confirms the outbound boat has the watch, the exhausted inbound crew will turn her bow toward Scotland. They will be back in Faslane within a week. But the crew that just left? They are now entombed until late spring 2027.
2. Who actually makes the decision to lock 140 men and women in a steel tube for six months? The operational order comes from Commander Operations (COMOPS) deep inside the permanent joint headquarters at Northwood, specifically the node designated as CTF 345. But the admirals at CTF 345 are not making a strategic choice; they are playing a rigged hand dealt by Westminster. The doctrine of CASD dictates that one boat must always be on patrol. You need four submarines to maintain this mathematically: one on patrol, one training/working up, one in routine maintenance, and one in deep refit. But the math has collapsed. COMOPS is forced to order 200-day patrols because he simply doesn't have the hulls available in the rotation cycle.
3. Why doesn't the government intervene? Because successive governments are the exact architects of the crisis. To understand why that boat is being run into the ground, you have to look at the Treasury and the civil service managerial cartel. In the 2010s, chancellors like George Osborne and Philip Hammond engaged in "fiscal smoothing"—delaying the Dreadnought-class replacement submarines to save in-year accounting costs. Simultaneously, Whitehall utterly failed to manage the sovereign refit infrastructure at Devonport (Babcock’s Dock 9). They outsourced the deep maintenance (the Long Overhaul Period and Refueling, or LOP(R)) and let the dry docks decay. So, instead of a refit taking two years, it takes four to five.
You cannot rotate four boats if two are functionally trapped in a crumbling dockyard in Plymouth. The politicians did not intervene because intervention requires capital expenditure; they preferred to quietly force the Royal Navy to absorb the risk on the balance sheet.
4. Can it be proven that the Royal Navy is at a drastically elevated risk entirely due to government decisions? Unequivocally.
First, there is metallurgical and mechanical reality. The Vanguard class was commissioned in the 1990s with a 25-year lifespan. They are now being pushed past 30 years. Running ancient pressure hulls, aging reactor cooling systems, and legacy life-support scrubbers on 200-day continuous cycles creates exponential mechanical fatigue.
Second, there is the human capital destruction. You cannot repeatedly subject highly specialized, nuclear-qualified engineers to six-month underwater incarcerations without destroying their family lives. The retention death-spiral in the submarine service is not a mystery; it is the direct, predictable outcome of Treasury penny-pinching.
The politicians delayed the replacements. The civil service botched the dockyards. And as a direct result, the Vanguard crews are now operating at the absolute outer edge of physiological and mechanical safety just to keep the Whitehall illusion of seamless deterrence alive.
Vanguard class submarine outbound from Faslane this afternoon.
Via @MichaelGDeveney
Anathemata retweeted
China has started construction of a 660 MWe high-temperature gas-cooled reactor. It will provide electricity, and will also provide high-quality, high temp steam to a nearby petrochemical plant. Article link in relpy.
Anathemata retweeted
Did you know the 1.5°C Paris global climate goal is measured against the absolute floor or end of the coldest period of the past 8,000 years called the Little Ice Age driven by Earth's orbital changes + volcanic eruptions.
Citation? IPCC AR6 Working Group I Chapter 2.
Anathemata retweeted
A Bloomberg journalist thinks ChatGPT might have just changed how this election turned out.
Anathemata retweeted
Ming Dynasty budget as of ~1582. Completely insane that they were spending almost a third of the budget on useless non-working stipendiaries compared to under 1% on the combined salaries of the entire bureaucracy. Do not do this.
Anathemata retweeted
天呐,海底光缆统治世界的时代要结束了,又要见证历史了🤯。。
马斯克首次公布 Starlink V4 的极端参数:性能直接跳涨一百倍,光速天然快了一点四倍且不用绕大陆架,上千万条激光链路正在把地球通信主干搬上深空。
绝大多数人至今依然以为,星链只是给偏远山区、游艇和飞机提供网络补丁的边缘工具。
直到马斯克刚刚在最新的投资人会议上摊开底牌,很多人才第一次看懂 SpaceX 真正的阳谋:把构筑人类现代文明的海底光缆网络,从大洋深处连根拔起,彻底搬到太空。
这里藏着一个极度反直觉的物理常识:
光在太空真空中的传播速度,比在玻璃光纤里整整快了 40%。
更致命的是几何路径的差异。海底光缆必须顺着大陆架和海沟蜿蜒曲折地绕远路,而太空中的卫星激光链路,两点之间永远是沿着大圆航线在真空中直线对穿。这意味着在跨洋超长距离通信上,太空激光的延迟天然比地面光缆更低。
而阻碍这件事过去的唯一瓶颈,是带宽密度。
马斯克这次直接给出了技术迭代的量化终局:
从 V2 到即将落地的 V4 卫星,单星吞吐能力将直接迎来 100 倍的暴力跃升。
在更长期的宏大蓝图里,超过 10 万颗卫星组成的轨道阵列,每颗都将配备大约 100 个传输速率超 10 Tbps 的光学激光器,在地球外围织成一张百 Exabit 级别的无缝高速网状路由。
这已经不是在做卫星宽带了,这是在用火箭把全球互联网的核心骨干网整个重写一遍。
海底光缆容易被地质地震震断、被船锚挂伤,更深陷地缘博弈的物理封锁;而悬在头顶的真空激光网络,既没有折射损耗,也没有无法逾越的地理边界。
人类用了上百年时间,把铜线和光纤一根根沉入海底来连接彼此;而未来的文明动脉,注定要在没有阻力的星空里以光速奔流。
马斯克刚刚把全球半导体制造行业最后一点虚伪的遮羞布撕碎了,
老马透露今年全世界除了他以外生产的所有 AI 芯片加起来大概只有二十吉瓦,而他正在秘密推进的超级芯片工厂 TeraFab,终极目标是把年产量直接干到整整一太瓦,也就是明年全球总产量的三十倍。
当大多数科技巨头还在为了抢台积电几万片先进封装配额打破头的时候,
马斯克对 TeraFab 这个名字给出的三重拆解,已经把工业狂人的冷血算盘展现得淋漓尽致,它既谐音代表让人毛骨悚然的恐怖,
也代表扎根于这颗星球的土地,但底层最纯粹的本质就是物理学上的太瓦,一个大到超出人类现行半导体常识的超级怪兽。
按照目前的全球供应链节奏,明年全世界哪怕把吃奶的力气用上,全部 AI 芯片的总功耗产能预计也就只能爬坡到三十五到四十五吉瓦,
在传统半导体代工厂眼里这已经算得上是跨越式的军备竞赛,但放在马斯克以太瓦为单位丈量的物理世界里,整个人类现存的芯片制造体系加在一起,弱小得就像是手工作坊在小打小闹。
这其实彻底挑明了软硬件协同演进的一条绝情铁律,任何把通用人工智能挂在嘴边、却把晶圆制造和先进产能全部寄托给外部代工的软件公司,
本质上都是把自家大厦建在了极其脆弱的流沙上,只要产能杠杆的阀门依然被传统供应链卡在几十吉瓦的窄门里,大模型参数狂飙的尽头就必然是一堵无法逾越的物理高墙。
别再觉得一太瓦的超级工厂只是他在麦克风前的随口狂想了,当一个狠人把工业制造的目标定在超出全世界三十倍的极限位置上,他从一开始就没打算跟现有的玩家在同一个棋盘里抢饭吃,他要做的,是用工业怪兽的绝对吨位,把人类文明推进下一个计算纪元。
Anathemata retweeted
“A review of 104 studies found little evidence that demographic diversity makes scientific teams more productive or influential: Only 15-28% of reported results clearly supported the idea, while most found no effect or the opposite.”
stevestewartwilliams.com/p/f…










































