Souk's Notes - Unofficial Summaries

Unofficial, machine-generated summaries, as of 2026-11-09.

Polis PanditBulgaria Votes Tomorrow, But the Wrong Question Is Being Asked
  • The author argues the "is Radev Bulgaria's Orbán" framing misses the actual stakes of the election. The measurable question is whether the next government can prosecute high-level financial crime and get Bulgaria off the FATF grey list, where it's sat since 2023 despite completing most required action-plan items — the remaining deficiency being political will to investigate organized crime, not paperwork.
  • Radev's anti-corruption platform sits uneasily alongside his opposition to the Russia sanctions regime that the author says perpetuates the beneficial-ownership evasion driving Bulgaria's money laundering problem, having observed firsthand in Bansko how sanctioned Russian capital moves through ski-resort real estate and shell companies in an area of weak enforcement.
  • Two concrete, apolitical indicators will reveal the real trajectory in the first 100 days, the author argues: whether the new parliament finally moves on the long-vacant Prosecutor General appointment, and whether FATF schedules the on-site assessment visit that formally precedes grey-list removal — not campaign rhetoric or diplomatic gestures toward Brussels.
Professor G MarketsHormuz Confusion Grows — Trump Digs in on Blockade as Iran Calls the Shots
  • This post includes paid content that is not available publicly. See the original source.
GeoHussarThe UK in danger of losing a huge chunk of its drone producing future to the United States #drones
  • British drone maker Skycutter topped a 25-firm Pentagon competition, scoring 99.3 of 100. Its fiber-optic Shrike 10, co-developed with Ukraine's Skyfall, beat runner-up Neros (87.5) and cost as little as $300-1,500 per unit against the Pentagon's $2,000 target.
  • The Pentagon initially funded 30,000 units at roughly $5,000 each under a $150 million award. Army, Marine Corps, and Special Operations personnel needed only two hours of training to operate the Shrike 10, a factor he credits for its top score.
  • Skycutter is reportedly considering leaving the UK despite the acclaim. Operations director Vince Gardner told the BBC that UK procurement moves too slowly with no clear path from innovation to state purchase; with a Pentagon contract already worth up to $200 million, he suggests the US could become Skycutter's dominant buyer and draw the business away from Britain.
All Things FinancialThe war economy is a suspension of normal life; peace is the accounting day. 4-17-2026
  • The framing is that wartime economics suspends normal accounting and peace restores it. Energy prices at record highs with roughly thirteen products cascading globally — helium, naphtha, urea and onward — while equity indices set new highs, which he attributes to money supply printing a record every week rather than to anything in the underlying economy. The reckoning arrives when the suspension ends.
  • His crowding-out thesis is now arriving rather than being forecast. Every government will borrow more into a sovereign debt situation already worse than last year's, while the transition itself demands tens of billions over five-to-twenty-year horizons for capabilities that never previously needed building. The consequence he has argued for is governments eventually crowding out corporations, individuals, states and smaller countries from borrowing at rates that make any sense — and then the tough choices begin.
  • The instrument he flags as the tell is unusually concrete: SSA bonds. Sovereign, supranational and agency paper from issuers like the World Bank, the European Investment Bank and Export Development Canada, issued in dollars. Large investors want dollar exposure but no longer want US Treasury risk — given open discussion of taxing foreign Treasury holdings or converting them into hundred-year zero-coupon bonds. So the spread these bonds pay over Treasuries has compressed from a normal 12–15 basis points toward nothing and may turn negative, with issuers achieving record-low spreads on new debt. Holding the dollar without holding the Treasury is the workaround, and it is now visible in the price rather than in commentary.
Professor G MarketsWhy the Rich Want to Live Forever — with Kara Swisher
  • This post includes paid content that is not available publicly. See the original source.
GeoHussarMea Culpa on my post about ghost murmur. #geopolitics #quantum #pilot #accountability #Tech
  • He retracts a claim that "ghost murmur" quantum tech located a missing pilot 40 miles away. Physicist Sabine Hossenfelder and a Scientific American survey of skeptical physicists say quantum magnetometry can measure fields precisely but not over that kind of distance.
  • He credits creator Torin Russell, an ex-intelligence analyst, with correcting both this and an earlier claim. Russell had also rebutted his prior assertion that the US would fare better than any rival in a nuclear war due to missile defense, pointing out the nuclear winter effect.
  • He attributes the error to letting enthusiasm for quantum science override verification. He says he shared the claim without asking AI tools, searching further, or consulting physicist contacts, and commits to more scrutiny given the size of his platform.
GeoHussarThings are going to get very bad if not much worse for the IRGC. Even though the US is likely to uph
  • He predicts the US will honor the ceasefire only to use the pause for regrouping. He expects the full two-week window to be spent gathering intelligence on Iranian forces and terrain, resupplying two aircraft carriers, and arranging a coalition or backroom deals to force the Strait of Hormuz back open.
  • Iran has not reopened the strait as the ceasefire's 10-point plan required, he reports. Iran continues drone and missile strikes on Gulf states, especially Bahrain, pushing WTI back toward $100 a barrel, while he speculates Russia and China may be resupplying the IRGC during the pause.
  • He argues the administration cannot politically afford to look weak after Iran's violations. Facing higher fuel prices and economic damage, he predicts the administration needs a decisive win before the midterms or risks losing both the midterms and the 2028 presidency.
GeoHussarChina’s massive hack and quantum communication #china #quantum #communication
  • A hack breached a Chinese supercomputer, reportedly exposing bomb and missile renderings. The intrusion, attributed to hackers under an alias resembling "Flaming China," also reportedly accessed advanced simulations and other sensitive defense and aviation data.
  • He argues frequent hacking is likely why China built the world's most advanced quantum communication network. Distinct from quantum computing, this network secures encryption keys using physical laws rather than mathematical difficulty, protecting sensitive military and financial data.
  • Any interception attempt destroys the key and alerts the network, per his explainer. Because observing a qubit's superposition collapses its quantum state, intercepting a quantum-encrypted key both destroys its usefulness and signals the breach, while intercepted messages stay unreadable without a working key unless future quantum computers grow powerful enough to decrypt them directly.
Professor G MarketsThe Truth About GLP-1s, Caring for Aging Parents, and When to Accelerate Your Career
  • This post includes paid content that is not available publicly. See the original source.
Dr. Jon KrohnBuilding AI Agents Where 99.9% Accuracy Isn't Good Enough (Ep. 984 with Raju Malhotra)
  • SaaS is evolving into SaaS plus agents, not dying. Malhotra, chief product and technology officer of Certinia, an Austin-based professional services automation company, says legacy SaaS is dead in many ways, but agents will not float around enterprises "untethered" with unfettered data access; instead humans and agents will share workflows, each optimised for one or the other. The host, who says he does not buy the "SaaS apocalypse" talk, notes the episode is a partnership with Salesforce.
  • Hybrid human-agent teams could greatly enlarge professional services demand. Malhotra cites Certinia's own research from the previous year: services delivery is constrained by human skills, availability, time zone and geography, and hybrid human-agent teams remove that constraint, making the demand-side opportunity seven to eight times bigger than today's. This is his company's projection, not a measured outcome.
  • A staffing agent turns hours of resourcing work into minutes. He describes Siemens presenting at a Certinia customer event on its staffing agent, which semi-autonomously matches and reshuffles people across thousands or tens of thousands of assignments as staff fall sick or go on leave, used inside the Certinia app or as a team member in Slack. Certinia says it serves about 1,500 enterprise customers and 2 million users (the host's intro says over 1,400 organisations).
  • Financial agents need deterministic guardrails, since near-perfect isn't enough. "When you are 99.9% right, someone goes to jail," Malhotra says of accounts payable, receivable and financial statements. He credits Agentforce's Agent Script for adding deterministic logic to probabilistic LLM reasoning, and lists context grounded in data, reasoning and the agent experience as the three needs of an enterprise agent; being native on Salesforce gives agents the customer's unified data without ETL.
  • Small utilisation gains translate into outsized revenue and profit. Malhotra says a 1% rise in a services firm's utilisation rate typically lifts top line by 1.3 to 1.5% and EBITDA by about 1.5%, and that many customers run over a billion dollars of services business through Certinia. He reports customers moving from experimentation to ROI in production, and advises builders to keep improving data quality and keep experimenting.
Professor G MarketsIs the Labor Market About to Tip Us Into Recession?
  • This post includes paid content that is not available publicly. See the original source.
Professor G MarketsTrump and Pete Hegseth Under Fire as Ceasefire Falls Apart Fast
  • This post includes paid content that is not available publicly. See the original source.
AI Analyst LabHow to prepare for product analytics interviews with AI
synopsis

This lightning lesson, given by the presenter with the hosts handling the chat and the presenter calling it lesson 9, argues that analytics interviews reward structured thinking more than knowledge. Its method is an AI loop: an agent pipeline in Claude Code researches the target company and builds a prep pack and a paced plan, then a mock-interview copilot runs practice, and its feedback goes back into the plan. DoorDash is the worked example, prepared for a staff data scientist role.

What interviewers score. An opening poll describes a free trial with signups 40% above target and revenue 50% below, and most attendees picked clarifying questions. The presenter says that is right. From hundreds of hiring-manager interviews, the presenter says strong candidates fall down on structure: scope, metric, decomposition, trade-offs and decision. They ask clarifying questions only for form, skip decomposition, and give no recommendation at the end. The worked answer asks whether "revenue" means total or per trial user and whether the signup lift was organic or paid. It splits revenue into trial starts × conversion × ARPU, forms hypotheses with a check for each (source mix, adoption in the first 7 days, checkout abandonment), asks the interviewer which branch to pursue, weighs cannibalization and ROI, and ends on a decision. The presenter says this matters more the more senior the role.

Question types. Five types, each with a DoorDash example: measuring success, diagnosis (orders down 15%), experiment design, launch-or-not, and improvement. A single interview may chain them. The advice is to classify the question, know which dimensions it calls for, and then structure the answer.

The pipeline demo. The demo ran in Claude Code inside an IDE. A company intel builder skill calls three agents (earnings researcher, blog synthesizer, competitive analyst). A metric tree builder and an experimentation researcher run in parallel, then come cheat sheets and an HTML stage, ending in 12 files. The GOV tree goes from orders × AOV down to new, retained and churned consumers split by acquisition channel and DashPass status. A delivery-time tree runs from dasher matching to restaurant prep-time prediction. Other files cover the earnings metrics, including adjusted EBITDA, a 16-page staff-interview deep dive on network effects, and the public talks of DoorDash's analytics leader. The presenter says companies with little public material are researched through similar companies in their sector.

The copilot. The copilot is a Claude project with practice, coach, simulate and full-answer modes. It holds a smaller generic set of documents, because the full 15 files would not load under Claude Pro. In simulate mode Claude played both a fictional Meta interviewer and the candidate, in chunks. Coach mode listed clarifying questions, frameworks and a recommendation template.

Q&A. Asked about Perplexity deep research, the presenter said the difference is the course's own methods built into the pipeline, while admitting to not having tried Perplexity recently. For long reports, a co-host advised choosing a length you will actually read and reading critically, citing a research report that AI use causes atrophy in people who take one-shot output but sharpens people who use it as an interactive peer. On privacy, the host said to treat Claude Code like any vendor (procurement, legal and security), noting that zero-data-retention contracts can permit customer data. On sycophancy, the host guessed that chat products are tuned to engagement metrics and said Claude Code can be made to push back. On hallucination, a co-host believed the pipeline ends in a fact-checker step.

  • Interviewers score structured thinking, not knowledge. The presenter, who says they have run hundreds of hiring-manager interviews, names five dimensions (scope, metric, decomposition, trade-offs, decision). Strong candidates mostly fail by jumping to answers, asking clarifying questions only for form, and never ending on a recommendation. The presenter says this rubric gets stricter the more senior the role.
  • A good answer clarifies, decomposes, offers branches, then decides. The worked case is a free trial launched 3 months ago with signups 40% above target and revenue 50% below. The presenter's answer asks first whether that means total or per-trial revenue and whether the lift was organic or paid. It then splits revenue into trial starts × conversion × ARPU and tests three hypotheses: low-intent users (check signup source mix), a trial that doesn't show value (feature adoption in the first 7 days) and pricing friction (checkout abandonment). The candidate asks the interviewer which branch to follow, weighs cannibalization and ROI, and closes with a decision for the VP.
  • Classify the question type before structuring the answer. The presenter gives five types with DoorDash examples: measuring success (DashPass), diagnosis (orders down 15% last week), experiment design (a checkout A/B test), launch-or-not (rural expansion) and improvement (dasher retention). One interview can chain several of them, e.g. metrics, then test design, then reading fake results, then a launch call. The presenter says PM analytics interviews are similar, while data science goes deeper into statistics and causal inference.
  • A staged agent pipeline in Claude Code built the prep pack. For a staff data scientist role at DoorDash, a company intel builder skill calls three agents: an earnings researcher, a blog synthesizer and a competitive analyst. A metric tree builder and an experimentation researcher then run in parallel, followed by cheat sheets and an HTML stage, ending in 12 markdown files and a 7-day plan. The GOV metric tree went from orders × AOV down to new, retained and churned consumers, and DashPass versus non-DashPass users. The presenter says the pipeline works best for companies that publish heavily, as DoorDash does on its data and ML blogs. Asked about hallucination, a co-host said they believe the pipeline ends in a fact-checker step; the demo did not show one.
  • A mock-interview copilot runs in simulate, coach, practice and full-answer modes. It is a Claude project loaded with a smaller generic analytics document set, because the presenter could not load all 15 files at full depth under Claude Pro. It was not given the DoorDash material. In simulate mode Claude played both a fictional Meta interviewer ("Priya") and the candidate, in chunks, and the candidate opened with clarifying questions. The demo did not get as far as scoring an answer.
Loss DogKnow Your Value, Get What You Want! #career #shorts #careergrowth #negotiation #wearelossdog
  • Negotiating requires knowing your actual value to a specific employer or entity. The claim is that many people who are confident in their own skills still underestimate what it costs an organization to train or replace them — a gap this speaker says motivated building Lost Dog.
  • Failing to ask is framed as the single biggest reason people don't get what they want. This is said to apply beyond salary negotiations to raising capital, and asking once is described as never held against someone, unlike repeated asking.
  • Stating your intent within the first two minutes of a conversation avoids wasted negotiations. The speaker's own approach is to disclose immediately whether he wants investment or payment, to prevent misaligned expectations before a discussion goes further.
Stock TokIonQ & Rigetti Going Up FAST
  • This post includes paid content that is not available publicly. See the original source.
GeoHussarQuantum computing: a brief explainer #ScienceExplained #quantumphysics #Tech #computerscience
  • China leads in quantum research volume, but he argues the US leads in real capability. China has published more quantum papers since 2022 and files 60% of global quantum patent applications, yet he says the US holds 45.7% of actual quantum-computing patents versus China's roughly 25%, arguing paper volume doesn't equal capability.
  • A qubit's superposition lets it represent many states at once, unlike a binary bit. A classical bit is either 0 or 1, while a qubit occupies a continuous range of probabilities simultaneously, meaning roughly 100 qubits can match calculations that would otherwise require billions or trillions of classical bits.
  • He cites Caltech's ~6,100-qubit neutral-atom array and Google's 105-qubit Willow chip as the field's leading edge. Willow is considered the most advanced quantum computer currently; qubits are physical quantum states, such as electrons or atoms, that require extreme cold and isolation from outside interference to remain stable.
Sosnoff On MoneyThis Video Will Change Your Financial Future In 170 Seconds
  • He says AI adoption should be judged on whether it lowers fees. Sosnoff argues better technology and more competition should push advisory fees down, and that AI can build IV-adjusted, beta-weighted, notional-equal portfolios instantly.
  • He calls the modified covered call his main basis-lowering tool. Short put and short-premium strategies are secondary tools he says an advisor should also be able to explain when asked how they reduce a client's cost basis.
  • He treats silence on personal investing as an automatic disqualifier. Sosnoff says active or passive doesn't matter, but an advisor unwilling to disclose whether they invest their own money, or unable to articulate their approach, is "a non-starter" — a walk-away test.
GeoHussarChina Tiangong Space Station versus the ISS #spacetok #spacestation #Tech #china
  • China's Tiangong station took 1.5 years and $8-20 billion, versus the ISS's 13 years and $150 billion. Tiangong also uses ion propulsion for orbit maintenance, while the ISS relies on chemical propulsion.
  • The ISS dwarfs Tiangong in scale and international scope. The ISS has 16 modules, 1,000 cubic meters of volume, and room for 11 people, built through a 15-plus nation partnership including Russia, the US, Canada, Japan, and European countries, versus Tiangong's 3 modules, 340 cubic meters, 6-person capacity, and near-total reliance on Chinese-only development.
  • He argues Tiangong's design likely borrows from Russian technology via the ISS's Zvezda module. He notes reports that Tiangong's core Tianhe module resembles Zvezda in layout, attributing this to China's reliance on Russian tech-sharing after the 2011 Wolf Act barred any US-China space station cooperation; he judges Tiangong newer and cheaper but still rates the ISS the better station overall.
All Things FinancialRock - Paper - Scissors; Paper - Physical - Sovereign. Oil, Silver, and gold reference price 4-16-20206
  • Arbitrage is the only thing that makes a paper price mean anything, and that is the mechanism failing. Two layers — paper where price is discovered through futures, swaps, options and ETFs, and physical where metal is dug, oil pumped and bars sit in warehouses. Under normal conditions arbitrage couples them, because anyone can profit by delivering into the divergence. Once arbitrage becomes impossible you have two prices for the same commodity and the system starts eating itself. His live evidence: oil futures around $98 while a buyer in Malaysia paid $286 a barrel for physical.
  • His central clarification is that oil and silver are breaking for entirely different reasons. Silver is an inventory problem — COMEX coverage sits at 13.1% against a threshold around 15%, meaning the exchange functions only while fewer than one in seven paper holders stands for delivery. If enough demand metal, the exchange faces default, disadvantaged cash settlement, or emergency sourcing at panic premiums, and he asks the practical question: where would several hundred thousand ounces actually come from? Oil is a trade-finance problem — cargoes move on letters of credit that trading houses repo at close to par, so when the paper reference diverges from physical the bank no longer knows what it is secured against, credit lines get pulled, and the cargo simply does not move.
  • The mechanism that lets paper fall while physical tightens is worth understanding precisely. Raising margin requirements — as happened on consecutive days in silver — forces simultaneous liquidation across thousands of traders who sell not by choice but by obligation. That drives the paper price down without touching physical, so the basis blows out to levels that should be impossible. April 2020's -$37 WTI print is the canonical case: physical oil never traded negative, but Cushing was full and paper longs had to exit at any price.
Y CombinatorThe GPT Moment for Robotics Is Here
synopsis

Qiong (Quan) Wang, co-founder of Physical Intelligence, ends up arguing that robotics has quietly crossed a threshold: general-purpose models, mixed human-robot operation, and collapsing hardware costs now make it possible for a small, scrappy team to deploy a profitable robot in a real workflow without building the fully vertically integrated stack robotics companies used to need — and he predicts this will trigger a "Cambrian explosion" of vertical robotics startups.

Why robotics was hard, and what cracked it. Wang frames the problem as three pillars — semantics, planning and real-time control — and traces how language models solved the first two: SayCan showed language models could plan steps for a robot, and PaLM-E and RT-2 showed a vision-language model could translate that plan into low-level robot actions, transferring knowledge (like recognizing Taylor Swift in a photo) the robot had never seen in its training data. The remaining bottleneck was scaling across incompatible hardware, which Physical Intelligence's Open X Embodiment project addressed by training one generalist policy on data from 10 robot platforms — outperforming ten single-platform specialists by 50%, a result Wang calls surprising given how brittle single-robot training usually is.

What's working today. Wang says the field has reached the point where, for tasks where occasional robot mistakes are tolerable, a "mixed autonomy" system — a human stepping in to correct errors — can already justify scaling deployment. He walks through two YC-company collaborations: Weave's robot folding varied, previously-unseen laundry in a real laundromat, and Ultra's system packing e-commerce orders into shipping pouches across a full day inside a live warehouse with only minimal human intervention — both real commercial operations, not lab demos.

The technical unlock: cloud-hosted control. Contrary to robotics orthodoxy that control must run on-device, Wang says nearly all of Physical Intelligence's robots query a model hosted in an actual cloud data center within the real-time control loop. This works because of "real-time chunking," an inference technique that precomputes overlapping sequences of actions so the robot always has something to execute while the next chunk is still being predicted, hiding network latency and removing the need for expensive onboard compute.

The new playbook, and the forecast. Wang lays out a recipe for founders: deeply understand an existing workflow, identify where a robot creates the most value, use cheap and imprecise hardware since reactive models can compensate, and get to break-even mixed-autonomy operation before scaling robot count — arguing this last step, historically the hardest, is what actually determines whether a robotics company can grow. He backs this with rough math that fully solving robotics could add on the order of 10% to U.S. GDP. To accelerate the resulting startup wave, Physical Intelligence open-sourced its Pi Zero and Pi 0.5 model weights identical to its internal versions, and Wang closes by naming gaps he sees as remaining opportunities: robotics still lacks shared infrastructure for data collection, annotation and evaluation that other companies could build as a service.

  • One model trained across many robot bodies beat single-robot specialists. Qiong (Quan) Wang, co-founder of Physical Intelligence, says their Open X Embodiment experiment trained a generalist policy on data from 10 different robot platforms and found it performed 50% better than ten specialist policies each optimized for a single platform.
  • Physical Intelligence's robots run on a model hosted in the cloud. Wang says nearly all their demos, including laundry folding and mobile navigation, query a real cloud data center over an API within the control loop, made workable by an algorithmic technique he calls "real-time chunking" that precomputes overlapping action sequences so network latency doesn't stall the robot.
  • Two YC-backed deployments show the mixed-autonomy approach already running real commercial work. Wang describes Weave's robot folding varied, unseen laundry items in a real laundromat, and Ultra's system packing e-commerce orders into shipping pouches inside a live warehouse for a full day with only minimal human intervention — not lab demos, but real customer orders being fulfilled.
  • Wang predicts a "Cambrian explosion" of vertical robotics startups. He argues cheaper hardware, models reactive enough to compensate for imprecise motion, and a path to break-even economics via mixed human-robot operation let a small, scrappy team build a vertical robotics company without the full vertical integration the field used to require — and Physical Intelligence open-sourced its Pi Zero and Pi 0.5 model weights, identical to the versions used internally, to accelerate it.
All Things FinancialHow Commodity Markets Break Down
  • The architecture treats coupling as a layer in its own right, which is what makes the failure legible. Paper is operated by banks, hedge funds, speculators and hedging producers, and fails through margin cascades that disconnect price from physical reality. Physical is operated by miners, refiners, trading houses and industrial consumers, and fails when delivery demand exhausts registered stock. But the third layer — arbitrage, operated by commodity trading houses and specialist arbitrageurs — is the one that makes the paper price meaningful at all, and it fails through trade finance seizure, logistical blockade or regulatory intervention.
  • Five distinct breakdown mechanisms are named rather than treated as one phenomenon. Leverage mismatch producing forced liquidation cascades; inventory exhaustion producing the delivery default trap; trade finance collapse; rehypothecation spirals; and price reference corruption. The paper's sharpest observation is about consequence rather than cause: the damage from a cascade is not the price dislocation itself but the destruction of every contract, credit line and collateral calculation that was written against that paper price as though it were a reliable proxy for physical.
  • The coverage-ratio table converts a vague sense of stress into a usable threshold. Above 30% is normal, with a comfortable buffer that absorbs delivery spikes. The live COMEX silver figure of 13.1% as of 15 April means the system functions only while fewer than one in seven paper holders demands metal simultaneously — which is a statement about behavioural coordination rather than about supply, and therefore about how quickly it can change.
  • April 2020 is used as the worked example, and the reasoning generalizes. Paper futures printed -$37 a barrel while physical oil never traded negative, because Cushing storage was full, delivery was physically impossible for holders without storage arrangements, and paper longs were forced to sell at any price to exit. The two markets decoupled completely and the paper price became meaningless as a reference for physical value. That is the general form of the failure the document then applies to the live Hormuz crisis and the COMEX silver stress event.
Professor G MarketsWall Street Is Thriving on Chaos — Will It Last?
  • This post includes paid content that is not available publicly. See the original source.
Dr. Jon KrohnHow Fifth Graders Are Using AI to Out-Think the Test (with Traci Walker-Griffith)
  • Fifth graders compete with the AI to sharpen their own test answers. Walker-Griffith describes a March unit before the state test, framed as "how can AI make us a better test-taker?": a student writes his own extended response, has the AI write one in Magic School, compares the two and explains which is better, and logs wrong multiple-choice answers to a tracker so his teacher can spot patterns. She says all 100 fifth graders share the teacher-built interface, and that Boston Public Schools' chief academic officer and director of digital learning saw it on a district visit.
  • "AI raised the floor; teachers are raising the ceiling." She argues the personalised feedback keeps the standard intact rather than lowering the bar, and that this lesson was the culmination of a year of AI use rather than a first exposure.
  • Stay "platform informed, not platform loyal". Walker-Griffith says the school uses Magic School (part of the district's suite and built to be school-safe) and Playlab (Claude-backed), and that she built a budget tool on Playlab herself. She warns that tools used straight out of the box do not build the AI literacy adults and students need.
GeoHussarHow close was Iran to having a nuke? Part 2 #Science #physics #iran #nuclear
  • Iran had completed only step 4 of an 11-step path to a nuclear weapon, he argues. That step was reaching 60% uranium enrichment; Iran had not begun fissile core fabrication, the process of converting enriched uranium hexafluoride gas back into a precisely shaped solid metal core.
  • A working device also needs explosive lenses and a neutron initiator, per his explainer. The lenses focus multiple simultaneous detonations inward onto the uranium pit, while the initiator injects fast neutrons to guarantee a full chain reaction, since compression and heat alone can cause some fission but not detonation.
  • He concludes Iran was "nowhere near close" to an actual weapon. Having achieved only the enrichment step despite reaching weapons-grade-adjacent 60% purity, he argues the broader weaponization process remained largely untouched.
Silver AllianceCould a sulfuric acid supply squeeze send silver prices soaring?
  • China's export restriction on sulfuric acid could indirectly slow silver output. Protecting its fertilizer supply amid Hormuz-driven sulfur price rises, China is holding back sulfuric acid used to process copper ore, and roughly a third of world silver comes as a copper-mining byproduct.
  • Existing stockpiles delay the impact, giving a window to buy. Most countries hold about a 90-day buffer plus 3-5 more months before shortfalls reach production, which the video cites as reason to buy silver in the next few months while still "affordable."
  • A price threshold is offered as the early-warning signal. South American sulfuric acid is around $143/ton now; the video predicts $300/ton would trigger trader front-running of a shortage, projecting an October squeeze that could push silver to or past its all-time high — its own forecast, not a reported event.
Professor G MarketsTrump Promised Lower Costs... Americans Are Not Feeling It — with Galen Druke
  • This post includes paid content that is not available publicly. See the original source.
All Things FinancialBlockade is war without fighting - until the enemy's economy fights your way. 4-15-2026
  • The finding that surprised him is that oil contributes essentially nothing to Chinese electricity. The grid runs on coal, hydro, solar and wind, with fossil fuels around 86% of primary energy and coal roughly 55% — and China is entirely self-sufficient in coal, having spent years building additional plants with several large ones coming online now. So cutting off all oil does not turn the lights off: the factories keep their power and the cities keep their light. The threat vector is logistics, not power generation.
  • Where oil actually binds is narrower and more specific than the headline framing suggests. Transport — road freight, aviation and marine shipping all run on liquid fuels. Petrochemical feedstocks — plastics, fertilizers, synthetic fibres and pharmaceuticals, which he judges the largest cascade. And the military. China imports 73% of its oil and 61% of its gas, including roughly 1.4 million barrels a day from Iran, so the dependency is real; it simply sits in different places than a power-grid analysis would predict, and that mismatch is precisely what makes it an exploitable leverage point.
  • The 90-day clock is the single most important number, and it cuts both ways. The first thirty days are strategic reserve deployment, rationing and spiking transport costs. Thirty to ninety brings feedstock depletion biting into industry, with export manufacturing slowing and freight disruption becoming visible. Ninety to one-eighty leaves reserves thin, military operations constrained and political pressure to resolve or escalate acute. Beyond that comes structural damage. The clock pressures Beijing toward resolution — but it also gives China enough runway to outlast American domestic tolerance for oil above $100. Whoever blinks first loses the bilateral negotiation.
GeoHussarThe vacuum of space and the American flag #space #outerspace #moon #debunked
  • Sound can't travel in a vacuum because it needs a medium like air, he explains. Without air, kinetic energy from an action instead dissipates as vibration and eventually heat and infrared radiation within the object itself, rather than propagating outward as sound.
  • He applies this to debunk the moon-landing flag-waving claim. The flag was extended by a horizontal rod, and with no air resistance, the kinetic energy from astronauts moving it had nowhere to dissipate quickly, causing prolonged movement rather than the instant stillness deniers expect; lower lunar gravity, one-sixth of Earth's, plays a minor supporting role.
  • He corrects the common attribution of the conservation-of-energy law to Einstein. He argues the principle — that energy changes form but is never destroyed — may instead originate with Joule in the 1840s.
Loss DogWhy you should be maximizing your HSA! #throwyouabone #shorts #hsa #wearelossdog #taxes #savings
  • HSAs are the only accounts offering a triple tax advantage. Contributions, growth, and withdrawals for medical expenses are all tax-free, but eligibility in 2026 requires a high-deductible plan with a minimum deductible of $1,700 individual/$3,400 family and a maximum out-of-pocket of $8,500/$17,000.
  • 2026 contribution caps are $4,400 individual and $8,750 family, plus a $1,000 catch-up over 55. Eligibility also broadened under 2025 tax changes: bronze and catastrophic ACA exchange plans now qualify, and pre-deductible telehealth visits no longer disqualify an account holder.
  • The "shoebox strategy" lets medical receipts sit unreimbursed for decades while the HSA balance grows in the market. After age 65, the account is said to convert to IRA-like treatment, allowing withdrawals for any purpose with no penalty, just ordinary income tax, while non-medical Medicare-era withdrawals remain penalty-free.
Professor G MarketsViolent Backlash: What the Sam Altman Attacks Signal for AI
  • This post includes paid content that is not available publicly. See the original source.
Dr. Jon KrohnHow One School Used AI to Become the Highest Performing in Boston (with Traci Walker-Griffith)
  • The school adopted AI to deliver writing feedback faster. Walker-Griffith says that in spring 2024 teachers could not get feedback to students fast enough to move outcomes, so the school, an innovation school, began using Claude as a teacher-facing tool and prompt-engineered it to give feedback on student writing. She calls the early prompting very frustrating but says the shared struggle made teachers better together, which she ties to Linda Hill's idea of "collective genius".
  • The model was chosen for seeming less prone to hallucination. In October 2024 the school took the problem to an AI Collaborative convening in Denver run with The Learning Accelerator and Playlab, and settled on Claude because it then seemed less prone to hallucination and could serve as a repository for the work; a year-long effort followed with the grades 5–8 ELA/humanities team.
  • Feedback had to be differentiated by how far behind students were. She says one good prompt only served one student, so rubrics and reading levels were built in; asked for their views, on-level students wanted more feedback while those furthest behind needed targeted small-group instruction. Pre- and post-assessments across three feedback iterations showed the largest gains among multilingual learners and students with disabilities, some of whom need over a year and a half of growth.
  • Student-facing AI came later, and is split by age. When Gemini was released in the district's Google suite in spring 2025, with a recommended level of grade 7 and up for student-facing use, the school worked with the district to use it in grades 5 and 6. Kindergarten to grade 4 do not see AI directly and teachers use Gemini gems for small-group planning; grades 5–8 use it student-facing. She credits the school's turnaround from under-enrolled and on the closure list to highest-performing in Boston and 13th in Massachusetts to constant learning.
Professor G MarketsTrump Spirals as Iran Blockade Triggers Recession Fears — with Sen. Chris Murphy
  • This post includes paid content that is not available publicly. See the original source.
GeoHussarAluminum helped us get past the Van Allen belts to reach the moon and back #space #apollo #radiati
  • Aluminum's strength-to-weight ratio let Apollo capsules carry thick shielding without adding launch weight. He presents this as the key reason astronauts crossing the Van Allen belts came back with low radiation doses, since thick aluminum alloy shielding could surround the capsule.
  • Aluminum's light nucleus kept secondary radiation low. With only 13 protons and 14 neutrons, aluminum mostly produces x-rays rather than the gamma rays and loose neutrons a heavier shielding material would generate when struck by the belts' protons and electrons — a secondary effect he identifies as Bremsstrahlung radiation.
  • He frames this as the answer to moon-landing skeptics who doubt astronauts could survive the belts. The Van Allen belts' inner zone of fast protons and outer zone of fast electrons are ionizing radiation capable of damaging DNA, but he argues aluminum shielding absorbed the brunt of the impacts.
Loss DogUnlock Explosive Opportunity: Master High Volatility Investing #shorts #investing #marketingstrategy
  • Higher volatility always means more opportunity, not more risk to be avoided, the host argues. He claims there is "no such thing as too much opportunity" because elevated volatility tends to coincide with very low basis or prices.
Dr. Roman YampolskiyWill AI take my job? Expert Explains!
  • Computer-based cognitive jobs are the most exposed to automation. The speaker says any purely symbolic work, such as accounting, tax prep, web or logo design and programming, will be "very easy" for AI to automate. Jobs needing physical presence, like plumbing and construction, are safe until humanoid robots become much more capable and affordable, "a few good years". So are roles where people prefer a human, such as senseis, gurus and hiking guides.
  • Human-level AI breaks the pattern of automation creating new jobs. Any new job created by advanced technology can also be done by AI, the speaker argues, so "a very high degree of unemployment" is expected. The advice is to read up on the state of AI in a field, and its expected needs 4–5 years out, before starting a college major.
GeoHussarThe IRGC has more to show for this war than the United States by far #war #f15 #a10 #iran #economy
  • Every day the IRGC survives while keeping the Strait of Hormuz shut counts as a win for Iran, he argues, at direct US expense. He cites Brent crude at $106 and WTI at $103, with the "real" price closer to $141/barrel — the highest since the 2008 financial crisis — as evidence of the war's economic toll on the US.
  • He reads the same-day loss of a US F-15E and an A-10 as a sign of Russian-assisted Iranian targeting, not coincidence. One F-15E crew ejected, with the weapons system officer still missing and a bounty on him, while an A-10 crashed under unclear circumstances; he notes Russia is reportedly sharing intelligence with Iran on hitting American targets.
  • He frames Trump's threat to exit NATO and reluctance to commit to toppling the IRGC as abandoning Gulf allies who begged the US to finish the war. He also points to rising US debt-interest costs despite wartime "safe haven" expectations as a sign markets are reading the war as destabilizing rather than reassuring.
Stock Tok62 Days of Waiting is OVER. The 1st 2 stocks we are buying now
  • This post includes paid content that is not available publicly. See the original source.
Polis PanditThe Iran Oil Shock Exposed Years of False Alarms
  • This post includes paid content that is not available publicly. See the original source.
All Things FinancialNo matter the strategy, what decides the war or peace is the ability to control the chokepoint 4-14-2026
  • He names a structural failure mode that he argues has never been formally entered into a planning document: allied financial defection. The dollar coercive architecture rests on two pillars — Iran cannot receive dollar payment for oil, and any third party routing around that faces exclusion from the US financial system. The second pillar functions only while allies fear dollar exclusion more than they fear the energy or economic cost of compliance. When a treaty ally buys sanctioned Iranian oil in euros or won, the legitimacy architecture collapses across three dimensions at once: allied political support, international legal standing, and domestic political will.
  • The evidence he assembles is a list of defections rather than an argument. Japan agreeing to buy Iranian oil in yuan with ships already en route. The UAE accepting yuan. Saudi Arabia agreeing to yuan and offering to accept euros. India resuming Russian crude once sanctions lifted, having stopped earlier only under threat to its banks. Russia and Venezuela running barter and yuan. And Rubio stating that sanctions will not be a topic of discussion in five years — which he takes as the architecture's own operators conceding the point.
  • His qualification cuts against the loudest version of the story, which is what makes it worth keeping. The world will not simply reprice oil in yuan, because China will not permit it. Most yuan trade is bilateral: deposit into Kunlun Bank and draw credit to buy Chinese goods. Beijing has no intention of opening the capital account to make the yuan a general settlement currency, because that would cost it control. He similarly deflates the crypto story — Iran uses it for the transit fee because it is not set up on CIPS, not as strategic adoption. What he does treat as structural is the Shanghai gold exchange now offering trading, clearing, storage and hedging as a physical venue where the Western equivalents are paper.
Crosstalk SolutionsWe Built Our Own 5G Carrier
  • Crosstalk has launched its own 5G data SIMs, Crosstalk Mobile. The company says the SIMs are truly unlimited on a nationwide 5G network, month-to-month with no hidden fees, and sold as an authorized MVNO with a direct carrier agreement rather than through gray-market resellers. Billing starts at activation, not at shipping.
  • Its pitch is support from network engineers, not a call centre. The company says it will help choose hardware, diagnose failed 5G failover, decide on directional antennas and aiming, lock to the best cellular bands for a site, and design bonded field deployments with Peplink SpeedFusion.
  • MSPs get volume pricing and out-of-band management. A partner program lowers pricing as more SIMs are deployed across client sites, and the company argues each SIM gives the MSP a back door to troubleshoot when a client's main internet is down, avoiding a truck roll, while earning recurring revenue.
  • Static IP addresses over 5G are available. The company says it can provide static IPs on cellular on request; it does not give prices for any plan in the video, and says more plans are coming.
Professor G MarketsThe Oil Blockade Isn’t Spooking Markets — Yet
  • This post includes paid content that is not available publicly. See the original source.
Dr. Jon KrohnBoston's #1 Principal: AI Is My Secret Weapon (Ep. 983 with Traci Walker-Griffith)
synopsis

Traci Walker-Griffith, principal of the Eliot School in Boston, a public K–8 school, ends up arguing that AI in children's education should be deployed carefully and intentionally rather than avoided. In her account it works when it is tied to a defined problem of practice, measured against data, kept away from the youngest children's direct use, and built on a culture of adult learning. She cites her own school's results as evidence and plans to spread the approach after retiring in June.

The school and its foundation. She arrived in 2007, when the Eliot was on the closure list with 150 students; it now has 800 across three buildings, and she calls it the highest-performing school in Boston and 13th in Massachusetts. Children from age 4 program KIBO and Bee-Bots, learning that input quality determines output, and start robotics in fourth grade.

Teacher-facing feedback with Claude. In spring 2024 teachers could not get writing feedback to students fast enough to move outcomes, so the school began prompt-engineering Claude to do it. In October it joined an AI Collaborative convening in Denver with The Learning Accelerator and Playlab, choosing Claude because it then seemed less prone to hallucination. Early prompting was frustrating, and one good prompt served only one student, so rubrics and reading levels were built in. Students on or above level wanted more feedback; those furthest behind needed targeted small-group instruction. Pre- and post-assessments across three feedback iterations showed the largest gains among multilingual learners and students with disabilities.

Splitting by age. After Gemini arrived in the district's Google suite in spring 2025, recommended for student-facing use from grade 7, the school worked with the district to use it in grades 5 and 6. In kindergarten to grade 4, teachers feed anonymised writing into Gemini gems to spot class patterns in 5 to 10 minutes. In grades 5–8 students work with AI directly: before the state test, fifth graders compare their answers with Magic School's and log errors for their teacher. She says the school stays "platform informed, not platform loyal".

Failures and measurement. Feedback failed by being generic, over-encouraging and over-trusted, and the school regrets not recording which prompts worked. It tracks writing growth, teacher time, a youth-outcomes survey, metacognition and ethical understanding of AI, conceding that metacognition remains very hard to quantify.

Home and books. Asked whether keeping younger children off AI should be the general rule, she does not endorse it outright: two-year-olds already have iPhones, so families must raise critical consumers and audit what children use at home. She insists on paper books and at least 30 minutes of reading at home and at school.

Answering Brookings. The host cites a January Brookings report concluding that generative AI's risks in children's education outweigh its benefits. She reads it as "deploy carefully", argues AI should amplify teachers rather than replace them ("if you don't feed the teachers, they eat the kids"), and calls for sharing practice so the US can reach the top of PISA.

  • A collaborative culture, not AI alone, turned the school around. Walker-Griffith says she took over the Eliot School in Boston's North End in 2007, when it was on the closure list with 150 students; it now has 800 students across three buildings (ages 3 to 14) and is, she says, the highest-performing school in Boston and 13th in Massachusetts. She credits a culture of collaboration and "failing forward", and frames AI as the latest tool on that foundation; she retires in June.
  • AI writing feedback produced the largest gains for students furthest behind. From spring 2024 her grades 5–8 ELA/humanities teachers prompt-engineered Claude, then the school's choice for seeming less prone to hallucination, to give feedback tuned to each student's reading level. She reports that pre- and post-assessments across three feedback iterations showed the biggest gains among multilingual learners and students with disabilities, some of whom need over a year and a half of growth.
  • Younger pupils never see the AI; older ones work with it directly. For kindergarten to grade 4, teachers feed scanned writing with no student identifiers into Gemini gems loaded with rubrics and standards, spotting class-wide patterns in 5 to 10 minutes instead of passing papers round. Grades 5–8 use AI directly: in a Magic School unit before the state test, a fifth grader writes his own answer, compares it with the AI's and explains which is better. Her line: "AI raised the floor, teachers are raising the ceiling."
  • Generic praise and over-trust were the failures that forced redesigns. She names three: generic feedback, too much encouragement, and over-trust, so teachers must check everything before it reaches a child. She also regrets not keeping a shared record of which prompts worked, since teachers experimented separately.
  • She reads the Brookings warning as "deploy carefully", not "don't deploy". The host cites a January Brookings report concluding that generative AI's risks in children's education currently outweigh its benefits; Walker-Griffith answers that children are not left alone in front of AI, and urges AI-enabled practice tied to a defined problem, deep teacher professional learning ("if you don't feed the teachers, they eat the kids") and sharing practice nationally. She hopes this could help lift the US to the top of the PISA rankings.
Professor G MarketsChina Steps In as Trump’s Ceasefire Unravels — with Eyck Freymann
  • This post includes paid content that is not available publicly. See the original source.
Dr. Roman YampolskiyWelcome to the Roman Forum
  • Yampolskiy launches the Roman Forum podcast on superintelligence and related topics. He introduces himself as a computer scientist, professor and safety researcher, "which just means I really worry about the future". The podcast will cover superintelligence, simulation, consciousness and the singularity, with guests who are scientists, philosophers and founders "making it happen".
Professor G MarketsThe Democrat Who Could Flip Texas and Pivoting Out of a Dying Industry
  • This post includes paid content that is not available publicly. See the original source.
Professor G MarketsTrump Blockades Hormuz as Ceasefire Collapses, and Orbán Loses Power
  • This post includes paid content that is not available publicly. See the original source.
Loss DogUnlock Trading Success: Master Capital Efficiency NOW! #tradingstrategy #trading #tomsosnoff #shorts
  • Capital efficiency depends on a broker actually offering minimum buying-power requirements, the host argues. Many brokers don't use the lowest buying-power requirements they're permitted to, which he says limits which strategies and underlyings a trader can access — and is one of several factors he weighs when comparing brokers.
  • Roughly 80-85% of underlyings should need only minimum capital under normal conditions, by his estimate, covering about two standard deviations of risk; the remaining 15-20% legitimately require more capital for risk purposes.
All Things FinancialA one-time utility becomes a national security when it can be turned off, blocked or weaponized. 4-13-2026
  • The ownership concentration is the fact that starts the episode, and it is startling. Four companies — Google, Meta, Microsoft and Amazon — own or lease most of the world's subsea cable capacity. Per an EU Commission report, they hold 90% of transatlantic routes and 70% of all international bandwidth globally, up from under 10% in 2012. Google alone has 33 subsea cables; Meta owns the longest system in the world, encircling Africa across 33 countries and three billion people. Amazon's routes run from Qatar, where the data centres were built on cheap gas — which is no longer cheap.
  • Three layers matter and only one of them is the pipe. The cable, the landing station, and the data node where traffic is actively handled. Nationalizing the node captures both the pipe and the intelligence-collection architecture simultaneously, which is why the realistic scenarios target the landing station first — it is the single leverage point that reaches both.
  • The mechanism already underway is nationalization without the word, and he explains why it accelerates. Requiring landing stations to be majority domestically owned, annually licensed, subject to real-time government access and staffed by cleared nationals — India furthest along, then Indonesia, with the EU moving regulatory rather than ownership. His explanation of the motive is the sharp part: under FISA the US does not intercept everyone's calls, it sits at the data nodes and watches traffic pass. Which means other governments have now understood that their health records, tax structures, corporate data and government records all transit infrastructure a foreign state can read. He expects most of these cables to be nationalized within a year or two, and reads Europe's migration off Microsoft toward Linux as the same reclaiming impulse arriving one layer up.
Mr. GlobalChrist Like: The GOP Has Lost Moral Authority
  • Republican silence on the Pope attack tests the party's limits. He reports that Trump posted a long attack calling the Pope weak, saying nobody had heard of him and that he was chosen only because Trump is president, then posted an image of himself as Jesus Christ. As of filming, four or five hours later, he had heard nothing from any Republican.
  • Staying quiet trades short-term safety for lasting loss of credibility. He argues the party's religious brand is foundational with Catholics and evangelicals, and that a party which won't defend a religious figure when inconvenient is "selectively faithful when it aligns with power" — which he says he always believed it was.
  • Republicans have little left to protect by deferring to Trump. In his reading, loyalty to Trump has already cost them: he expects them to lose the House definitely and the Senate most likely, so Trump cannot do them more damage than he already has.
  • Faith voters will leave quietly rather than loudly. He expects suburban Catholics and older churchgoers who stay off social media but vote to drift away slowly, and warns that normalizing this out of fear makes it hard to draw any line later.
  • Any pushback will be strategized, not spontaneous. He predicts the leadership will clear only safe-seat or retiring members to object while those in tight races are told to stay quiet, with reluctance to criticize a president in wartime as a further brake.
Dr. Jon KrohnWhy AI Adoption Is a Human Problem (Not a Technical One)
  • Staged autonomy exists for humans to catch up, not the AI. Asked by the host about a four-phase adoption model running from assessment to scaled autonomy, Glickman says no system should be switched straight to full autonomy: teams must understand the problem and how a human would solve it, then understand how the AI succeeded before they are willing to let go. He warns that impressive coding agents without guardrails, context and human oversight can do more harm than good.
  • Enterprises are won by showing, not telling. Glickman says pain and backlogs are so large that if it works "it's a no-brainer", and that his company's head agent, Eve, signs on as just another developer: it follows existing processes, runs in development, and submits audited, documented pull requests to the CI/CD pipeline. He argues the risk is no worse than hiring a new employee, since existing controls already stop rogue developers, so the client's only loss is that it might not work.
  • Buy outcomes rather than building your own agents. He says the main obstacle is disbelief, often from bad experiences with coding agents or failed in-house agent builds, which he calls a fallacy. Businesses never cared how a consultant staffed a migration, only about the outcome and cost, so firms should "pick a winner" for non-core work instead of chasing a technology moving out of reach.
Professor G MarketsDon't Try to Beat This Market — Here's What to Do Instead
  • This post includes paid content that is not available publicly. See the original source.
GeoHussarThere are over 10,000 starlink satellites in orbit right now.. #space #starlink #satellite #inte
  • SpaceX's Starlink constellation of over 10,000 satellites accounts for 65% of everything in Earth orbit, dwarfing every state-run program combined. The closest rival, Europe's OneWeb, has only about 600 satellites; Russia, China, Japan, and India have nothing comparable.
  • Starlink has become integral to the Ukrainian military, while Russia's access has been fully cut off for about a month. SpaceX sent thousands of terminals to Ukraine after the 2022 invasion; Russia, never officially granted access, relied on smuggled terminals for drone coordination before losing even non-military access.
  • He calls extending global broadband access, including to Iranian protesters who used smuggled terminals in January, one of the most consequential projects underway. He estimates 40,000–50,000 satellites would be needed to fully serve the roughly 2.2 billion people currently without connectivity.
All Things FinancialOil is fought over on sea lanes; Data is fought over at control points. 4-12-2026
  • His framing question is not whether tolls are possible but whether they are about to generalize. Panama has charged since 1914 and Suez since 1869, so toll-taking at chokepoints is ordinary. The live question is whether the current environment triggers the model spreading across all eight major chokepoints simultaneously. And the constraint is always the same: geography creates the toll potential, but only sovereign enforcement capacity converts potential into revenue. Poland can enforce because it has a real military; the Red Sea scores maximum toll viability on every dimension except the one that matters, because no stable sovereign controls it.
  • Four archetypes, distinguished by stability rather than by size. Formalized tolls — explicit, legal, arbitrable, and highly stable, as at Panama, Suez and the Bosphorus. Implied tolls — priority partners receive priority service, with access traded for supply agreements and security guarantees rather than dollars, which he rates the most stable of all precisely because it is deniable. Weaponized tolls — closure or threatened closure extracting political concessions rather than fees, episodic and destabilizing, as with Iran and the Houthis. And the sabotage tax — infrastructure damage raising cost for everyone while generating revenue for no one, the least stable form, which is what Russia is running in the Baltic.
  • His specific forecasts are datable, which makes them checkable. Malacca formalizes within ten to fifteen years, with Indonesia and Malaysia applying demographic and economic pressure on Singapore — which has the institutional architecture to run a toll but no army or navy to defend one. The North Atlantic toll is denominated in security guarantees rather than cash, with Iceland extracting sovereignty protection instead of fees. And Hormuz weaponization is strategically self-defeating on a ten-year view as bypass infrastructure matures — Saudi Arabia already building across, with Israel offering a route the Arab states are unlikely to take.
All Things FinancialUCI Chokepoint Futures
synopsis

The paper's conclusion is that toll-taking is becoming the norm across all eight major maritime chokepoints, but the transition most likely runs through Bloc Fragmentation (45% probability) rather than a cooperative Polycentric Toll Order (38%), because F5 (Security) — sovereign enforcement capacity, not geography — is the binding constraint everywhere; and beneath the physical layer, undersea cables have already become a covert data chokepoint that the Five Factor framework must now score separately from physical infrastructure.

Four toll archetypes. Formalised tolls (Panama, Suez, eventually Malacca) are stable and transparent; implied tolls (Singapore, Djibouti) trade access for security guarantees rather than fees; weaponised tolls (Hormuz, Red Sea) extract political concessions and are judged strategically self-defeating since they accelerate the bypass infrastructure that erodes the toller's own leverage; sabotage taxes (Baltic Sea) generate no revenue at all, only destroyed value.

Eight chokepoints scored. Hormuz and Red Sea both score critical on food and energy but score no stable toll collector; Malacca is the most mature implied-toll operation and is judged likely to formalise within 10–15 years as Indonesia and Malaysia demand a share; Taiwan Strait is unique as a Technology-factor chokepoint (90% of advanced chips) rather than an energy one; Djibouti's multi-power basing model is called the most durable toll structure anywhere.

The data layer. UCI-GBL-002 splits Technology Factor 3 into physical infrastructure (3A) and intelligence capability (3B), arguing control has moved from oil-style volume control to content control via three layers — physical repair, connectivity/landing, and covert interception — with the US, UK, China, and Russia already operating at Layer 3. Sri Lanka's Galle Depot, uniquely high on 3A and zero on 3B, is recommended to monetize that neutrality explicitly via a proposed "Galle Neutral Certification" rather than trade it away for an intelligence partnership with any single power.

  • The four toll archetypes are formalized with stability rankings, and the ranking is counterintuitive in a useful way. Formalized tolls are high stability. Implied tolls — priority partners receiving priority service, paid in supply agreements and security guarantees rather than dollars — score very high stability precisely because they are deniable and therefore sustainable. Weaponized tolls are low stability because they are episodic and destabilizing. And the sabotage tax is the least stable of all, since it funds no one while destroying system efficiency. In every case Security is the binding constraint: geographic position creates toll potential, sovereign enforcement capacity converts it to revenue.
  • The probability matrix scores all eight chokepoints across all five factors and produces verdicts rather than narrative. The Red Sea scores high on almost everything and still cannot extract, because there is no stable sovereign controller — the entry's own line is that toll potential is enormous while sovereign capacity is absent. Hormuz is weaponized and ongoing. The Baltic is a sabotage tax being executed now. Malacca is the formalization target, with Singapore holding the institutional architecture.
  • The document's real contribution is the third layer, which the oil analogy misses entirely. UCI control decomposes into physical infrastructure, connectivity, and data sovereignty — and the data layer carries an intelligence premium and a deniability advantage that no commodity chokepoint has, because reading traffic leaves no trace and denies nothing. That prompts a proposed framework revision splitting the Technology factor into 3A and 3B to capture the distinction between physical technology capacity and data-layer control.
  • Three macro futures are offered rather than a single forecast, each with per-chokepoint scenarios at three, seven and fifteen years. A polycentric toll order, bloc fragmentation, or digital bypass — with the structural inversion of the Pax Americana era as the through-line, and derived concepts including a neutral depot premium for states that can credibly certify non-alignment, and a Data OPEC scenario in which chokepoint holders coordinate on data transit the way producers once coordinated on oil.
GeoHussarUkraine, using sea drones as motherships for their fiber optic drones. #Science #drones #fiberopt
  • Ukraine has paired sea drones with fiber-optic drones as motherships, extending jam-resistant strikes far beyond their normal range. Standard fiber-optic drones trail cables 30–50 km long and can't be jammed since they aren't radio-controlled, but Ukraine's naval drones can carry that tether dozens, even up to 100 km, further out before launching.
  • He argues this opens the entire Russian-held Sea of Azov coastline, including Crimea, to Ukrainian strikes. Ukraine has also used sea drones to sink or disable a Russian submarine, and while Russia has nearly caught up on land-based fiber-optic drones, it has not matched the sea-drone-mothership approach.
  • He is skeptical of Russian claims to already have an equivalent fiber-optic sea drone. He compares it to earlier Kremlin media claims that the Kinzhal missile was hypersonic, noting every missile becomes hypersonic during its terminal phase regardless.
GeoHussarBefore today, the US’s last lunar Mission was in 1972. #SpaceExploration #space #moonlanding #as
  • The 1969 Apollo 11 moon landing was a direct response to the 1957 Sputnik launch, not a standalone goal. Sputnik's October 4, 1957 launch triggered the space race; the US didn't answer with its own satellite, Explorer 1, until January 31, 1958, and the Soviets pulled further ahead with Yuri Gagarin's 1961 orbital flight and Gherman Titov's 25-hour flight.
  • Kennedy's 1961 pitch to Congress aimed to overtake a Soviet Union that was outright winning the space race, not merely competing. The Soviets targeted a crewed moon landing by 1967–68 but were stalled by repeated N-1 rocket failures, managing only a robotic lunar landing in 1966.
  • The roughly $300 billion, ten-mission Apollo program wound down once Soviet lunar competition ended. NASA then consumed about 4% of the federal budget; all missions after Apollo 11 landed successfully except Apollo 13, which had to abort, and NASA's budget was cut afterward as its role seemed less urgent.
Dr. Jon KrohnWhy AI Should Say “I Don’t Know” More Often (with Matt Glickman)
  • Reasoning models made reliable self-reported confidence possible. Glickman says that when reasoning models arrived "early last year", the under-discussed win was that they could be asked how confident they were that they had done everything asked, answering, for example, "90% confident" and naming the one point needing clarification. Genesis now asks for confidence at every step of complex missions: show the artifact if confident, retry or escalate the missing pieces if not.
  • Unasked, agents bluff the way humans do. He says a model will otherwise guess a formula it could not find and move on unnoticed, just as people make leaps of logic unless asked how they reached a conclusion. He dismisses threatening models for accuracy; the host mentions a study finding aggressive prompts gave about 10% more accurate answers, which Glickman says is not worth it.
  • Captured clarifications compound and pull agents up the stack. Glickman argues that an answer given once must be applied next time; stored securely, those answers build an understanding of a business's semantics and flows. He says customers now ask Genesis for work further up the stack, such as board presentations grounded in how the business actually operates.
Mohamed El-erianThe Weekly Look at the Global Economy and Markets
  • A historic threat from Trump to Iran gave way to US-Iran ceasefire-extension talks in Pakistan, even as March US CPI tripled to 0.9% monthly (3.3% annual) on war-driven energy costs. Core inflation stayed contained at 0.3% monthly and 2.6% annual, suggesting no broader spillover yet, while Saudi Arabia reported losing 600,000 barrels a day of production capacity and a further 700,000 bpd foregone to pipeline damage.
  • US consumer confidence hit a record low even as the country's stagflation risk remains milder than elsewhere. The University of Michigan's current-conditions gauge fell to a record 50.1 and expectations to their lowest since 1980, with one-year inflation expectations surging to 4.8%, while he notes several European countries saw energy-price protests and some developing nations face outright shortages.
  • The week ahead centers on follow-up US-Iran talks in Pakistan, which face seven unresolved issues beyond the Strait of Hormuz and Israel's attacks on Lebanon. Early Sunday signals pointed to no agreement, as the IMF's Spring Meetings prepare to present three scenarios that Managing Director Georgieva said would all show lower growth and higher inflation.
GeoHussarA walk-through as to how close Iran was to having a nuke #ScienceExplained #chemistry #uranium #n
  • He describes uranium enrichment as a long industrial chain Iran controls end to end, from its own domestic ore deposits to centrifuge cascades. Mined ore is milled into yellowcake (80% uranium oxide), converted to a gas with fluorine, and spun through thousands of centrifuges to progressively concentrate uranium-235 over repeated cycles.
  • Enrichment past 20% is unambiguously weapons-grade territory, and 60% is a short step from bomb-usable material. Civilian power only requires 5% enrichment; he states Iran held 440 kg of uranium enriched to 60% even before Operation Midnight Hammer in 2025, and reaching 90% from 60% is comparatively quick and easy.
  • He frames having enriched uranium as only one of several remaining steps toward a usable weapon, and offers to cover those additional steps in a follow-up video depending on interest.
All Things FinancialMalaysia
  • Malaysia holds three non-overlapping chokepoints at once, and only the combination is interesting. Maritime geography at the Strait of Malacca, heavy rare earth separation, and semiconductor back-end packaging at roughly 13% of the global share. No single one of these would lift Malaysia above a standard toll state read. Together they make it structurally the most complex leverage state in Southeast Asia and an entry in its own right rather than a footnote to the regional analysis.
  • The Lynas plant at Kuantan functions as an operational monopoly asset for the entire West. China controls approximately 91% of global rare earth refining, and LAMP is the only large-scale non-Chinese heavy rare earth separation facility in existence — serving non-discretionary demand in defence, EV motors and wind turbines with no bypass available at scale. The seven-to-ten-year timeline to replicate a comparable facility means the position is structurally durable regardless of what Malaysian policy decides, which is an unusual thing to be able to say about a single plant.
  • The dual-track architecture is simultaneously the thesis and the risk, and the entry is clear that it may not hold. Malaysia supplies roughly 10.6% of China's rare earth imports while anchoring the Western hedge through Lynas — serving two adversarial supply chains without formal commitment to either. That is the maximally efficient toll posture: extract rent from all parties. But it carries the structural vulnerability that both parties may eventually demand exclusivity, and specifically that Chinese regulation could force China-linked extraction to be refined domestically only. The resolution of that tension is named as the primary structural risk to the entire Malaysia thesis.
All Things FinancialMarkets are efficient until nations decide they are not. Financial theory is elegant; geopolitics decisive 4-11-26
  • He revisits the Intel call and disagrees with the people now agreeing with him. He recommended it when no conventional investment framework supported the position — the consensus said break it up, the technology was poor and bankruptcy was plausible. Now the same analysts are recommending it on the numbers, which he thinks is still the wrong framework. His actual thesis was never financial: Intel began moving roughly a week after the war started, because the world absorbed that the United States could not hold Hormuz. Which raises the question that follows automatically — whether Washington would defend Taiwan sixty miles from a China with over 250,000 missiles and a genuine navy and air force, when Iran achieved closure with a fraction of that.
  • The concentration point is the whole argument compressed. Every major American semiconductor name except TSMC depends on TSMC. His standing question from two years ago was what Nvidia would be worth the day after Taiwan and China combine by any route — conquest, agreement or attrition — and his answer was less than a dollar. Against new price targets running from $67 to $150 and one house arguing Intel overtakes Nvidia within months, his own view is that the convergence is more likely to come from Nvidia falling than from Intel rising.
  • Japan is doing the same arithmetic with public money, which he takes as corroboration. Another $4B approved for Rapidus, stated explicitly as guarding against external shocks including Sino-Japanese tensions — because if Taiwan ends up with China, Japan may simply lose access to advanced chips. He then turns to Malaysia as the same lesson one layer down: understanding a single company is no longer sufficient when the supply chain runs through a toll strait, a single rare earth separation plant and a packaging cluster. Financial theory is elegant; geopolitics is decisive.