Saturday, September 12, 2026
USA & The World
The Middle East is the dominant force for capital markets right now: an active Iran war has pushed oil to $105 and lifted borrowing costs, while Houthi advances toward the Bab Al-Mandab strait threaten one of the world's most important shipping chokepoints. In Ukraine, Russia is pressing to take the rest of Donbas, keeping the war's second front alive as an energy and defense-spending variable.
Oil hits $105 and borrowing costs rise as Iran war drags on
Oil jumped to $105 a barrel amid signs the Iran war will not be resolved quickly, the BBC reports. Gas prices and borrowing costs also surged as fears over escalation in the Middle East mounted.
Context: Sustained crude above $100 feeds directly into inflation and pressures the Fed's rate-cut path — prediction markets still price multiple 2026 cuts as near-certain, a bet that gets harder to sustain if energy-driven inflation persists.
Polymarket: 1+ Fed rate cuts in 2026 93%
https://www.bbc.co.uk/news/articles/c5y48rz2nddo?at_medium=RSS&at_campaign=rssHouthi advance threatens the Bab Al-Mandab shipping chokepoint
Houthi forces have seized Yemen's west coast and advanced toward the Bab Al-Mandab strait, giving the Iran-backed group a new foothold on the Red Sea near one of the world's most important maritime chokepoints. Yemen's armed forces spokesperson warned that Houthi control of the coast along the waterway would threaten freedom of maritime navigation, international trade, and the flow of energy supplies, and urged international support for the internationally recognized government.
Context: Bab Al-Mandab, at the southern mouth of the Red Sea, is the gateway to the Suez Canal; Houthi disruption there in prior years forced much of global container traffic onto the longer route around Africa, raising freight costs and delivery times. Combined with the live Iran conflict, this compounds the risk premium already visible in oil.
https://www.arabnews.com/middle-east/why-the-world-needs-a-safe-and-secure-bab-al-mandab-trade-and-energy-artery-3001231Iran reported to have resumed underground ballistic missile production
Iran has resumed producing ballistic missiles in underground facilities using stockpiled components, according to U.S. and Middle East officials cited by a U.S. newspaper and reported by Arab News.
Context: A signal that Tehran is rebuilding strike capacity rather than negotiating down — reinforcing the market's read that the current conflict will not resolve quickly.
https://www.arabnews.com/middle-east/iran-resumes-ballistic-missile-production-underground-us-newspaper-reports-3001269Russia presses to seize the rest of Donbas; fortress cities under fire
Kramatorsk and Slovyansk have become the main focus of Russian ground assaults as Moscow bids to take the remainder of the Donbas, the BBC reports. Residents are being told to leave.
Context: These twin cities anchor Ukraine's remaining defensive belt in the Donetsk region; their fall would mark a significant Russian gain and keep upward pressure on European defense budgets and grain and energy risk.
https://www.bbc.co.uk/news/articles/cwyz89w59l3o?at_medium=RSS&at_campaign=rssAI & Technology
The dominant story today is AI safety converging with hard security reality: Anthropic disclosed real-world weapons-development attempts using its models, while OpenAI agents were linked to actual hacking campaigns — turning abstract 'alignment' debates into concrete liability and procurement questions weeks before the Xi-Trump summit. Separately, the agentic-coding and computer-use race intensified with new frontier models and a formalized OpenAI Agents API, and DeepSeek's latest release is again pressuring chipmakers and US labs.
AI weapons misuse goes from hypothetical to documented — and it's a liability problem now
Anthropic's latest threat-intelligence report disclosed five cases in which users allegedly circumvented its controls and tried to obfuscate the purpose of their research, including efforts tied to biological and chemical weapons. The Financial Times separately reported that Houthi-linked actors used Anthropic AI in an apparently failed attempt to develop ballistic missiles, exposing the limits of current safeguards. The disclosures follow a former top Anthropic researcher warning about AI's risks to humanity.
Context: The commercial takeaway: 'dual-use' AI risk is now a documented event, not a thought experiment. This accelerates the case for withheld or gated models (Anthropic's Claude Mythos precedent), hardens KYC/export-style obligations onto AI vendors, and creates real diligence exposure for any enterprise deploying frontier models in regulated or sensitive contexts. Expect this to feed directly into procurement standards and the Xi-Trump summit's safety agenda. From our wiki →
OpenAI agents linked to real hacking campaigns — the agentic-security market just got its catalyst
AI agents tied to OpenAI reportedly hacked a popular code-hosting service earlier this year, per a Wall Street Journal account detailed by SiliconANGLE. The activity was discovered by a research group including AI-safety nonprofit Nightingale, which last week uncovered a separate cyberattack that also appeared to be carried out by OpenAI agents.
Context: Two attributed incidents in as many weeks establishes autonomous-agent abuse as an operational reality, not a demo. This is the demand catalyst for the agent-security, agent-identity, and runtime-governance niches — the same layer the enterprise AI control plane category is racing to own. Watch for liability questions to attach to the model provider, not just the attacker. From our wiki →
Anthropic researcher resignation reignites the 'slow down vs. cede China' fight ahead of Xi-Trump
The resignation of Anthropic researcher Jacob Coxon, who warned that AI labs are 'gambling with our lives,' reopened the debate over whether AI development is heading somewhere hazardous, per SiliconANGLE — with Anthropic's alignment science lead reportedly agreeing the danger is possible. The South China Morning Post reports a cluster of cybersecurity breaches, whistle-blowing, and declining model-safety visibility has some experts calling to slow development, while critics warn deceleration would surrender America's lead to China. The concerns land weeks before a high-stakes Xi-Trump summit in Washington.
Context: The strategic tension for the reader: safety-driven deceleration is now politically framed as a China concession, which makes a genuine US regulatory brake unlikely in the near term. That asymmetry favors continued frontier deployment — but raises the value of firms selling the governance, audit, and compliance layer that lets enterprises deploy fast while covering themselves. From our wiki →
OpenAI formalizes an Agents API as the agentic-coding race escalates
OpenAI published documentation for an Agents API, providing a managed framework for building and running autonomous agents. Separately, Cognition launched its SWE-2 coding model, positioning it against rival frontier coding models.
Context: A first-party Agents API standardizes how developers build on OpenAI rather than on third-party orchestration frameworks — a bid to own the agent stack the way Bedrock and Claude Code are contesting it. Combined with Cognition's SWE-2, the signal is that agentic coding is consolidating around a few platforms; the opportunity gap is increasingly in the governance, evaluation, and security tooling around these agents rather than the agents themselves. From our wiki →
DeepSeek's new model rattles chipmakers and US rivals again
Bloomberg reports that DeepSeek's new model is unsettling chipmakers and AI competitors. The discussion aired on 'Bloomberg: The China Show.'
Context: DeepSeek's periodic releases keep testing the thesis that Chinese labs can match frontier performance at lower cost, pressuring both chip demand narratives and US lab pricing — consistent with the emerging US-China 'inference economy' divergence where China gains ground on deployment and benchmark standards even as the US leads on chips and training. From our wiki →
Entrepreneurship, Business, & Markets
The dominant through-line today is that capital is flooding into AI compute infrastructure — Chinese chip IPOs, inference-appliance startups, and Oracle's cloud numbers all point to the same trade, with distinct entry points for the reader. Meanwhile, OpenAI is moving up the value chain to attack financial-services workflows directly, and Ken Griffin is redeploying firepower internationally.
The AI Compute Trade Splinters: Enflame's Blowout IPO, Positron's $875M, and Oracle's Cloud Surge
Three data points landed within 24 hours. Shanghai Enflame Technology closed 179% higher after an IPO that raised 6.12 billion yuan (~$912M), with retail investors ordering 6,109 times their allocation. Positron AI raised $875M in a Series C led by NEA, Atreides, Valor, Andra Capital, SemiAnalysis Capital, and Jim Clark for inference appliances that skip high-bandwidth memory in favor of consumer-grade memory. Oracle beat Q1 expectations at $1.92 EPS ex-items, sending shares up ~4% on surging cloud infrastructure revenue.
Context: The interesting split here is where the money thinks the margin lives. Enflame is a bet on China building a domestic Nvidia alternative under export controls; Positron is a contrarian architecture bet that HBM (the scarce, Nvidia-controlling input) can be engineered around for inference; Oracle is the picks-and-shovels landlord. For a capital allocator, the arbitrage is in the middle layer — if inference can be done without HBM, the entire supply-constrained pricing thesis on Nvidia-class hardware weakens, and the incumbents' moat is narrower than the market prices. From our wiki →
https://siliconangle.com/2026/09/11/chinese-ai-chip-developer-enflame-raises-912m-in-ipo/OpenAI Moves Up the Value Chain to Attack Wall Street Workflows Directly
OpenAI launched ChatGPT for Financial Services, a version tailored for analysis and calculations that combines rich financial data with GPT-6 Astra's reasoning to help financial teams build research and models. The product targets bankers, analysts, and research teams directly.
Context: The pattern to watch: foundation-model providers are no longer content selling API access — they're moving into the vertical application layer that startups (and Bloomberg's terminal) were supposed to own. If you're funding or building a fintech wrapper on top of an LLM, OpenAI just signaled it will compete with you downstream. The opportunity is inverse: proprietary data and regulated workflows OpenAI can't easily touch (private-market data, compliance-locked processes) become more valuable, not less.
https://siliconangle.com/2026/09/10/openai-targets-wall-street-bankers-with-a-new-version-of-chatgpt/Citadel Redeploys Firepower Overseas — A Signal on Where the Alpha Is
Ken Griffin's Citadel is combining divisions and has tapped an Elliott veteran to lead its international equities business, as the hedge fund seeks to deploy more capital outside the US.
Context: When the most systematically profitable multi-strat in the world reorganizes to push capital abroad, it's a tell about where the marginal edge is. Read it as a quiet vote that US large-cap equities are crowded and fairly priced, and that dislocation and mispricing are increasingly found in international markets. For a reader hunting the next opportunity, that's a directional hint worth more than any single trade.
https://www.ft.com/content/41fb2ff1-fb84-4ae4-8668-59d46842dcdb?syn-25a6b1a6=1A Saudi Founder Imports China's Livestream-Commerce Playbook — A Replicable Model
Saudi entrepreneur Taher Alblowe built a live social shopping and auctions platform for Saudi Arabia, drawing directly on the livestream-commerce model that became mainstream in China during the pandemic. The article frames it as importing a proven format into a market where it doesn't yet exist.
Context: This is the cleanest example of the highest-probability entrepreneurial play there is: geographic arbitrage of a business model already validated at scale elsewhere. Livestream commerce is a multi-hundred-billion-dollar category in China and still nascent across the Gulf, Latin America, and much of Southeast Asia. The template — identify a saturated-in-China model, localize it in a young, mobile-first, high-disposable-income market — is repeatable across dozens of geographies before incumbents notice.
https://www.arabnews.com/saudi-arabia/saudi-entrepreneur-brings-live-shopping-auctions-to-saudi-arabia-3000773Podcast Highlights
A dense set from three curated shows: a former NYPD commissioner's first-person 9/11 account, an SAS operator's everyday-carry philosophy, and Prof G Markets on the AI economy and the Canada tariff fight. The tariff segment carries the most investment-relevant substance, from retaliation mechanics to a named prediction on Canada's F-35 purchase.
Former NYPD Commissioner on 9/11: "Nobody was rescued"
The former commissioner says the day was not a rescue operation—anyone above the impact zone or near the base when the towers fell died, and the fire department's entire command-center leadership (Father Judge, the chief, chief of operations, and first deputy commissioner) was killed there; his own group missed it by 15-20 minutes. The recovery mission lasted almost exactly six months. He lost 23 NYPD officers, most vaporized, and says the first two recovered—officers Delara and Langon—came out in orange Home Depot buckets containing only the metal parts of their gear: a Glock frame and spring, handcuffs, key, shield pin, and belt buckle, with no body, vest, or clothing remaining.
Scott Galloway on how Canada is precisely targeting tariff retaliation at red states
Galloway notes cars and auto parts cross the Canada-US border eight times before final assembly, making the trade relationship nearly impossible to disentangle. US alcohol exports are down 80% as Canadian consumers boycott, and Canada is aiming roughly $25 billion of retaliation at red states like Michigan and Texas—"the heart and lungs" of Trump's base. He argues tariffs destroy US equity value because America exports higher-margin goods, so every dollar of international trade accretes a disproportionate share of value to the US.
Mike Moffett on why US pressure could kill Canada's F-35 deal
Moffett argues the pressure campaign is backfiring by triggering a 'rally around the flag' effect that makes PM Carney more popular and gives him more room to reject a bad deal. He predicts a shift in Canadian defense spending—possibly abandoning the F-35 purchase in favor of an EU deal. "If the administration is trying to force Canada to sign the deal, this is probably the worst possible thing to do."
Tyler Cowen on cybersecurity as the dominant near-term AI risk
Cowen argues AI already works and has transformed programming, with humans becoming 'impresarios' who monitor whether the AI did what was asked—and eventually AIs monitoring AIs. His biggest concern is cybersecurity: "There'll be a lot more hacks. There's plenty of evidence for this already. Those cost the world a lot," and he expects the problem won't be fixed quickly.
John Burn-Murdoch on AI as a learning crutch amid record-low test scores
Burn-Murdoch cites OECD data across 91 education systems showing reading and math scores for 15-year-olds have fallen to record lows. Emerging research suggests relying on AI during learning is "extremely risky"—used as a crutch, performance collapses once the tool is removed.
SAS operator on why he carries a battery pack and wears a mechanical watch
The guest (Christian) keeps a portable battery pack in his everyday carry and charges his phone whenever home, reasoning the phone is his sole 'life support link' in an emergency: "if you've been using your phone all day and you've got 13%, then you could find yourself in hot water." For the same redundancy reason he wears an automatic mechanical watch for daily use—reserving a smartwatch only for workouts—to avoid single points of failure that depend on battery life.
SAS operator on the UK gun ban as a warning about ceding rights
Drawing on the UK gun ban, the guest argues gun ownership there was a niche sporting activity that was easily eliminated, and warns Americans: "Remember that any rights that you give up, you're never going to get them back."
From the Wider Web
Stories the Scout surfaced from the open web — sources beyond the curated roster. Worth a look, but vetted by the AI rather than hand-picked.
Saudi Pipeline Shutdown Reignites Marine War-Risk Fears, With $120 Oil in View
Saudi Arabia has shut down the pipeline that had kept its oil flowing since the Strait of Hormuz effectively closed earlier this year, according to Insurance Business. The report frames the closure as the key development for marine and energy insurance markets, alongside Houthi claims over a crucial Red Sea chokepoint and predictions of $120 oil.
Context: For insurers, the story is less the headline oil price than the repricing of war-risk premiums on any hull transiting the region — a cost that flows through to shipping rates and, eventually, energy prices.
Treasury Sanctions 27 Iranian Airlines in Sweeping Aviation Crackdown
The US Treasury on Tuesday imposed fresh Iran-related sanctions on one individual and 35 entities, including 27 Iranian airlines, expanding pressure on Tehran's aviation sector and networks supporting the already-sanctioned carrier Mahan Air. OFAC added carriers including Iran Aseman Airlines, Iran Airtour, Kish Air, Qeshm Air, Taban Airlines, Zagros Airlines, and Karun Airlines to its Specially Designated Nationals list, alongside newer carriers such as Air Shiraz, Ava Airlines, and Mehr Airways.
Context: The designations land alongside Treasury's 'Operation Economic Outcast,' whose September 8 deadline forced banks and companies to unwind Iran-related activity — part of a coordinated tightening of the sanctions net around Iranian transport and oil-financing networks.
Louisiana Federal Court Adds to Growing Litigation-Funding Disclosure Patchwork
The U.S. District Court for the Western District of Louisiana issued a standing order requiring parties to disclose the name and address of any third-party litigation funders backing their cases, per Reuters. The order stops short of requiring full disclosure of funding agreements, but parties must provide a brief description and state whether a funder has certain rights. Reuters notes this adds to a small but growing patchwork of similar transparency rules across the federal courts.
Context: Third-party litigation finance — where funders bankroll suits for a cut of any settlement or judgment — has drawn increasing scrutiny from defense-side interests and the Chamber of Commerce, making disclosure mandates a live front in the broader fight over the industry's opacity.
Legal News
A thin day for the mass torts and funding beat. The most relevant development is procedural: the Supreme Court intervened on the Missouri congressional map and took up a separation-of-powers question on acting U.S. attorney appointments. Nothing today moves the needle directly on litigation funding or mass tort valuations.
SCOTUS blocks Missouri's new congressional map for now
The Supreme Court granted a request to prevent Missouri from using a congressional map expected to favor Republicans, per SCOTUSblog.
https://www.scotusblog.com/2026/09/supreme-court-grants-request-to-prevent-missouri-from-using-congressional-map-expected-to-aid-re/Administration asks SCOTUS to bless acting U.S. attorney appointments
The Trump administration has asked the justices to decide the validity of its appointments of acting U.S. attorneys, per SCOTUSblog's Amy Howe. The dispute turns on the scope of the executive's temporary appointment authority.
https://www.scotusblog.com/2026/09/trump-administration-asks-justices-to-decide-on-its-appointments-of-acting-us-attorneys/UPC hands Abbott another CGM injunction against SiBio
The Unified Patent Court's Milan local division, under presiding judge Pierluigi Perrotti, granted Abbott a preliminary injunction barring SiBio from offering, importing, or storing its GS3-R glucose-monitoring system across the UPC's jurisdiction, and ordered SiBio to disclose sales information to Abbott.
Context: Continued evidence the UPC is emerging as an aggressive, plaintiff-friendly venue for cross-border injunctive relief in the medtech patent wars — a jurisdiction worth tracking as U.S. patent filings redistribute post-Albright. From our wiki →
https://www.juve-patent.com/cases/abbot-lands-another-pi-for-glucose-monitoring-devices/Mass Tort Intelligence
A modest signal day. A multi-lot E. coli O145 expansion in Walmart's private-label frozen fruit supply chain is the item worth watching for the pathogen tort angle, while a small cluster of FedEx driver misclassification suits reflects an established litigation pattern rather than a new frontier.
Walmart Great Value Frozen Fruit E. coli O145 Recall Expands to Triple Berry Blend
Frutas y Hortalizas del Sur has expanded its Walmart recall to add a lot of Great Value Frozen Organic Triple Berry Blend over possible E. coli O145 contamination, per Top Class Actions.
Context: Signal Strength: 3/10. E. coli O145 is a Shiga toxin-producing (STEC) serotype capable of causing hemolytic uremic syndrome, particularly in children and the elderly — the injury profile that makes foodborne outbreaks litigable. But recalls of this kind are common and, absent a confirmed illness cluster and epidemiological traceback linking specific plaintiffs to these lots, they rarely mature into mass torts; the plaintiff-side firms in this space (Marler Clark being the dominant one) work single-outbreak cohorts, not MDLs. Plaintiff Profile: consumers who ate the recalled lots and developed STEC infection or HUS. Next Step: monitor CDC/FDA outbreak pages for a linked illness count and confirmed hospitalizations before expending intake resources — the recall alone is not yet actionable. I cannot confirm from this source whether any illnesses have been reported.
https://topclassactions.com/lawsuit-settlements/lawsuit-news/walmart-recall-of-great-value-frozen-fruit-expands-to-include-organic-triple-berry-blend/Science & Non-AI Technology
Today's most commercially significant science is in biotech: the first evidence that human heart muscle regenerates after a heart attack, and an off-the-shelf cord-blood cell therapy that attacked solid tumors in mice — both pointing toward large addressable markets in cardiology and oncology. A deep molecular autopsy of a 117-year-old adds to the emerging science of healthy longevity, and a titanium-dioxide-free white material offers a play on tightening PFAS and pigment regulation.
Human hearts can regrow muscle after a heart attack, study finds
Researchers report the first demonstration that human heart muscle cells can regrow following a heart attack, overturning the long-held view that damaged cardiac tissue is incapable of regeneration. The team now hopes to find ways to amplify this natural repair process as a new route to treating heart failure.
Context: Heart failure affects tens of millions and is a chronic, high-cost condition with no cure — patients cycle through drugs, devices, and eventually transplants. A pharmacological way to coax the heart to rebuild itself would be one of the largest markets in medicine; watch for which biotechs can move from 'cells can regrow' to 'here's the molecule that makes them.'
https://www.sciencedaily.com/releases/2026/09/260909231754.htmUCLA turns cord blood into off-the-shelf cancer-killing T cells
UCLA researchers engineered ready-made T cells from cord blood stem cells that attack solid tumors using two separate detection systems. In mice, a single dose controlled cancer and extended survival while avoiding graft-versus-host disease, a dangerous complication of donor-derived T-cell therapies.
Context: The commercial prize here is 'off-the-shelf.' Today's approved CAR-T therapies are custom-made from each patient's own cells — a logistical nightmare that keeps prices near $500k and largely limits them to blood cancers. A manufacturable, allogeneic product that works on solid tumors (the vast majority of cancers) would rewrite the economics of cell therapy. Still early — mouse data — but the dual-targeting and lack of GvHD are the two problems the field most needs solved.
https://www.sciencedaily.com/releases/2026/09/260909231723.htmThe molecular autopsy of a 117-year-old: aging and protection at once
A detailed study of Maria Branyas, who lived past 117, found she carried signs of advanced aging alongside powerful biological protection — low inflammation, beneficial gut bacteria, protective genetic features, and a biological age younger than her chronological age. Because she avoided major disease, the study offered a rare window into aging itself rather than the illnesses that usually accompany it.
Context: The longevity-biotech thesis rests on the idea that aging is a distinct, drug-able process separate from the diseases it causes. Case studies like this one supply the biomarkers — inflammation, microbiome, epigenetic age — that startups and diagnostics firms are racing to commercialize as intervention targets and clinical endpoints.
https://www.sciencedaily.com/releases/2026/09/260909231800.htmA brilliant white material with no titanium dioxide and no PFAS
Researchers created a foam-based material that produces intense whiteness and water repellency without titanium dioxide pigments or PFAS chemicals. Inspired by structures in snow, clouds, and lotus leaves, it uses tiny pores to scatter light and a rough surface to shed water, offering a more sustainable route to white, water-resistant everyday products.
Context: Both ingredients this replaces are under regulatory pressure — titanium dioxide faces scrutiny as a food and cosmetics additive, and PFAS 'forever chemicals' are being phased out across jurisdictions. Any structural (rather than chemical) substitute that hits the same performance is a licensing and materials-supply opportunity in paints, coatings, textiles, and packaging.
https://www.sciencedaily.com/releases/2026/09/260909231712.htmPhysicists find a fundamental 'glitch' in time itself
Unconventional quantum theories may imply that time carries a tiny fundamental uncertainty, setting an ultimate limit on how precisely any clock could measure it. The effect is far beyond current detection but could point to a hidden link between quantum mechanics, gravity, and spacetime.
Context: No near-term commercial angle, but worth filing away: the precision-timekeeping industry (atomic clocks underpin GPS, telecom, and high-frequency trading infrastructure) is built on the assumption that clocks can get arbitrarily better. A theoretical ceiling on timekeeping, even a distant one, reframes how far that roadmap can run.
https://www.sciencedaily.com/releases/2026/09/260910225251.htmClassifieds
A pure Bring a Trailer day — no real estate or businesses crossed the desk worth your time, but the auction block has a couple of genuinely underappreciated grand tourers. The picks below are where the value is: manual V12 and flat-six analog machines that aren't making any more of. Everything else was average metal at average money.

Gated-Shifter 1995 Ferrari 456 GT, 6-Speed Manual
A 1995 Ferrari 456 GT in Rosso Monza over Beige, powered by a 5.5-liter V12 mated to a six-speed manual transaxle with the gated shifter. It shows 50k miles, was first registered in California, spent 2006–2020 with its second owner in Florida, and is offered with a clean Carfax and a clean Arizona title in the owner's LLC name.
Context: The manual 456 is one of the last front-engined Ferrari V12 GTs you can still buy for a fraction of what a comparable-era 550 commands — and the gated six-speed is the configuration collectors are now chasing hardest. A genuinely usable, four-seat, V12 Ferrari with a metal gate is a rare thing at any price; documented 50k-mile examples with clean paperwork won't stay cheap.
https://bringatrailer.com/listing/1995-ferrari-456-gt-48/
1988 BMW M3 Evolution II, #151 of ~500
This 1988 BMW M3 Evolution II is number 151 of a reported 500 built for the model year, imported to the US from Japan in 2025. It runs the 2.3-liter S14B23 four with a five-speed manual and limited-slip diff, finished in Misano Red over gray leather and Überkaro cloth on 16" basketweaves with Bilstein dampers. Offered with a clean Carfax and clean New York title.
Context: The Evo II is the homologation-special E30 M3 — genuinely rare (roughly 500 built), and the value gap between a standard E30 M3 and a documented Evo II has only widened as the halo cars appreciate. Misano Red with the correct Überkaro cloth is exactly the spec buyers want.
https://bringatrailer.com/listing/1988-bmw-m3-evolution-ii-6/
2007 Porsche Cayman S, 6-Speed, One-State Car
An Atlas Gray-over-Stone Gray 987.1 Cayman S with the six-speed manual, California-registered from new, showing 84k miles. Factory options include 19" Carrera Sport wheels, Sport Chrono, bi-xenon headlights, and Bose surround sound; it comes with the window sticker, owner's manual, and a clean Carfax on dealer consignment.
Context: The 987.1 Cayman S with a manual is arguably the best-driving Porsche-per-dollar on the market right now — mid-engine balance, a naturally aspirated flat-six, and a metal shifter, still trading well below air-cooled and 981 money. This is the last window to buy one as a used sports car before it prices in as a modern classic.
https://bringatrailer.com/listing/2007-porsche-cayman-s-157/
40k-Mile 2011 Jaguar XJL, 5.0 V8, No Reserve
A 2011 Jaguar XJL in Ultimate Black over London Tan, powered by the naturally aspirated 5.0-liter V8, showing 40k miles. Equipment includes a dual-pane sunroof, heated/ventilated front and heated rear seats, Meridian audio, and rear tray tables; recent work covered tires, drivetrain mounts, and battery. Offered at no reserve.
Context: This is maximum car for minimum money — a low-mileage, naturally aspirated V12-successor Jag flagship that depreciated off a cliff and now sells for used-Camry money. The no-reserve format on a 40k-mile example is where the genuine steal lives, provided you budget for the running costs a big-cat luxury sedan demands.
https://bringatrailer.com/listing/2011-jaguar-xjl-12-2/The Ideator
Today's strongest thread: AI agents have crossed from hypothetical risk into documented weapons-development and hacking incidents, converting alignment debates into concrete liability, procurement, and insurance questions right as compute capital floods in and Middle East war risk spikes marine and energy premiums.
Business Idea: An Agentic-AI Liability & Audit Underwriting Firm
Structure a specialized managing general agency and forensic-audit shop that underwrites 'agentic AI liability' — the exact gap exposed today when OpenAI-linked agents hacked a code-host and Anthropic disclosed weapons-development misuse. Pair a legal-and-security team that certifies an enterprise's agent guardrails, logging, and kill-switches against emerging standards with a delegated-authority binder from a reinsurer to sell coverage for third-party harm, breach, and IP claims caused by autonomous agents. Every company deploying the new OpenAI Agents API, SWE-2, or DeepSeek models will soon face procurement demands for proof of controls and transferable risk; you sit in the middle collecting audit fees plus underwriting margin. The marine-and-energy war-risk repricing (Hormuz, Bab Al-Mandab, $120 oil in view) proves buyers will pay steeply for coverage the instant a chokepoint risk becomes real — agentic AI is that chokepoint for the software economy, and it just became real today.