Aggregator

“There was shock, definitely,” said the source, who said staffers were buzzing about the crime before the paper finally covered it on Monday.

While Eklund came into camp as an underdog to make the roster, his inclusion was the obvious move by the end of it.

To Build Golden Fleet Frigate Fast, Two Shipyards Are Needed

Authored by Brent Sadler & Chase Ouellette via RealClearDefense,

In 1798, while at war with France, an exasperated British Lord Nelson exclaimed, "Were I to die at this moment, 'want of frigates!' would be stamped on my heart." It's a perennial problem: Frigates formed the backbone of navies in Nelson's time and do so today as well, and there seldom seems to be enough. The U.S. Navy's ambitious Golden Fleet shipbuilding plan aims to fill a frigate gap but will require shipyard capacity that today the nation lacks.

The Navy's 2026 shipbuilding plan plans a fleet of 66 frigates. By 2031, two Constellation-class frigates are to be completed at the Marinette, Wisc. shipyard and a modified U.S. Coast Guard National Security Cutter at the Ingalls, Miss. shipyard. But to meet the Navy's goal, a sustained delivery of more than three frigates per year will be needed by 2037. Such a production schedule is unlikely with one shipyard, and the plan all along was to have two frigate-producing shipyards.

In December 2020, just months after the initial frigate contract, the Navy reported to Congress its intention to stand up a second shipyard to expand production to three ships by 2023 and four ships from 2025. In its 2025 shipbuilding plan, the Navy laid out a proposal to expand frigate production to a second shipyard by 2030.

The Constellation frigate program and a repurposed National Security Cutter will fill the need for a multi-session surface combatant - anti-submarines warfare and limited air defense. Nearly seven years after the program's inception, zero frigates have been built.

Complicating frigate production has been persistent design uncertainty; one adjustments to original FREMM design for Constellation, and uncertainty over how the Navy will turn the National Security Cutter into a frigate. For Constellation, the intent was to build a proven allied design with little to no changes, but modifications resulted in the ship retaining only 15% of its original design. Today the design is over 95% complete, but cost and displacement have ballooned to approximate a destroyer.

The axe fell in November 2025. Then-Secretary of the Navy Phelan cancelled Constellation because it was 80% the cost of a destroyer while providing only 60% of their capability. The thinking then was to substitute this with a repurposed cutter.

The previous delivery of 10 National Security Cutters seemed to validate a pivot towards HII's Ingalls shipyard. That said, modifications will be needed before its design is Navy-ready: modest air-defense systems, torpedo magazines, added fuel storage, additional crew accommodation. But it has been years since the last National Security Cutter was built. As such, additional workforce at Ingalls will be needed and supply chains restarted, slowing renewed production of a reproposed cutter.

On the other hand, the Constellation program does benefit from Fincantieri investments of $800 million and workforce expansion of 850 that since cancellation has begun to be let go. Reversing this is still possible with a finalized design that could allow for initial delivery quicker than the Ingalls repurposed cutter. But both options today face capacity constraints.

In an August 13 memo, President Trump announced a framework to increase domestic shipbuilding capacity using the "Finland Model" that seeds domestic shipbuilding growth through initial overseas orders. This works by making conditional ship orders that see the initial hulls built at foreign shipyards while concurrently developing American shipyards to construct later ships. This model is inspired by from the 2024 ICE Pact icebreaker construction effort with Canada's Davie Shipbuilding with shipyards in Finland and Texas.

Adapting the Finland Model to frigate production offers way ahead that could leverage shipbuilding capacity in Japan and South Korea - the world's second- and third largest shipbuilders, respectively. Both nations build capable frigates that have emerged as potential options for meeting urgent American frigate needs: South Korea's Chungnam-class and Japan's Mogami-class. Already, Australia has inked their own Finland Model 11-ship deal with Japan, with the first three Mogami-class vessels being constructed in Japan and follow-on production in Australia.

Thoughts of partnering with Turkey on frigate production, however, would be hindered because it hasn't made industrial investment commitments like Japan and South Korea have. Japan has committed a portion of the $550 billion of U.S. investments to modernizing America's shipbuilding industry. South Korea's Hanwha Group has followed its $100 million purchase of its Philadelphia shipyard in 2024 with a $5 billion infrastructure plan to further develop the facility, and Seoul has committed $150 billion to the Make American Shipbuilding Great Again.

While the Navy may opt to acquire foreign frigates under the Finland Model, further leveraging existing American shipbuilding capacity at Ingalls and Marinette shouldn't be overlooked to further accelerate domestic frigate construction. The Navy needs frigates, and unconventional approaches required to be clear, but only a plan that seeks added domestic shipbuilding capacity will be sustainable.

Brent Sadler is a 26-year naval veteran and current senior research fellow for naval warfare and advanced technology at the Heritage Foundation. Chase Ouellette is a member of Heritage's Young Leaders Program.

Tyler Durden Wed, 09/30/2026 - 19:30

I am a lifelong Jewish Democrat and civil libertarian. And I have left the Democratic Party. So should you. 

Jameis Winston feels like he can’t win even when the Giants are victorious. 

1. Old man yells at robot by Mike Solana, Editor in chief  Hot on the tail of former abundance proponent Ezra Klein’s proposed hamstringing of the model labs in the pages of the NYT, America’s foremost old communist, Bernie Sanders, unleashed his “Ban Artificial Superintelligence Act,” prompting immediate endorsement from Doom Prophet Eliezer Yudkowsky, with...

Gavin Newsom took an oath to "support and defend" the US Constitution when he became Governor of California. But you wouldn't know it from his habit of signing unconstitutional bills into law.

Prosecutors initially claimed the Jane Doe never accused anybody of a crime -- but newly obtained police interview transcripts show otherwise.

McBride is one of the best values in the league. But assuming he has another productive season, he is in line for a dramatic pay raise. 

Dr Ning Xu denies he broke the rules of Nikon's annual Small World In Motion contest.

EDC head Anthony Shorris, entrusted with the indispensable, herculean task of promoting citywide business growth, sounds like he’s swallowed Zohran Mandani’s socialistic Kool-Aid.

A power outage left the San Francisco Ferry Building and parts of the surrounding waterfront without electricity Wednesday morning.

Benchmaxxed: Google's New Gemini 4 Aces SAT, Struggles With Actual Job, "Skeptical Employees" Admit

After a year in which Google's AI roadmap resembled a Waymo stuck in a roundabout, the search giant on Wednesday finally unveiled Gemini 4 "Argon", its long-awaited flagship model. The market cheered, briefly: according to Goldman's closing equities color, GOOG traded +2% after hours on the "Argon" announcement. 

Then Bloomberg reported that some of the people who have actually use the thing - as in Google's own engineers - aren't nearly as impressed as the leaderboard. The stock promptly faded. 

According to Bloomberg's Julia Love and Davey Alba, Gemini 4 "performed well on benchmarks widely used to gauge model efficacy" but "does less well when employees actually put it to work," particularly on coding. One insider said the model "isn't particularly adept at front-end design" - the part of software that decides how apps and websites look and feel. Which is a bit awkward for a company whose entire business is, well, things you look at on a screen.

Google, naturally, disagrees. It told Bloomberg it would be "inaccurate" to say Gemini 4 is underperforming in coding, and pointed back to last week's remarks by DeepMind boss Koray Kavukcuoglu, who said "it's a certainty that we are always gonna be at the frontier." Another Google employee said there is "large consensus" internally that the model is frontier-class. Which is the kind of thing one tends to say when there isn't.

The industry has a word for this. "Benchmaxxing" is when engineers optimize a model to crush standardized tests rather than to do useful work - the AI equivalent of the kid who memorizes every SAT prep book, posts his 1600 on LinkedIn, and then can't do his own laundry. Two people familiar with Gemini 4 told Bloomberg the model appears affected by exactly this.

Surge AI founder Edwin Chen put it more bluntly, calling benchmark-chasing "an incredibly pernicious problem" and comparing it to bragging about your kid's SAT score. We'd add that a generation of Silicon Valley's finest minds has now spent billions teaching machines to do what they themselves did in high school: optimize for the test, then act surprised when the real world grades on a curve.

The irony is that Google's own researchers have documented how quickly optimization turns into gaming. As we reported on September 3, a DeepMind experiment found that when math problems got hard, 9% of AI agents outright cheated and another 5% cheated "after hesitation", gaming a shared knowledge base that rewarded successful submissions. Turns out teaching to the test works on silicon too.

A $400 million detour

Gemini 4 is also the model Google shipped instead of the one it promised. At I/O in May, Google pledged Gemini 3.5 Pro for June; the date slipped (Bloomberg first reported the delay on July 16, citing tech that "fell short of internal goals") and the project was eventually abandoned altogether. Bloomberg Intelligence's Mandeep Singh estimates a frontier training run can cost as much as $400 million - before paying the researchers, many of whom have since left.

And leave they did. We noted in June that Google was losing more Gemini researchers to Anthropic, following the earlier departures of Nobel laureate John Jumper and transformer co-inventor Noam Shazeer. Then in August came the big one: Jeff Dean exited after 27 years to launch his own startup, taking several senior researchers with him and knocking 5% off Alphabet stock. Demis Hassabis subsequently kicked himself upstairs to chairman, handing day-to-day DeepMind operations to Kavukcuoglu - who now gets to defend Argon's front-end skills to Bloomberg.

Why this matters: the ROIC math doesn't grade on benchmarks

This would be a nerd-bro squabble over leaderboard screenshots if it weren't for the money involved. Gemini underpins nearly everything Google sells - Search, Maps, Gmail, Chrome, each with over a billion users - and Alphabet is one of the hyperscalers footing the largest capex bill in corporate history.

Per Goldman's Ryan Hammond (full note available to pro subs), US hyperscalers are on track to spend roughly $800 billion in 2026; consensus expects $1.1 trillion in 2027, while Goldman's house view is even higher at $1.2 trillion and $1.4 trillion for 2027 and 2028. Hammond estimates hyperscalers need about $300 billion of annual AI revenue just to break even on 2026-27 spending, and that end users may ultimately need to spend around $1 trillion a year on AI applications for everyone in the stack to earn decent returns.

Separately, Goldman's Eric Sheridan, whose ROIC framework we flagged last week, calculated that assuming a 15% ROIC target and ~$42 billion of capex per gigawatt, the six big US hyperscalers need to generate roughly $1.42 trillion in cumulative revenue during 2028-30 - about $11.6 billion per GW per year - with a range of $908 billion to $1.89 trillion depending on assumptions. As we tweeted at the time, even in Goldman's worst-case scenario where ROIC on capex goes to zero, they'd still need $920 billion a year just to cover depreciation and running costs.

Here's the rub: none of that revenue gets paid in MMLU points. It gets paid by developers and enterprises choosing whose model to build on. And Bloomberg notes Gemini 4 is "a very large model" - and big models are expensive to serve, which means pressure on margins at exactly the moment the capex hurdle is rising. Goldman's TMT desk flagged this week that the industry is already in a price war: OpenAI cut Luna pricing by 80% in July and usage rose roughly tenfold, while "Big Short's" Steve Eisman is openly asking what happens to margins if cheaper Chinese and open-weight models push prices down further. Bringing an expensive, benchmark-optimized heavyweight to a knife fight over token pricing is a bold strategy.

Meanwhile, the competition isn't waiting

The timing couldn't be worse. On the same day Google unveiled Argon, OpenAI was busy at its Developer Day rolling out "Dots" - persistent, always-on agents that live inside ChatGPT, Slack and Teams. Goldman's Sean Johnstone noted OpenAI is "increasingly competing with Microsoft 365, Google Workspace and traditional enterprise software - not simply other AI models." Per Axios, cited by Goldman's desk, OpenAI's ARR is nearing $70 billion, up from ~$40-41 billion as recently as mid-August, with enterprise sales more than doubling since July. OpenAI is now reportedly seeking $30 billion at a $1.4 trillion valuation.

And earlier this month Meta released its new agentic platform, Muse, which promptly took the app charts by storm - so thoroughly that Amazon blocked it - and sparked a 25%+ rally in META. Goldman's Sheridan now frames agentic commerce as one of AI's biggest monetization opportunities, noting that "similar to how Google captured search intent, successful AI platforms may capture shopping intent." Read that sentence again if you're long GOOGL.

To be fair, Gemini 4 has real strengths: insiders say it stands out at multimodal work like extracting metadata from video, on safety and cybersecurity (it reportedly beat OpenAI's Astra on a security benchmark), and it can spit out up to 1 million tokens - roughly 750,000 words - in one go. Whether anyone needs a 750,000-word answer is a separate question; we suspect the answer is "only for benchmarks."

Bottom line: Google has the distribution, the TPUs and the balance sheet. What it doesn't have is much time. Every quarter it spends explaining leaderboard scores, is a quarter that OpenAI, Anthropic and Meta spend convincing developers, businesses and consumers that the future of search and software runs on their platforms. Because the $1.4 trillion question isn't who tops the leaderboard - it's who gets paid.

Tyler Durden Wed, 09/30/2026 - 19:04

Ari Emanuel‘s tell-all book “Roll the Calls” will debut at No. 2 on New York Times best seller list — getting edged out by Princess Diana’s brother, Charles Spencer‘s surprise tome, “Swan Song: Diana, My Sister” at No. 1. Emanuel presumably would’ve been in the top spot, but Diana’s brother played the spoiler with his...

Emanuel presumably would've been in the top spot, but Diana's brother played the spoiler with his tome.

Will farmers want AI tools to help judge when to pick fruit, or is their own intuition enough?

How far will the socialist antisemite wave roll in America? The Los Angeles mayoral race will be a key test.

About a month before his team reported to spring training this year, Dodgers manager Dave Roberts sat on the balcony of an upscale restaurant near his offseason home in San Diego, recounting his journey from franchise scapegoat to championship leader. The more Roberts spoke about his professional evolution, the more it became clear that he...

Justin Herbert looks like a shell of his former self this season

McDonald's is the latest to use AI-powered algorithms to set pricing.