Jeffrey Miller @millermusing

Only times in history where the Semiconductor Index gained more than 230% in a 14-month span: 1) December 1998 - February 2000 2) April 2025 - Today That's the entire list.

Clearly, Fable is doing a lot of work, and unleashing a ton of agents. To review a short technical note, it released 31 agents, coded simulations to verify my results, did "adversarial reviews". Eventually, it only made the assumptions slightly more rigorous. It is all good. For a four-page technical note+a little code, though, it consumed all my Pro session tokens, *plus* $17 worth of credits. It is ridiculously expensive. I have 20-page reports that are way more complex than this. I can see how Anthropic has entered the phase of market-clearing prices, yield management, and pre-IPO. I recall Boris Cherny saying in a podcast, "run Opus [4.6], not Sonnet. It's worth it". I feel comfortable saying that running your top-shelf model is *not* worth it anymore. Decreasing returns, on most tasks. Like in the real world, some people can be real smart, but real expensive.

Private credit fund manager explaining why Bloomberg shows 75, but their marks are 98

These people have been waiting an hour to get into Madison Square Garden. Here is a first person account: “They routed everyone down to 34th and 8th. Police barracade. Through metal detectors to a point about 200 feet from MSG. Then they came and apologized and said they were confused and instead we needed to exit the barracade and go to 34th and 7th. We walked there and then a different set of police told us to go to 32nd and 6th. There are hundreds, if not a thousand people, wandering around in giant herds not knowing where to go or how to get in. The police keep saying they don’t know the protocol.”

The top producer of Fortune 500 CEOs isn't Harvard or UPenn. It’s no degree at all.

A look back at the stocks that traded above 10x sales at the dot-com peak. What happened next: – Cisco: ~25x sales, P/E above 200. Crashed -90%. Finally broke its 2000 peak in December 2025. 25 years and 8 months later. – Intel: ~13x sales. Crashed -82%. Finally broke its 2000 peak in May 2026. Almost exactly 26 years later. – Microsoft: ~25x sales. Crashed -65%. Took 16 years and 8 months to make a new high (October 2016). – Qualcomm: ~30x sales. Crashed -88%. Took roughly 20 years to break even. – Sun Microsystems: ~10x sales. Crashed -97%. Acquired by Oracle in 2009. – JDSU: ~50x sales. Crashed -99%. Broken into pieces. – Yahoo: ~50x sales. Crashed -97%. Sold to Verizon for a fraction. – Lucent: ~10x sales. Crashed -99%. Eventually absorbed by Nokia. – Nortel: ~15x sales. Bankrupt in 2009. Then there's the famous mega survivor. Amazon traded at ~30x sales at the peak. It still crashed -97%. The investor who bought at the top held through a 97% drawdown before eventually making money roughly a decade later. The lesson isn't that every 10x sales stock ends in zero. It's that even the eventual winners crash 90%+ first, and break even only after a generation. Cisco. Intel. Microsoft. Amazon. The four greatest tech survivors of the dot-com era. Average time to break even on price alone: roughly 19 years. Inflation-adjusted, the math is uglier. You have to be very right, very early, and willing to hold through unimaginable pain. Most people aren't.

Thierry from arvy 🇨🇭 @ThierryBorgeat

2 weeks ago

51% of the S&P 500's market cap is in stocks trading above 10x sales. Half the index. In 2002, after Sun Microsystems crashed 90%, CEO Scott McNealy famously said this about his own stock at 10x sales: "At 10x revenues, to give you a 10-year payback, I have to pay you 100% of

134 488 2K 785K 750

I say this all the time with writing: Just because you took the time to write it, doesn't mean anyone owes you their time to read it. Effort and output are irrelevant. The only thing that matters is the value it provides someone else. The same is true for technology.

Jen Zhu @jenzhuscott

2 weeks ago

Massive output uptick due to agentic AI. Complete flat adoption.

470 980 7K 2.5M 2K

Today: Juan Soto signing a $765 million contract and won’t run out ground balls 70s: Pete Rose signing a $100,000 contract and running out his fucking walks like Usain Bolt because he had money riding on that shit Advantage: 70s

Easily still the funniest clip from his second term for me

The SE corner of Montana is closer to Texas than it is to the NW corner of Montana.

Disney CEO seeing “Obsession” and “The Mandalorian & Grogu” both tracking to $330,000,000 even though the indie horror cost $750,000 while the Star Wars reboot cost $165,000,000 (with another $200,000,000 more marketing)

Film Updates @FilmUpdates

3 weeks ago

‘OBSESSION’ has grossed around $148M worldwide. Now eyeing a $280-330M finish on a $750k budget. One of the most profitable movies of all time.

356 3K 53K 16.6M 3K

Nicolai Tangen has interviewed all the most important CEOs and investors in the world (that's what $2T AUM buy you: lots of access). He has also interviewed a fair number of HF managers. Here are my favorites. Even so, it's almost 4 hrs. But I recommend actually listening, over the gemini summary. Chris Hohn: nbim.no/en/news-and-in… Paul Singer: nbim.no/en/news-and-in… Marc Rowan (not a HF, still interesting): nbim.no/en/news-and-in… Stan Druckenmiller: nbim.no/en/news-and-in… Ken Griffin: nbim.no/en/news-and-in…

Heading into the weekend with the S&P 500 riding a 9-week winning streak for the first time since 2023

Jamie Dimon, complaining about the Clarity Act and Coinbase CEO Brian Armstrong this AM: “He’s spending hundreds of millions of dollars in Washington in this thing.” Maria: “He said he’s representing the whole —” Dimon: “He’s full of shit.” Maria: “…well.”

I just can’t see how firms decided to reward employees for maximizing use of inputs rather than maximizing output.

Techmeme @Techmeme

3 weeks ago

Sources: Amazon has shut down an internal leaderboard that tracked employees' use of AI tools after workers tried to boost their scores with needless tasks (@rafeuddin_ / Financial Times) (Visit Techmeme dot com for the link and full context!)

48 110 1K 1.8M 292

Read the second sentence too(about model stability). And the third (about sources of instability). But a fourth should be added—on describing the target phenomenon sufficiently well, out of sample. It seems that a lot of economic theory does not concern itself too much with data in sample, either. An excuse to extract some “stylized facts”?

Fatih Kansoy @kansoy

6 years ago

Have you seen a better inroductory? Solow (1956) : 𝑨𝒍𝒍 𝒕𝒉𝒆𝒐𝒓𝒚 𝒅𝒆𝒑𝒆𝒏𝒅𝒔 𝒐𝒏 𝒂𝒔𝒔𝒖𝒎𝒑𝒕𝒊𝒐𝒏𝒔 𝒘𝒉𝒊𝒄𝒉 𝒂𝒓𝒆 𝒏𝒐𝒕 𝒒𝒖𝒊𝒕𝒆 𝒕𝒓𝒖𝒆. 𝑻𝒉𝒂𝒕 𝒊𝒔 𝒘𝒉𝒂𝒕 𝒎𝒂𝒌𝒆𝒔 𝒊𝒕 𝒕𝒉𝒆𝒐𝒓𝒚. Any other example? #EconTwitter

12 68 490 0 155

"It was the time sheets that made people look into him...." I wish people could see what the Army finance office will do when a soldier gets overpaid $20.00. Then I would like to show them the oversight that our Travel System (DTS) has for a soldier to get paid for spending one night in a hotel. Next I would show them the weekly Battalion meeting where every leader can see the names of every soldier in their unit who owes money for travel cards/DTS/paychecks. Finally I will show them the withdrawal and clearing process for the various OPFUNDS that are used during overseas operations. The army will crush a soldier for owing a single penny and this dude signs for $40 Million in Gold bars and puts them in his house....and the ONLY reason they found them was because he cheated on his time sheets. We have soldiers and their families living in trailers and we will go through their finances with a fine tooth comb. We have Illegals, Child care center operators, and senior executive band people bilking the Government for MILLIONS/BILLIONS and people seem to discover it by "chance". Every last fraudster caught with over a million dollars should be dropped from a space shuttle and allowed to re-enter the earth's atmosphere so we can watch them turn into a ball of flame for our entertainment.

BNO News @BNONews

3 weeks ago

CIA official arrested for taking home $40 million worth of gold and $2 million in cash after filling out fraudulent time sheets - NYT

167 677 3K 769K 248

I didn’t realize computer inflation had gotten this bad:

keyboard_arrow_left Previous keyboard_arrow_right Next

In 1918, the Bolsheviks instituted the nationalization of urban housing in Moscow & Petrograd. Large apartment buildings were expropriated & converted into kommunalki (communal apartments). Owners were dispossessed, usually being assigned a small room in the building they formerly owned. Families who'd owned apartments with multiple bedrooms were accused of "bourgeois excess"; their "underuse" of property was likened to theft from the "proletariat." Their homes were seized & parceled out to poorer working-class families. Just in case you're curious as to where this could be heading.

Brecca Stoll @breccastoll

3 weeks ago

NOW: Mamdani says his admin will transfer ownership from bad landlords to non-profits. “For buildings that have suffered chronic neglect, we will work to transfer ownership to responsible stewards. Stewards that include community land trusts, non-profits, or even the tenants

6K 3K 12K 4.4M 1K