Cloudwood
banner
heaventree.bsky.social
Cloudwood
@heaventree.bsky.social
Just another stochastic process... Creative entrepreneur and investor. NYC.

A Trump in pursuit of the Nobel Peace Prize could be, on some things, an improved Trump.
Whatever Ukraine has on Trump, good work!
September 24, 2025 at 1:17 PM
Someone said even if undoc immigrants commit less crimes, they’re still adding to the pool of crimes and theirs are crimes we can avoid by kicking them out - but math says any one person is more safe w more immigrants. The proof:
July 7, 2025 at 11:37 AM
Is Musk announcing his new America Party like Darth Vader turning on the Emperor? Will they both fall into the reactor?
July 5, 2025 at 8:45 PM
Popped in to Barnes and Noble during a walk after dinner. “Irresistible” The NYT declares apparently about one well known novel. But I am resisting it.
June 17, 2025 at 11:20 PM
Having an edge on a longterm view really helps navigate risk management and risk taking on short term trades. $Gold
May 29, 2025 at 1:22 PM
Interesting to think about someone's narcissism level relative to their ambition to accomplish things. As we usually overestimate our own importance, someone very humble may have even humbler ambitions thus still be relatively narcissistic as a ratio of self-regard/ambition!
May 28, 2025 at 7:11 PM
“Is math invented or discovered?” is one of the reality-altering questions, like “What is consciousness?” & “Is there a spoon?” My answer is the universe is systematic & logical so a systematic & logical descriptive system appears inherent but is only (partially) consistent w its properties.
May 22, 2025 at 3:49 PM
What if Navarro is crudely right and reacting to Trump's ignorance and infuriating lies gets in the way of analyzing the situ. OK, ok, I don't think so, even if some of Navarro's ideas have vague merit. But the disaster may not play out in the way expected.
May 1, 2025 at 12:46 PM
All the caveats* but let's face it the complaining/(quality fin analysis) ratio on Bluesky is quite low while still exceptionally high on the non-poseur remnant of fintwit (finX) *Elon is evil, Trump is evil, Maga are morons, etc.
April 11, 2025 at 12:27 PM
[Friday 11 am]

Don: You’re interrupting my Big Mac
Scott: You watching yields?
D: No, I have President things to do.
Scott: They’re blowing out again. Pull it ALL back.
D: Admit I made a mistake? Never. I prefer Armageddon.
S: Will you let me still be Treas Sec?
D: Sure.
S: OK. Worth it.
April 10, 2025 at 5:01 PM
Quantery: Many times the fidelity of the model matters more than the accuracy. EG, sometimes all you want from a model is to transform the data out of the space used by every Tom, Dick, and Harriet, so you can run simple analysis in an uncrowded field. For that, you most need fidelity.
April 6, 2025 at 3:34 PM
The personality tuning of ChatGPT to each user impacts not only how info is presented but the info itself. The "truth" is to some degree tuned to you. This is self-evident if you see how it differently answers similar questions by diff users.
April 3, 2025 at 3:46 PM
Yesterday even as it declined price action (spx futures, es) was still running ahead of my model so it looked like some upside was on the way. Today running less ahead, more sideways in pre-market.
April 1, 2025 at 1:26 PM
ES (SPX Futures) outperforming my model by a few dozen points even as it continues reaching for downside. Which is either good for a tradeable long or bad.
March 31, 2025 at 1:48 PM
McKinley was a great acquirer of territories and he’s Trumps model President for 2.0. No wonder he’s blabbing on about getting Canada and Greenland-to get us out of this mess, need to come up w something non destructive for him to acquire pronto.
March 29, 2025 at 4:10 PM
We only know about the possible existence of dark matter & dark energy because of interaction w gravity. Interesting to wonder: 1. They may otherwise interact w each other & their own fundamental forces, constituting a parallel dimension w same space/time dimensions as us or diff ones! 2. What if...
March 25, 2025 at 5:37 PM
Reposted by Cloudwood
The people who did this (despite a judge directly instructing them not to) need to go to jail.
“One young man sobbed when a guard pushed him to the floor. He said, ‘I’m not a gang member. I’m gay. I’m a barber.’ I believed him. *** He “began to whimper,” as his head was roughly shaved, “folding his hands in prayer as his hair fell.” He “asked for his mother & cried as he was slapped again.”
What the Venezuelans Deported to El Salvador Experienced
Exclusive photos of the arrival of Venezuelan detainees deported from the U.S.
time.com
March 22, 2025 at 10:13 PM
I honestly don't know what can break through the Trumpy brains. Facts, logic, anger, sarcasm, etc. don't go in; they "just impact on the surface." Maybe the only thing w a shot is great storytelling. Someone who can tell a diff story than Trump and be loved for it.
March 17, 2025 at 5:01 PM
March 12, 2025 at 2:04 PM
Huh. So Americans are willing to bend over and take it from Trump/Musk but turns out Canadians aren't...
March 10, 2025 at 4:59 PM
Rallies are born in fire and die in ice, and the ivols are opposite: die in fire, born in ice. Yes, implied can be sticky once it gets going but it doesn't stick to the peaks, it sticks to higher valleys where a new rally can be born again (to mix up metaphors more than once) . Exceptions apply ofc.
February 6, 2025 at 7:19 PM
It's fascinating to watch VRP plunge immediately with the tariff pauses and then stay down tho the ivols remain elevated (tho ofc down from recent levels)... This is also why some form of skew is interesting to keep an eye on.
February 4, 2025 at 1:16 PM
Repetition and habits create the illusion that we're a singular entity, with a stable identity, moving through time.
January 31, 2025 at 6:05 PM
Reposted by Cloudwood
Researchers are finding that the architecture of LLMs has inherent limitations. But this is not the end of chatbots. @anilananth.bsky.social reports: www.quantamagazine.org/chatbot-soft...
Chatbot Software Begins to Face Fundamental Limitations | Quanta Magazine
Recent results show that large language models struggle with compositional tasks, suggesting a hard limit to their abilities.
www.quantamagazine.org
January 31, 2025 at 3:39 PM
I'm not saying I can compute Z-Scores in my head but having spent hundreds or thousands of hours staring at terminals with Z-Scores and abs values, I pretty much can.
January 31, 2025 at 3:11 PM