samsachdev.bsky.social
Founder of Ziporama. I like math and Alf.
Ziporama: https://www.ziporama.com/
Alf: https://en.wikipedia.org/wiki/ALF_(TV_series)
229 posts
86 followers
32 following
Active Commenter
comment in response to
post
This has to be understood in relation to the median age. EG, the median age of homebuyer in 2000 was 37. So, how far are 37 and 56 from the median age of all Americans in those respective years? A z-score (or the equivalent) is needed. A measure of how above or below the mean age of all Americans.
comment in response to
post
Now starring in "My Dinner with Dr. No".
comment in response to
post
2/much more valuable than the ability to deliver burritos -- and the equivalent -- to college students.
comment in response to
post
I/Interesting. There is no foundation model for robotics. This means that each robot in every unique setting is made and programmed only for that situation. So, these robots were hand-trained on these paths. And it also means that they were hand-trained on all use cases. The data from this may be
comment in response to
post
Wouldn’t this, what the article argues for, depreciate the USD? Also, implementing capital controls would harm Wall Street, which relies on foreign investments into, say, private equity firms, in addition to depreciating the USD. IMO, for these reasons, this is politically unrealistic.
comment in response to
post
334 million (Americans) times $5,000 is $1.67 trillion. The US gov't spent $6.1 trillion in fiscal year 2023. So, Musk is claiming that DOGE can cut 27% of the federal gov't.
comment in response to
post
Okay, I get it. Have worked out the algebraic solution. Can post it if you want (since you haven't and the other comment also hasn't.)
comment in response to
post
I assume the answer you should be the expected value. That is, if this experiment is repeated as many times as combinatorially possible, what's the largest number of pennies in one distinct pile, right?
comment in response to
post
From my experience, it's related to floating point errors as a system is upgraded. That is, if a previous system used 16 bit floating point and the new system uses 32 bit fp, then it's possible to get rounding errors which result in what you described. Integration testing is designed for this, btw.
comment in response to
post
From what I've read, missing fields in COBOL are stored using ISO 8601:2004 standard. That is, they're given the date May 20, 1875. This isn't unique to COBOL. It's simply a way to get around null errors in databases. What's shocking is that Elon and his team are so incompetent they don't know this.
comment in response to
post
2/If one president can do this much damage, then one president shouldn't have this much power. The system needs reforming.
IMO, this is what I would focus on. How executive power of the presidency needs to be limited and how to do to this.
en.wikipedia.org/wiki/Peter_p...
comment in response to
post
1/What's happening with the fed. govt seems to be a version of Peter Principle: the most incompetent will rise to the highest position. It seems that it was luck that the US hasn't had a president this incompetent and destructive. Trump, Musk, etc., then, are symptoms. The problem is the system.
comment in response to
post
Does the universal approximation theorem apply to linear probes? Specifically, does the space have to be Euclidean, specifically a compact subset X of R^n to R^m and which is also continuous?
en.wikipedia.org/wiki/Univers...
comment in response to
post
A strong counter-argument against a weakening dollar is what happened in 2018 with Trump’s tariffs against China: the renminbi depreciated 10% against the US dollar:
www.chathamhouse.org/2024/11/dona...
comment in response to
post
Doesn't Arrow's impossibility theorem predict that it's (close to) impossible for a third party to become viable in a winner-take-all voting system?
en.wikipedia.org/wiki/Arrow%2...
comment in response to
post
Asteroid collisions are often expressed as joules of energy with respect to the Hiroshima nuclear explosion. This released 10^ 18 joules of energy. The asteroid that caused Chicxulub crater, 66 million years ago, released 10^23 joules of energy. This is 100k times more energy than 10^18, and so on..
comment in response to
post
CNN interviewed a few Muslim voters who voted for Trump and their opinion on Trump's comments on Gaza. Spoiler alert: all wouldn't change their vote.
www.youtube.com/watch?v=6sIa...
comment in response to
post
2/responsible for a 4% decrease in exports over this time, which resulted in a loss of about 150k jobs.
Most of the paper is a defining a casual mechanism to demonstrate that it was the steel tariffs that were the cause.
coxlydia.com/papers/cox_s...
comment in response to
post
1/This chart misses the point. Steel and aluminum are important components in many manufacturing processes. This paper (see link) studied the effects of Bush's steel tariffs in '02 and '03. It found that industries that had a 1% increase in input prices had a 3% decline in exports. This was
comment in response to
post
I do know there have been quite a few recently. This is one from Andrew Levin that found, using BA, in favor of the zootonic hypothesis: www.nber.org/papers/w33428
Am familiar with Michael Weismann's analysis (which is very long and technical). Feel free to share any others.
comment in response to
post
2/evidence of zootonic pathogens causing pandemics. There's also a lot of evidence of lab leaks. The two should simply be compared to each using Bayesian analysis.
This is my only point.
comment in response to
post
1/From a Bayesian prospective, I don't this matters (unless there's a data set of many labs where this is true). My point: a Bayesian analysis allows comparing the likelihood of a zootonic versus a lab leak hypothesis, compared to the likelihood that neither are correct. That is, there's a lot of
comment in response to
post
Worth mentioning: the median price of 12 eggs was, at one point, $1.20 in 2019:
fred.stlouisfed.org/series/APU00...
comment in response to
post
All I've done is presented evidence. You've neither denied it or confirmed it. And yet you're confident that you're right. I'd say that's the definition of a conspiracy theory. (In contrast, I haven't made any statement of a theory, only evidence.)
comment in response to
post
4/CDC, more than 30 workers were exposed to an accidental lab leak of anthrax that was improperly handled.
I could go on.
My point: there's an enormous amount of evidence of a lab leak and Bayesian analysis should be considered. It's not something that should inherently dismissed as trivial.
comment in response to
post
3/Klotz and Sylvester warned of pandemic lab leak dangers. In 2017, Trump lifted the ban on gain-of-function research. Then, Marc Lipsitch and Carl Bergstrom wrote publicly this could lead to a dangerous lab leak of a modified pathogen.
There's also a lot of evidence of lab leaks. in '14 at the
comment in response to
post
2/show that there's been a growth factor of ten of BSL-3 labs in China. The book "Pandora's Gamble" demonstrates that pathogen lab leaks are common, including in the US. WHO warned in '16 that the most likely source of SC1 would be a lab leak. In 2012, in a Bulletin of the Atomic Scientists article,
comment in response to
post
1/This has become a partisan (and racial) issue when it's just a health policy topic. This post suggests that there's no basis for the lab-leak hypothesis. There is. There, however, is no direct evidence of a lab leak. In any case, the following is worth pointing out. EG, Demaneuf and De Maistre
comment in response to
post
Musk is replacing the demand curve for this company's products with a demand curve for his fascist politics. In France, in '24, sales of Teslas were down 63%.In Germany (same time period) they're down 41%.Auto sales are down 5% in both countries. Musk has become the definition of irrational behavior
comment in response to
post
If you're curious, this is a large branch of economics, called inflationary psychology. How does the psychology of inflation create inflation or create the sense that there is inflation when there is a low rate of it? This paper is a good, broad overview of it: www.riksbank.se/globalassets...
comment in response to
post
2/that there will be a government shutdown, on March 15 or in 2025 (two respective events):
polymarket.com/event/us-gov...
comment in response to
post
This is a flawed comparison. As one ages, one's view of the world changes. A better comparison would be to compare each generation at the same age. What did, eg, Gen X think of the constitution when they were the same age as Gen Z, and so on. Or, how different was Gen X from Gen Z at the same age?
comment in response to
post
This somewhat fits a base 2 log scale well. For instance, 300 is 8.2 in log base 2. 1500 is 10.5. 300k is 18.19. It might be useful to make predictions of other values that fit within this scale.
comment in response to
post
2/constructive dialogue of how these institutions can be reformed. One example, Musk shouldn't be able to simply ask for access to a sensitive computer network and be given it. This a huge flaw in the bureaucracy. Musk is simply the symptom of something very broken.
comment in response to
post
1/Yes! My colleagues who are applied mathematicians, stats and computer science professors have been talking about this frequently. Outside of this, I haven't seen any mention of this. Also, there's a lot of anxious discussion about Trump attacking institutions but I haven't seen evidence of
comment in response to
post
2/asking. The response in the media has been instead to simply react and bemoan the horrors of Trump, Musk, etc. The long-term strategy needs to be more than just simply react but also anticipate and think of solutions of better designed institutions. Musk has just found a flaw in the system
comment in response to
post
1/There are deep flaws in American democracy and how technology is handled. If this happened at a startup or a corporation, the most likely response would be to ask how the system can be designed so that this never happens again. One shouldn’t be able to get access to a sensitive network by just
comment in response to
post
2/problems. Also, many mathematicians/computer scientists/physicists/Insta influencers -- from my experience -- list him as a way they found interest in math when math in school was tedious.
Here's a sample book: www.amazon.com/Mathematical...
comment in response to
post
1/Since you didn't state the age of your son (a reasonable assumption) but did mention engineering, I assume he's around high school age. Martin Gardener was an American mathematician who wrote a lot of math and logic puzzle books. You don't need formal math education to solve almost all the
comment in response to
post
3/there's an absorbing state.
This can be solved algebraically but it's a little difficult (at least I think so).
Also, I know zero about biology. You should ask these type of questions on a biology or math stack exchange. You'll likely to get experts answering your question.
comment in response to
post
2/(this should be a compliment of the first) you can then define the equation for the likelihood of the first step occurring: p_i = p*p_i + q*p_i, where p_i is the 1st step and p and q are the probabilities of occurring and not occurring. Then, code this as a recursion function and see if
comment in response to
post
1/If you don't know, the conventional way to determine if a Markov process has an absorbing state (it's a Martingale) is to model it as a symmetric random walk. You need to model it as a recursive difference equation. As long as you have the probability of the first step occurring and not occurring
comment in response to
post
The US-Mexico border is 1954 miles. That's a solider every 5 miles (spaced uniformly, obviously unrealistic). Also, no media outlet seems to have the metrics that are going to decide what's going to happen in one month. At best, this is agreement to reach an agreement.
comment in response to
post
You’re a good man, Sean.
comment in response to
post
3/the US government needs in order to prevent this from happening again.
comment in response to
post
2/If I were Krugman, someone who progressives respect, I’d think more broadly than simply pointing out what’s happening. It may be time to admit that something catastrophic regarding American democracy may happen. If so, it should also be time to think of how to respond and what reforms
comment in response to
post
1/From my experience working at startups, if a position can be abused, either for self corruption or abuse of others, it eventually will. The equivalent seems to be happening with the federal government. Trump, Musk, etc, are taking advantage of procedural lapses in law and regulation.
comment in response to
post
Could both be considered martingales? If so, then the way to analyze these conditional expectations is through their stopping times. Specifically, what is the event that causes the respective sequence to stop for each sequence.
comment in response to
post
2/it’s impossible to know if they were wrong or not. Also, one also needs GS’s predictions for similar events. For reasons that may have to do with who their investors are, 20% may be an accurate probability event of happening.