Simulation is largely what traditional engineers do - I mean how many classes have you taken on finite element methods, discretizing PDEs, etc.? It's not web dev.
Fair. I think this is about the extent of my training, which was as an Applied Mathematics and Econ undergrad about 15 years ago: Partial differential equations : an introduction / Walter A. Strauss
> https://libcat.canterbury.ac.nz/Record/1093497/TOC
Maybe my idea of NASA was too encompassing. I figured that, apart from the engineering work, general sim would require optimizations and productionalization similar to how we have AI Engineers focused on the practical implementation of ML systems apart from the core model R&D.
I got a bit hooked on Econ for awhile which held my attention through an MS, which is when I learned about computers and then applied that into DS and development.
Most of my simulation experience is in stochastic systems and modern digital twins where agents sometimes have asymmetric information. I can see how I'm of no practical use to NASA now, but it still stings. What a bummer existing and not doing anything cool with life. A warning to youth!
I think you are underestimating your ability to contribute and also putting NASA on too much of a pedestal.
I'd argue your background is extremely valuable, but not easily traversible to NASA at the moment.
If you are deeply interested in the space, working with the newer startups in geospatial/hyperspectral imaging (be it climate or defense usecases) or CV space.
In a lot of cases, NASA is basically just acting as a coordinator between multiple vendors who are doing "the cool stuff" with less bureaucratic minutiae and stress from what's going on in DC.
Lots of interesting players in the ClimateTech and DefenseTech space who would like your background, and indirectly or directly they all work with NASA anyhow.
Thanks. I did find a space jobs site last week, and some jobs looked like they aligned closely. That's probably why I was surprised the nasa reqs weren't as broad.
I wasn't really looking for a change; I have 1 and 3 year olds and am fully remote, and the flexibility with sicknesses is really a benefit. I think it was mostly a shock to my system that I may never do anything "cool" with my life.
Not in undergrad (a single upper division class), but yes in grad school. I did a lot of applied mathematics in undergrad and only took the min required upper division probability/stats class. I didn't find it interesting at the time. But when I got to Econ grad school there was a massive focus on econometrics, and I learned it from first principals.
For languages: SAS in undergrad econ/Matlab for math classes, STATA primarily in grad school, and I pivoted to R and then python when I hit industry.
And regardless of your feelings on his political positions, in his extremely lengthy time in the House and Senate, he has only gotten three bills he sponsored passed - two renaming post offices, and a VA benefit increase.
Despite having a position of power for a very long time, he has been completely ineffective at wielding that power to achieve any of his goals.
If he was not able to change policy in any way as a Senator, how would he be able to do so as a President?
He can veto a bill then get it overridden. He has already proven 100% he lets the more conservative parts of Congress walk all over him - he can have the best ideas in the world but that won't change a thing.
If you want someone to make people discuss ideas - great, you can be at a think tank. The point of electing someone to political office is to get bills passed, so that things actually change.
Your claim is his acting as a voice for America abroad has not benefited the US?
"Walk all over..." when he is clearly out numbered not just in Congress but by voters. You want him to show up with a flamethrower and show what he's really made of?
You're not at all engaged in a sincere discussion. Coming off like an intentional astroturfer just out to propagate Bernie hate
Don't get me wrong I am not a Bernie Bro. Just aware there is a world outside him working against him this whole time too.
He can do that without being an elected congressman.
If you are elected to congress, your job is to get bills passed.
If you like his politics, there are other people like Elizabeth Warren that have remarkably similar political positions, yet are some of the most highly effective politicians in the sense of enacting policy.
Oh, but she is a woman. So better support Bernie instead.
The conspiracy of people who hate the left are the ones who prop of Bernie, because he is a joke. The more the left supports Bernie, the more people like Warren struggle to get elected, and the authoritarian likes that because Warren is actually a formidable foe, so they want to prop up ineffective people like Bernie instead.
> If you are elected to congress, your job is to get bills passed.
This is a vast oversimplification. Your job is to represent your consistuents. In many cases, this means your job is to stop bills from getting passed, especially in the current political situation.
> The more the left supports Bernie, the more people like Warren struggle to get elected
This is a very strange take. Sanders and Warren are mostly close allies and rarely compete. Both are successfully elected Senators, from separate states. Warren declined to run for President in 2016 and appeared to be supporting Sanders. In 2020 they both ran for President, but guess what, neither one of them won the Democratic nomination. In any case, it's important to recognize that elections are popularity contests and not competency tests, as should be obvious from our current President.
The issue is not even "the left." Sanders is more popular than Warren, indeed more popular than almost any politician of any party (including male politicians, if you insist on making this about gender), among political independents. Because of his popularity among independents, he's the most popular politician in the US and would have a better chance of winning the Presidency than any Democrat (of any gender), but the Democrats nonetheless refuse to nominate him. If Warren were equally popular among independents, then Democrats should nominate her, but she's not. Of course this lack of popularity among independents is not specific to Warren: most non-Democrats dislike Democrats.
In 2016, Sanders put up an unexpectedly stiff challenge to Clinton, who was considered an overwhelming frontrunner at the beginning of the race. The natural next step for the left would be to build on that momentum and push Sanders over the top in 2020. In my opinion, it's quite delusional to expect that the left would for some bizarre reason abandon Sanders in 2020 and throw their support behind Warren instead. That would make little sense. Why start over from scratch? In any case, I doubt that Warren would have fared better. The establishment doesn't want a leftist, no matter who, and they quickly conspired to consolidate around Biden, who didn't even pick Warren as his running mate.
that makes a lot of sense. unfortunately github doesn't allow multiple accounts per person. at least it didn't last time i checked. i hope they change their policy for AI agents though.
"You must be a human to create an Account. Accounts registered by "bots" or other automated methods are not permitted. We do permit machine accounts:
A machine account is an Account set up by an individual human who accepts the Terms on behalf of the Account, provides a valid email address, and is responsible for its actions. A machine account is used exclusively for performing automated tasks. Multiple users may direct the actions of a machine account, but the owner of the Account is ultimately responsible for the machine's actions. You may maintain no more than one free machine account in addition to your free Personal Account.
One person or legal entity may maintain no more than one free Account (if you choose to control a machine account as well, that's fine, but it can only be used for running a machine)."
Repeating something that you heard someone say is the literal definition of hearsay. Typically courts want to hear about facts from people who actually know those facts, not someone who heard someone talking about those facts.
This would fall under the "statement against interest" exception to hearsay, though, because obviously the person who originally said the thing isn't going to want to admit in court that they were committing a crime.
Reporting what you heard someone say is the literal definition of hearsay.
If you want to use someone saying something as evidence in court, they need to say it to the court as directly as is practical. If the person saying it isn't going to say it directly to the court, then it needs to be justified with one of the exceptions to the hearsay rule.
In this example, it would be allowed because the person saying it wouldn't be willing to admit to a crime in court.
It's a statement not offered to prove the truth of the asserted statement - non-hearsay.
It would be hearsay if offered as evidence that you had meth in your pocket. It would not if offered in evidence you were enquiring about the legality, to show intent.
Reading the decision, they seem to say that online word processing software would be viewed similarly, since the TOS may allow Google to see your data, it is not confidential, so use of Google Docs or Gmail would not be considered privileged?
DARPA sponsors lots of things that aren't specifically about weapons or killing people - medical treatment, logistics, etc. that are useful for defense/war but generally applicable.
Sure, Boston Dynamics is a bit more obvious there, but merely having DARPA funding doesn't mean it's about killing people.
Every time I've rewritten something from Python into Java, Scala, or Rust it has gotten around ~30x faster. Plus, now I can multithread too for even more speedups.
Python is absurdly slow - every method call is a string dict lookup (slots are way underused), everything is all dicts all the time, the bytecode doesn't specialize at all to observed types, it is a uniquely horrible slow language.
I love it, but python is almost uniquely a slow language.
Algorithms matter, but if you have good algorithms, or you're already linear time and just have a ton of data, rewriting something from a single-threaded Python program to a multithreaded rust program I've seen 500x speedups, where the algorithms were not improved at all.
It's the difference between a program running overnight vs. in 30 seconds. And if there are problems, the iteration speed from that is huge.
To be fair, Python as implement today is horribly slow. You could leave the language the same but apply all the tricks and heroic efforts they used to make JavaScript fast. The language would be the same, but the implementations would be faster.
Of course, in practice the available implementations are very much part of the language and its ecosystems; especially for a language like Python which is so defined by its dominant implementation of CPython.
Fair! I guess I didn't mean language as such, but as used.
But a lot of the monkey-patching kind of things and dynamism of python also means a lot of those sorts of things have to be re-checked often for correctness, so it does take a ton of optimizations off the table. (Of course, those are rare corner cases, so compilers like pypy have been able to optimize for the "happy case" and have a slow fall-back path - but pypy had a ton of incompatibility issues and now seems to be dying).
Python has a JIT compiling version in GraalPy. If you have pure Python it works well. The problem is, a lot of Python code is just callouts to C++ ML libs these days and the Python/C interop boundary just assumes you're using CPython and requires other runtimes to emulate it.
You don't even need to go all V8, you could just build something like LuaJIT and get most of the way there. LuaJIT is like 10k LOCs and V8 is 3M LOC.
The real reason is that it is a deliberate choice by the CPython project to prefer extensibility and maintainability to performance. The result is that python is a much more hackable language, with much better C interop than V8 or JVM.
But that's not saying the proofs are an issue - usually the spec you can reasonably prove in lean or another prover, say TLA+ or Z3 depending on your kind of program - has to be overly simplified and have a lot of assumptions.
However, that is powerful.
It doesn't mean your program doesn't have bugs.
It means this big scary complicated algorithm you think works but are skeptical doesn't have bugs - so when you encounter one, you know the bug is elsewhere, and you start really looking at the boundaries of what could be misspecified, if the assumptions given to the prover are actually true, etc.
It eliminates the big scary thing everyone will think is the cause of the bug as the actual cause.
This has been insanely valuable to me lately. It is also something I never really was able to do before the help of AI - vibe coding proofs about my programs is IMO one of the killer apps of AI, since there aren't a ton of great resources yet about how to do it well since it is rarely done.
This surprises me. Formal verification so far has been a very niche thing apart from conventional type systems. I didn't think lack of vibe coding was much of a bottleneck in the past. Where do you use it?
The problem is implementing anything approximately twice is a hard sell… this is no longer true, though - TLA+ models are cheap now. You should be using them when writing any sort of distributed systems, which is basically everything nowadays.
Roughly anything that, say, has the complexity of a leetcode medium level problem that isn't already an extremely well known algorithm.
Any moderately complex thread safety thing with a few moving parts (e.g. there are multiple mutexes involved in various parts of the system, verify no deadlocks).
The lack of vibe coding has been a bottleneck for literally everything before.
When I see people say the hate vibe coding, I think "why do you hate formal verification? Because you could be spending your time on formal verification instead of removing "code smells" that don't hurt anything from vibe code."
Something under-appreciated: If you pretend a company is paying out 100% of profits as dividends (which it theoretically potentially could, and is useful as a financial modelling tool), then the inverse of P/E, E/P, is an interest rate on the price of the stock.
Ideal P/Es thus shouldn't be flat, they should be tracking long-term bond rates. This isn't an empirical observation, just a theoretical one of what "ideal" should be. But one should rationally expect P/Es to go up when interest rates drop.
It is disappointing to me that even Shiller doesn't really consider this much.
reply