I know red wine is just as bad for me as any other alcohol, which is to say quite bad, and in any dose. I meant I've largely quit because of that. But I just can't quit you, Red Wine.
Lofty valuation with a PE ratio of over 90 - even considering the growth and margins. Also almost no revenue growth in 2022/23 according to quickfs.net. Too expensive for my taste.
You're being misled by the site. Those aren't really their 2023 results, but their FY2023 results. And Nvidia's fiscal years are just nuts, basically FY2023 is 2022.
Their 2023 results will be reported as FY2024, and aren't out yet. But just the first three quarters showed 50% more revenue than the entire FY2023.
You are right, I was wrong. I have compared financial years, not actual years. There seems to be strong improvements with respect to revenue (66% TTM increase) and earnings per share (334% increase TTM).
a) Outright cheating in the interview is probably a bad sign about the candidate (whereas the interviewers might be ok with, say, using GenAI, which is increasingly common in many sectors). But impersonating a candidate, or having a friend sit in with you, would always be a bad sign.
b) Not necessarily, some companies would care if their contract writer stole/copied/repurposed/reused content from a competitor/regurgitated copyrighted material. Think also GenAI hallucinations, fabricated citations, etc. But sure, most cost-conscious companies won't give a hoot, so they could probably get by with cutting corners.
I think what parent meant was that Euler spent a few decades in Russia because of the funding provided by the empire. He spoke fluent Russian, even though there was a large German-speaking community there.
But it was typical for scientists to travel far for money. Some of the Bernoullis, a family famous for mathematicians, also worked in Russian for quite a while.
Does it really matter who payed 'em and what languages they spoke?
I wonder what he would have considered himself - he spent more time in St Petersburg than anywhere else but was born in Basel and spent a good bit of time in Berlin. For what it's worth, the Opera Omnia are published by the Swiss Academy of Sciences.
For modern mathematics, I think Alexander Grothendieck was one of the most influencial talents in the last century. He unified large branches of mathematics in a very short period. He was more a bird than a frog [1], however.
Seriously? Oh god, please no. We need less of all that BS. People are thinking that are being "tutored" by AI, when in fact is just the output of a number crunching program. Reading books will get you way way closer to solid knowledge than the output crap of this so called "AI".
I don't think such an attitude is warranted. While I'm skeptical about the potential of LLMs as AGI (whatever that means), being able to summarize subjects and even come up with test questions can be very valuable for learning. I am concerned about the confabulation aspect, though. I wonder if there could be a uncertainty metric for that.
Happy to disagree here. I am hopeful that students get apps which provide them with
- fast feedback loops
- better adjustment to their learning speed
- more patience
- gamification
- oppurtunity to ask endless questions
Of course, only as an addition to the current mix. For math problems, this will be easier than for other contexts.
I like text books for pop science or really hard things - like university level education. But I am surprised to see them as an option for basic education.
Is Euler really well known in public?
Einstein is way more well known than Euler, I'd guess.
Even if you will ask most famous mathematician, it is unlikely that he is named. Probably Pythagoras.
This is purely an indictment of modern mathematics education. Anyone doing ordinary high school math should have had their ears full of Euler's contributions for at least a couple of years towards the end. Math teachers who don't mention Euler need to be put onto a mock Königsberg and forced to search for a solution.
Sure, I didn't really mean the general public but meant scientific/mathematically oriented public. When I studied math in college, I found Euler everywhere to the point that it was hard not to stop and wonder how the hell this guy had so many fundamental discoveries. By contrast, I have awareness of present day figures like Tao and Erdos, but I don't really understand their work and perceive it to be specialized enough that it is unlikely to be widely understood in the way that say e^(pi * i) + 1 = 0 is.
Riemann died young, too. I only really know him for the Riemann sum formulation of integrals (I have yet to learn complex analysis), but I wouldn't be surprised if he has some popular reach through the Riemann hypothesis.
Another example: the notion of an n-dimensional geometric structure with intrinsic, variable curvature that serves as a mathematical foundation for the theory of general relativity originated with Riemann[1].
Round musket balls in hand packed smooth bore dueling pistols at 20 paces is inherently inaccurate - ten shots in a row from a pistol clamped in a vise will have a spread greater than the profile of an oppenent staning sideways.
Bullets with rifling came about circa 1820 (ish) but were not widespread by 1832 and traditional dueling pistols remained the norm for quite some time, the element of chance likely factored in as part of the hand of god influencing outcomes.
For some in the eighteenth century, duelling with less-accurate, smooth-bore weapons was preferred as they viewed it as allowing the judgement of God to take a role in deciding the outcome of the encounter.
There's not a lot of detail regarding the duel of Galois, his opponent isn't known for certain, nor the precise reasons, let alone the type of guns and ammunition used.
I am mathematician by training. The algorithms relied mostly on undergrad level mathematics when I took some courses six years ago, not easy, but I think there are harder algorithms.
I think it is qualitatively different from programming, because you try to find a reasonable good fit for data you know to guess new future data. Classical programming relies on rules and decisions. ML is closer to numerics, statistics, simulations. Guess the function from the data vs define a function and programm it.
You can easily get into territory that is harder than undergrad. Like it’s completely possible you’d need something like rejection sampling, functional time series, Jeffrey’s priors, etc
But misunderstood, see below…