sega_sai

22 ms·

331.
by sega_sai 2y ago
It is such an absurd take. That essentially implies that out of three options 1) Do nothing 2) Try to have detailed review of different programs 3) Stop everything The latter is the most beneficial.
332.
by sega_sai 2y ago
Yes, I did scroll through the list of stopped federal programs (listed by NYT), and didn't see much science in my area at least (physics), but according to Terence it is clear that some things in STEM are stopped despite that.
333.
by sega_sai 2y ago
yes, sure, reviewing thousands of programs will take very little time.
334.
by sega_sai 2y ago
The missed opportunities of highly beneficial for society work for hedge-funds for example. Jokes aside. Research is ~ 5% of US budget https://ncses.nsf.gov/pubs/nsf25313 and mathematics is 1% of the that. Also, yes, c
335.
by sega_sai 2y ago
The problem was stated in the beginning as 'Consider the problem of generating N points on a disk of diameter D. The function must be deterministic and the points somewhat uniformly spread around on the disk. Importantly, I want to gen
336.
by sega_sai 2y ago
You would fix the seed once in the beginning before sampling. Also you would need to sample exactly N-times as there no rejection required. as the points will all be within the circle by construction and will have uniform distribution
337.
by sega_sai 2y ago
A bit of a strange ad-hoc solution. First as other pointed out, one can generate the Sobol sequence or similar, but even with a random sample one can generate the random numbers deterministically. I.e Sample r^2 ~ Uniform(0,r0^2) and angle
338.
by sega_sai 2y ago
There is zero chance that this kind of Justice department actions will be continued in the current administration
339.
by sega_sai 2y ago
I am sorry, but the comparison of 'stack overflow' and tutorials to LLMs is bizarre. The amount of time to get to the answer from LLMs is drastically shorter. And claiming that the they only 'cache thing' is just wrong.
340.
by sega_sai 2y ago
It seems there is a battle of two opposite view-points. One is that LLMs are just dumb autocompletes with no ability to understand anything. Another is that LLMs can already right now be substitutes for programmers. I personally thing it
341.
by sega_sai 2y ago
I have just tried ollama's r1-14b model on a statistics calculation I needed to do, and it is scary to see how in real time the model tries some approaches, backtracks, chooses alternative ones, checka them. It really reminds of human
342.
by sega_sai 2y ago
I would have thought that any kind of contract would require a signature or something rather than agreement by text (but obviously I'm not a lawyer)
343.
by sega_sai 2y ago
Interesting, but it does not seem to be an overview of gradient optimisers, but rather gradient optimisers in ML, as I see no mentions of BFGS and the likes.
344.
by sega_sai 2y ago
I am not an expert, but it is unclear to me. A specialized robot vs general purpose robot is a scale question. I.e. if the task is big enough(i.e there is need for millions of these robots) and there is enough money, than yes specialized r
345.
by sega_sai 2y ago
I am waiting to see what will happen with NASA, NSF... Obviously it's less important than health, but still thousands of people, lives.
346.
by sega_sai 2y ago
There are a lot of issues in academia, but that number (40%) is not one of them. Not everyone is suited for an academic job, and doing the PhD is not enough to understand what it is to do independent research/find your niche what you a
347.
by sega_sai 2y ago
For the record -- the law for TikTok divestment was not passed on its own, but was instead included in the foreign aid (including Ukraine) package https://edition.cnn.com/2024/04/23/tech/congress-tiktok-b
348.
An Extremely Detailed Map of the 2024 Election(nytimes.com)
3 points by sega_sai 2y ago|0 comments
349.
by sega_sai 2y ago
Eurasia -> Eastasia
350.
by sega_sai 2y ago
At least this paper has only 43 citations over last 10 years, which is really nothing for Nature, which means it's basically irrelevant. (Obviously it is still a good idea to identify cheaters)
351.
by sega_sai 2y ago
It was not ice, but fibers from the sun shield. The ice issue was resolved by heating the satellite. The stray light issue affected spectra measurements, but not the astrometric side of the mission
352.
by sega_sai 2y ago
I've just tried to make it run something, and I just could not force to include the python code inside ``` ``` quotation marks. It always wants to put word python after three quotes, like this: ```python .. code.. ``` I wonder if th
353.
by sega_sai 2y ago
I meant that you tend to spend free time on that as opposed to treating it like 9 to 5 job. and again it is important that if you do that, it is because you just want to see what comes out of your experiment/learn a new thing etc rathe
354.
by sega_sai 2y ago
As a professor many years after the PhD my advice is to do the PhD only if you are genuinely excited / cannot stop yourself from doing research. Only then it will outweigh the negatives of difficulties of getting the jobs, somewhat low
355.
by sega_sai 2y ago
I have not fully read the whole page, but I wonder if this type of tree of expressions leading to a proof can be explored efficiently with the same idea as Monte-Carlo Tree Search in chess. I.e. surely you can explore all the possible combi
356.
Apple says it will update AI feature after BBC complaint(bbc.co.uk)
3 points by sega_sai 2y ago|1 comments
357.
by sega_sai 2y ago
It is a somewhat interesting take. What I am interested in as a person teaching a computing course, what is the best way to force people to understand/interact with the code coming from the LLM. I.e. when I give computing problems to s
358.
by sega_sai 2y ago
After watching the video, I think what he says in it is fair. It is perfectly reasonable to want to have more viewers of the videos. Not only that gives more profit, but also that's the whole point of making educational videos in the f
359.
by sega_sai 2y ago
Overcoming social anxiety and shyness by Butler was one of the books that was pretty good.
360.
by sega_sai 2y ago
I don't find this convincing as an introvert. Dealing with shyness or being an introvert required me reading a few books and techniques how to deal with it. In the end that helped, but it was not as easy as just 'try harder'

More ›