hmm I think i am still bitter about that I did not get an opportunity to stay in physics for a long term. I thought I was over it, but it may take much more time for me to be over this
February 9, 2026 at 7:35 PM
hmm I think i am still bitter about that I did not get an opportunity to stay in physics for a long term. I thought I was over it, but it may take much more time for me to be over this
friends and family are what really matter
February 9, 2026 at 6:24 AM
friends and family are what really matter
haha really without the help of friends, my attempt to transition could've been much more difficult or even almost beyond impossible.
February 9, 2026 at 6:24 AM
haha really without the help of friends, my attempt to transition could've been much more difficult or even almost beyond impossible.
one exciting aspect of going to industry is that once i make enough money, i can come back to physics to work on whatever problem that interests me even without a university affiliation
February 6, 2026 at 10:48 PM
one exciting aspect of going to industry is that once i make enough money, i can come back to physics to work on whatever problem that interests me even without a university affiliation
even if i change my mind to stay in physics, learning how to make machine learning models seems to be going to be an extremely useful skill.
February 6, 2026 at 10:40 PM
even if i change my mind to stay in physics, learning how to make machine learning models seems to be going to be an extremely useful skill.
ooh i for the first time bought gpu with cuda. excited to have fun with this guy as i build some home NNs
February 6, 2026 at 7:59 AM
ooh i for the first time bought gpu with cuda. excited to have fun with this guy as i build some home NNs
although thinking about finding a job is super stressful, learning ml and python is extremely fun i would say. knowing that something has a solution, and i can find a solution in a reasonable time makes me super excited..
February 5, 2026 at 7:38 AM
although thinking about finding a job is super stressful, learning ml and python is extremely fun i would say. knowing that something has a solution, and i can find a solution in a reasonable time makes me super excited..
To do science one may really belive in it, no matter how folks laugh at the dream. In that regard, many I know are laudable
February 3, 2026 at 10:00 AM
To do science one may really belive in it, no matter how folks laugh at the dream. In that regard, many I know are laudable
should really try to us ai to find ``nice" string vertices.. it would be so cool if ai can find some patterns to see how to organize string theory
February 1, 2026 at 12:00 AM
should really try to us ai to find ``nice" string vertices.. it would be so cool if ai can find some patterns to see how to organize string theory
Haha finding a job is super stressful...
January 31, 2026 at 7:52 PM
Haha finding a job is super stressful...
how do people choose which activation functions to use? is there even a design principle for all of this?
January 30, 2026 at 6:34 AM
how do people choose which activation functions to use? is there even a design principle for all of this?
updated my (academic) cv of failure. probably the last update for this guy?
www.dropbox.com/scl/fi/3p1h6...
www.dropbox.com/scl/fi/3p1h6...
www.dropbox.com
January 30, 2026 at 4:46 AM
updated my (academic) cv of failure. probably the last update for this guy?
www.dropbox.com/scl/fi/3p1h6...
www.dropbox.com/scl/fi/3p1h6...
Reposted by Manki
Science and Technology Facilities Council told staff it needs to reduce its spending by £162 million by 2029-30, requiring “substantial savings” to be made. 😬
www.researchprofessionalnews.com/rr-news-uk-r...
www.researchprofessionalnews.com/rr-news-uk-r...
UK physics research and science facilities face ‘substantial’ cuts - Research Professional News
Exclusive: Science and Technology Facilities Council seeks £162m cost savings, with existing projects facing axe
www.researchprofessionalnews.com
January 29, 2026 at 3:16 PM
Science and Technology Facilities Council told staff it needs to reduce its spending by £162 million by 2029-30, requiring “substantial savings” to be made. 😬
www.researchprofessionalnews.com/rr-news-uk-r...
www.researchprofessionalnews.com/rr-news-uk-r...
"이 대통령은 친중, 한국은 베네수"…미국 정부에 개입 요청한 쿠팡 | 중앙일보
쿠팡, 韓정부 차별 주장하며 美에 개입 요청
www.joongang.co.kr
January 29, 2026 at 4:41 AM
Reposted by Manki
I often find myself explaining something that is easy to take for granted, but should be evident: There is no reason why the program of Figuring Out Nature should proceed in time steps that are less than or comparable to the duration of one scientist’s career. (1/n)
Is Particle Physics Dead, Dying, or Just Hard? | Quanta Magazine
Columnist Natalie Wolchover checks in with particle physicists more than a decade after the field entered a profound crisis.
www.quantamagazine.org
January 27, 2026 at 4:57 PM
I often find myself explaining something that is easy to take for granted, but should be evident: There is no reason why the program of Figuring Out Nature should proceed in time steps that are less than or comparable to the duration of one scientist’s career. (1/n)
i am having too much fun with claude code
January 27, 2026 at 9:52 AM
i am having too much fun with claude code
man i am really really excited....... i should spend significant amount of time with the help of ai to create a package for SFT
January 26, 2026 at 7:48 AM
man i am really really excited....... i should spend significant amount of time with the help of ai to create a package for SFT
i think it is not a dream to write a code package to do SFT calculations. this year, i am going to use ML to write an SFT package for calculations. I think it looks really possible to do all kinds of string loop calculations that i finally understand how to do in principle but couldn't in practice
January 26, 2026 at 7:23 AM
i think it is not a dream to write a code package to do SFT calculations. this year, i am going to use ML to write an SFT package for calculations. I think it looks really possible to do all kinds of string loop calculations that i finally understand how to do in principle but couldn't in practice
shit copilot is insanely good.
January 26, 2026 at 7:22 AM
shit copilot is insanely good.
Reposted by Manki
Damn proud of Minnesota.
January 26, 2026 at 2:01 AM
Damn proud of Minnesota.
hmm chat gpt is actually not so bad as a calculator and information summary tool..
January 24, 2026 at 6:38 PM
hmm chat gpt is actually not so bad as a calculator and information summary tool..