nixonaddiction

i did manual labor part time for 6 years during high school and college. i liked it a lot. i had opportunities to move up to management, but i refused them because i preferred the more physical aspects of the job. also it would suck for my team to lose me as i was just faster than everyone else.

my coworkers would always tell me, the youngest, about how much their bodies hurt, how tired they were, how hard it was to make ends meet. they either wanted me to stick around because i made their jobs easier, or wanted me to get my degree and go. ive seen accidents happen, but none terrible. people were always so worried about my safety. i got lucky and was never injured.

i dont know. ive been in my current role for 4 months now. its not bad. i dont dislike it. i cant tell if i miss my old job or if im just lonely in a new place.

i really miss zoning out and working with my hands though. there are moments where you have to lock in, i remember moments where i was working with something dangerous and how acutely aware of the position of my body i was. but there were moments where you did a repetitive task for a while. and i always liked that aspect. the physicality.

i think about going back to it a lot actually. maybe im just nostalgic. should give it a few years to get used to the cubicle life. i really do miss it though. might do what my father did and compulsively renovate his basement to get the high of building something with his hands.

i dont think this works but i do think the patterns look really cool. its like a hawaiian shirt but worse. im lowk obsessed i lowk will be purchasing once i have the funds.

i think financially, frontier labs are going to crash. this is purely from a financial standpoint. the question is when, this article posits 2027/28. i need to better verify the claim. (its easy to be right, its hard to time being right.)

however, when frontier labs become financially insolvent (and they will eventually, the question is when) i think smaller local models will end up taking over. and i think there is a bet to be placed on smaller models here. the demand for ai isnt going away. peoples workflows would be crushed without it. the question actually becomes, who either purchases frontier labs, or what cheaper alternative replaces frontier labs?

ive never been able to use zettelkattsen. i like having daily notes instead of different notes for different ideas. its easier this way. my obsidian vault is a hybrid of diary and research and i am trying to separate the diary and research but its not easy tbh. (i suffer from a crushing urge to note that i "just ate a sandwich, was good" while liveblogging research and debugging.)

that being said, i got a big kick out of feeding my daily notes into my local model. i had some more diary-like entries that i would never share with anyone, but it was extremely funny to have my local model take a 3000 word recap of what i did in a day and summarize it like "writer reflects on their job: they worry about their performance. meals: the writer ate oatmeal, and a sandwich. family: the author contemplates their fathers absence growing up."

ive developed a really large corpus of written work over the years and am hoping to fine tune a model using my diary as the training set. i think this would yield hilarious results. keeping it all local though, i drop my passwords in there at times and i dont want that leaked.

well. to an extent. one always has to sell themselves. view selling as "convincing people to give me money." that i feel is pretty similar regardless of if i am applying for academic grants or applying for vc funding or running political fundraisers. (the latter being my area of experience.) theres kinda no way to avoid "selling" when you have an novel idea you want executed.

what i mean is less the goal being making money, more the goal being getting money to fund research. theres no way to avoid "selling" in life. (alas.) what i would be selling would be my research topic to venture capitalists, instead of a product to people. (obviously making some mentions of "well um this could be a sellable product one day" and like, a halfassed marketing campaign in order to better sell my idea to vcs.)

company failing isnt incompatible with the goal of doing research. if i can get a few years of funding by at least putting up the illusion of a sellable product it could be worth it. maybe i'd even get lucky and end up with an actual product that makes money. but uh, comfort in failure. the process of learning over the promise of success.

sorry i sound so evil and scammy rn ahhhhh promise im not trying to be on forbes 30 under 30. just considering vc funding as an alternate route to academia with regard to getting funding for research. (i will need to dance for dollars either way. but vc funding is seeming like it has less uh, annoying interpersonal academic hoops to jump through.) but id need a few years working under people and gaining experience before i would try at this lol.

can vc money just fund basic research?

im slowly getting confused on what founding a company even is. i am in grad school right now. a lot of funding and opportunities come from guys who are like "we need you to innovate we need you to act like a startup." its honestly kinda annoying. i just want to do basic research. i dont want to sell anything. i just want to build things. but the way this article frames it- is founding a company just getting funding for a research project? but using different words? could i just pull up with a list of my research interests, convince people its a sellable product, and score funding? obviously id need to deal with financial overhead myself (a pain) but hypothetically. lets say i complete my degree, maybe spend a year or three in industry networking and better understanding the problem space. then pitch myself to vcs with hope of funding whatever topics im interested and have my company function like a research group except instead of publishing papers we build products? like no profit no big deal, i never intended to profit. which, is this not what openai is doing?

unethical to lie about my intents (i want to do research not start a company) but i feel the end goal (more funding for research) could be worth it.

i was trying to build llms from scratch at 17. failed miserably because i did not know linear algebra. ended up in a different but adjacent field. when chatgtp got big suddenly there were so many people doing llms and i didnt want to compete like that. so much of what was happening was hype, and that really turned me off. am back to building llms from scratch, but like, its a journey teaching myself all the theory on top of my job. i have decent fundamentals, but i need a better grasp of all the advancements in the field in the past 5 years before i would feel comfortable designing anything. baby steps, essentially. am working on better understanding all the layers of an ai while implementing a rag on my local model as an experiment. 17 year olds should learn whatever theyre interested in but need fundamentals in order to do anything advanced.

people dont hire ambitious young women because theyre afraid theyll fall in love with them???

for a lot of the article i was mentally replacing "ai" with "calculator" and going "yeah we will just need longer education times. more schooling etc."

but the phrasing of "who pays for the increased schooling times?" is a good one. i think "debt" can be a decent way to conceptualize the cost and repayment of training someone. feels evil to say, but viewing people as firms you can invest in and expect returns upon. you know not all loans will be repayed, but hopefully they'll average to a profit. (risk management etc.)

student loans are. a decent example. the government/private enterprise gives money to pay for education, then this is repayed, providing a financial incentive for paying for someone else's longer schooling timelines. firms investing in training can be viewed as an extension of student loans. but then ah, there are countless stories of how debtor/creditor relationships can be exploited. indentured servitude etc. there are a lot of complications coming to mind. also "altruistic" people who give without expectation of repayment. or the divide between like, communal vs individualistic cultures. (individualism, i argue, encourages the formalization of debt, as opposed to a more communal culture where the expectation of repayment is informal.) you could do math on how many people pay vs how many people benefit, who is the biggest stakeholder, etc.

but i am on my lunch break and need to get back to my work. good article tho. good topic to bring up.

apparently it helps reduce risk of mouth cancer..... what im hearing is rotate this and zyns in my mouth, see what happens.

i mean my line is always "when." people have been saying ai is a bubble and will crash since the beginning. the question has always been "when." i think the best framing of the ai crash is not in terms of software, but in terms of real estate and infrastructure. the real cost is hidden in construction contracts and people buying up land (and chips, transformers, generators, etc), not in technology. this will be the 2008 real estate crash, not the 2000 dot com bubble. (altho this point is not my most well researched point, i tbh need to look more into hard numbers of where the most risk is distributed, ai software startups or data center contracts. but i also am in grad school i have papers to publish and no time.) im also gonna push back on the "diseconomy of scale" point. while its true that the best models rn use the most resources, we are building better smaller models as well. i recently set one up on my 12 gb vram and while its not nearly as good as claude, it works. i make it write really simple code for me. and we are still improving small models. i think theres some level of hope to economy of scale. obviously there are going to be physical limits, but im sure there will be a way to get economies of scale to work.

“wahhhh this is bad” ok sure but how. what will the downstream effects be. how can we model increased ocean temperatures and how they will affect weather patterns or whatever? gives me no info on the implications, let alone info on the implications with rigor.

yeah yeah nicotine bad. let me have this one vice. i used to pop 6mgs during exams and life was awesome. i <3 smokeless nicotine. have tried pouches, patches, lozenges, gum. the gum is my favorite. had a 12mg once that stuff burned my gums and i swore off higher mg pouches. i honestly dont get how people use them.

yea im doing my masters in dft research so ik abt this. depends what u want 2 simulate! chemists more do molecular dynamics type stuff and will use experimental data for fitting data etc. like uh what surface of a metal water will react with from thermodynamics or something. (that isnt my field lol i just know a lot of catalysis guys.)

truly ab initio methods involve figuring out electronic properties from scratch like ionization energy or bandstructure. the real issue is that we dont have exact relations for the exchange and correlation terms. we can know the kinetic energy and charge screening, but we dont know how the electrons are interacting with each other. generally the xc term is treated as a function of electron density or its gradient (see: lda, gga, meta-gga) but there are so many different ways to approximate that. different models are good for different applications also, like transition metals vs organics. and then theres the issue of basis sets (most people use gaussian basis sets that have been tuned over many years but theres also plane waves and finite element methods) which can also change results. and even once u have a decent approximation of density you can try perturbative methods (GW family, delta scf i count also) to try and improve the approximation. i am rambling and typing this on my phone. essentially yes, but often calculations are a little inaccurate. but more accuracy has a higher computational cost, which makes it hard to run larger simulations. tradeoffs of engineering. hope this was coherent.

i learned to code primarily from really old guys so i started on vim. i was forced to use vim because i was doing a lot of remote hpc work. now i like vim. use vim/tmux combo to improve my mouseless workflow. i just need a better way to look at all my files and pdfs without touching my mouse. i should try sublime as i cant have any real comparison between the two without having tried both. but i much prefer mouseless workflows.

ive been using obsidian for about 4 years. my main obsidian is my diary. recaps of everything i did that day, and working through my thoughts. sometimes i draft emails in thre. i dont reread it unless i need to know what i was doing on a specific day 3 years ago. once a daily note is finished i do not touch it again. it is mainly there for historical purposes. i have a separate obsidian for work notes. code im working on, todolists, etc. i am constantly editing these documents as needed. that obsidian functions more like a file management system. i have folders for different code projects, with markdown files explaining what i am doing and what projects relate to other projects. i mainly slip up when i write something in the wrong document. but i can just copypaste it over. the system is still evolving though, i just entered grad school and my needs changed rapidly, and i am shifting my information management system to match that.

ok so google paid companies like apple to make google the default search engine. you cannot claim everything was just because their product was good. you can argue that striking the deal with apple was just a smart business move, but google isn't winning just because of their r&d.

"healthcare should be for everyone" is a great claim to make. but then the question is implementation. how will you get rid of the current system and replace it with a more equitable one? people are generally hesitant to make changes unless things are really bad. i like to think of this in terms of chemical bonds - people are bonded to their current systems, and wont break those bonds unless they are under enough stress that bond breakage is favorable. and once you start arguing for destruction of the current system, the morality gets fuzzy. do you support accelerationism, or a more gradual change? and then once you are in the weeds of implementing a fairer healthcare system, things are just genuinely terrible. i am very uninvolved in the healthcare system, but you need organizational structures, supply chain, etc. someone somewhere will probably try and be selfish about things which will make everything harder. structures will have to be built to deal with legal minutia. and meanwhile there are all these other preexisting systems used to the former system that struggle to make the switch instantaneously? every question is complicated and awful once you think about implementation. nothing is ever self evident. imo!

im a nuance enjoyer when it comes to politics too but i wouldnt say i know adequate amounts about economics, politics, game theory, etc. i might know slightly more about my preferred fields than the average person, but im still woefully incompetent. so im always hesitant to lay judgement. especially because politics is such a complex system. its difficult to make the probabilities the author speaks of unless you make a bunch of assumptions. which is terrible and miserable. things get even worse when you think about things at a global vs local political level, which are just completely different in dynamics.

i hate rationalists because it's like. you cant logically reason your way out of this one buddy. the system is far too complex for rationalism to work. sometimes its easier to just align with the groupthink and focus on other things you deem more important. hanging out with friends vs spending all day in your room teaching yourself about tribal relations in central africa so you can have your own unique opinions on us foreign policy.

just sounds like teaching more students philosophy, but now the dominant philosophy is systems theory. which i super agree with. systems are relevant everywhere from literature to engineering. i agree with everyone else who has said true polymathy can't be taught.

i really doubt it was an elaborate plan. more, people were annoyed by apple's monopoly, engineered a solution, and apple proceeded to shoot themself in the foot by trying to hold on to the said monopoly. beeper mini is probably leaning into provoking apple after witnessing the reaction, which makes sense. it gives them press. we'll see how this plays out. i know the justice department is building a case, and they're probably salivating at this evidence, but the 2024 election could result in less antitrust friendly actors working against the case. possibly. i need to research the justice department more.