|
jtrn 1 hour ago | next [–] I am extremely skeptical about this research. Firstly, the "plastic is bad" narrative is too easy to believe, just because plastic is ugly. It's also starting to follow the pattern I have seen in clinical psychology, where many studies find promising initial results, but no one bothers to conduct the real-life randomized controlled study that shows an actual clinical effect. And after a couple of decades of dogma, someone points out that the emperor has no clothes. Suddenly, the whole field looks silly, be it EMDR, classical psychotherapy, mindfulness therapy for severe mental illness, or the abuse craze in the 1990s. The longer it goes without concrete and clinically significant findings, the larger I think the probability of the findings being wrong becomes. I also find it strange that so few of the studies I have read ever comment on the fact that our system might be fully capable of removing the nanoparticles by itself, just as it removes everything from dust to methylmercury. We do not know if this is the case, but the fact that nobody is addressing this further strengthens my fear that there is a lot of confirmation bias going on. Every time I post something like this, I get a lot of angry responses, so I can try to preempt some of them by saying: I am not asserting that microplastics are safe. But the pattern of lots of pilot studies, and few studies that significantly prove the theory, is very recognizable to me. Until someone either conducts a naturalistic experiment with lots of people exposed to large doses of microplastics and compares them to a control group, or we expose some larger animals to microplastics over a long time in a true randomized controlled study, I'm going to remain skeptical. reply
|
#
?
Nov 18, 2023 22:28
|
|
|
|
| # ? Apr 28, 2024 22:03 |
|
|
Ono-Sendai 5 minutes ago | next [–] While obviously a brilliant scientist, sometimes I think Einstein set physics on the wrong path for 100 or so years. The concept of the photon seems unnecessary and confusing to people, if not just plain wrong. Lorentz ether theory seems like a better theory (or interpretation of the empirical data at least) than special relativity. The last major contribution of Einstein is general relativity, which I suspect is incorrect too. (Obviously I can't offer a better theory though) reply
|
|
#
?
Nov 19, 2023 02:58
|
|
|
hn thread: obviously I can’t offer a better theory
|
|
#
?
Nov 19, 2023 03:10
|
|
|
lordamercy people who don't understand how science works but can't stop acting like they do get my goat
|
|
#
?
Nov 19, 2023 03:11
|
|
|
Subjunctive posted:hn thread: obviously I can’t offer a better theory beat me to the punch by seconds
|
|
#
?
Nov 19, 2023 03:12
|
|
|
Tired: Complaining that string theory is "not even wrong" Wired: complaining that quantum mechanics is "not even wrong" Inspired: complaining that relativity is "not even wrong"
|
|
#
?
Nov 19, 2023 03:35
|
|
|
look just because something has been confirmed in literally every experiment ever conducted to test it that doesn't mean its right or anything
|
|
#
?
Nov 19, 2023 03:37
|
|
|
i mean, every scientific theory we have is definitely wrong. but all the empirical results will be captured by future theories and current theories will be expressable with the concepts of future theories, and future theories will be able to explain the successes and failures of current theories thinking that anything we have today is right full stop is probably not a great strategy. but i mean, they're clearly closer than what we had 10/50/100/200 years ago and all that. and it's not like we're sure how are theories are wrong, if we knew that we'd change them to be right.
|
|
#
?
Nov 19, 2023 03:42
|
|
|
Internet Janitor posted:my guess is his idea of "cramming" is to pop a bunch of adderall and then tell a personal assistant to summarize textbooks to him while he half-listens and furiously composes tweets on his phone i think it mostly consists of surrounding himself with yes-men who will do their absolute best to convince him that the thing he just said made sense so that he thinks he understands what they're talking about
|
|
#
?
Nov 19, 2023 06:45
|
|
|
 why all you geniuses believe in the dumbest poo poo https://news.ycombinator.com/item?id=38331493 the dude also appears to have some jordan peterson posts because of course
|
|
#
?
Nov 19, 2023 14:36
|
|
|
david-gpu 3 hours ago | parent | context | flag | favorite | on: Persuasion through status rather than argument The elephant in the room is the widespread social acceptance of "girls only" and "women only" clubs/associations/events. To me they elicit the exact same response as "whites only" organizations would, except I've never actually seen the latter IRL. The saddest part to me is that this form of segregation is championed by people who identify as socially progressive.
|
|
#
?
Nov 20, 2023 17:47
|
|
|
fritz posted:david-gpu 3 hours ago | parent | context | flag | favorite | on: Persuasion through status rather than argument 
|
|
#
?
Nov 20, 2023 18:01
|
|
|
concordDance 6 days ago | next [–] If it helps, he doesn't want to hurt your trans friends, it isn't a terminal value. If you could convince him that trans people and potentially-trans people will be better off with people respecting trans pronouns and teaching about transness in schools then he would change his behavior. reply
|
|
#
?
Nov 20, 2023 22:13
|
|
|
are they talking about Musk?
|
|
#
?
Nov 20, 2023 22:16
|
|
|
i thought the horse obsession bit was especially funny
|
|
#
?
Nov 20, 2023 22:16
|
|
|
purpleblue 33 minutes ago | prev | next [–] Ahhh, so I guess this really was just a shakedown after all, like some banana republic despot. Color me stupid for thinking that the US cared about law and order and not about making money through the justice system. reply guess the article US Seeks More Than $4B from Binance to End Criminal Case
|
|
#
?
Nov 20, 2023 22:28
|
|
|
fritz posted:Ono-Sendai 5 minutes ago | next [–] christ hn has some dumb takes but "ether theory is better than relativity" made me have to lean back in my chair and contemplate what i'm doing with my life for a second.
|
|
#
?
Nov 20, 2023 23:02
|
|
|
Subjunctive posted:are they talking about Musk? you know it!
|
|
#
?
Nov 20, 2023 23:03
|
|
|
Neito posted:christ hn has some dumb takes but "ether theory is better than relativity" made me have to lean back in my chair and contemplate what i'm doing with my life for a second. any physics topic on hn brings out the cranks but anti-relativity makes a change from the normal parade of anti-dark-matter
|
|
#
?
Nov 20, 2023 23:12
|
|
|
Qwertycoatl posted:any physics topic on hn brings out the cranks but anti-relativity makes a change from the normal parade of anti-dark-matter what
|
|
#
?
Nov 20, 2023 23:55
|
|
|
quote:Fundamentally ethics are a luxury. I might have some if I ever become rich and financially independent of the rest of society, but until that point I will not have any. this is so on-the-nose i'm inclined to believe it might be a troll, but then again, hacker news...
|
|
#
?
Nov 21, 2023 18:49
|
|
|
Neito posted:what any time there's an article on hn that mentions dark matter (or an article about cosmology at all) a load of hners come out of the woodwork to complain about it
|
|
#
?
Nov 21, 2023 20:24
|
|
|
I will pick and choose only those parts of modern physics that feel intuitive
|
|
#
?
Nov 21, 2023 21:51
|
|
|
fritz posted:thepasswordis 12 hours ago | parent | context | flag | favorite | on: Ilya Sutskever "at the center" of Altman firing? someone should edit the opening of terminator so that a skull is reading this post until a terminator’s giant metal foot crushes it
|
|
#
?
Nov 21, 2023 22:36
|
|
|
we can make that clip with the power of AI 
|
|
#
?
Nov 21, 2023 22:37
|
|
|
mcbishop 50 minutes ago | parent | context | flag | favorite | on: OpenAI API and ChatGTP Outage Its outage shows me how much I now rely on ChatGPT during novel programming. I've been smashing that refresh button like a lab rat with a pellet lever. reply
|
|
#
?
Nov 22, 2023 01:27
|
|
|
I can't tell if they're talking about using concepts new to them in code or if this is a new term for thinking up amazon book stuffing prompts
|
|
#
?
Nov 22, 2023 01:30
|
|
|
a simply remarkable percentage of programmers fear and hate programming, and cannot code their way out of a wet paper bag chatgpt didn't create this phenomenon, but it sure seems to have made people more enthusiastic about outing their own inability to grasp simple concepts
|
|
#
?
Nov 22, 2023 02:09
|
|
|
nothing wrong with fearing and hating programming, though
|
|
#
?
Nov 22, 2023 03:11
|
|
|
important to understand the thing you fear though
|
|
#
?
Nov 22, 2023 10:54
|
|
|
james_david 48 minutes ago | prev [–] The subtitle of the article is cats who fetch. I'm sorry to be a pedant, but language matters when we talk about other living beings, and I think it's important that we recognize them as such, rather than, say, a box that opens, or a Bluetooth speaker that ran out of battery. reply cnity 43 minutes ago | parent | prev [–] This isn't pedantic, it's just wrong. "Who" is for humans. reply james_david 35 minutes ago | root | parent | next [–] This is the point I am making. Human exceptionalism is built into our language and it is wrongheaded. Anyone who has known a cat will be aware that they are full of personality, very much unlike a box or a Bluetooth speaker. reply
|
|
#
?
Nov 24, 2023 17:47
|
|
|
if you eat a burger you're eating a person - you never thought about that did you???? luckily me and my giant brain are here to help you
|
|
#
?
Nov 24, 2023 17:56
|
|
|
josh04 posted:is altman even an engineer? if you asked me what the ceo of OpenAI did on a day-to-day basis i'd assume "schmooze, lazily tell people to work harder". 
|
|
#
?
Nov 24, 2023 18:16
|
|
|
Is "he's too much of a dweeb not to be an engineer?" what you're trying to say?
|
|
#
?
Nov 24, 2023 19:57
|
|
|
this is the ideal male body you may not like it but this is what peak performance looks like
|
|
#
?
Nov 24, 2023 20:00
|
|
|
he needs another collar layer imo
|
|
#
?
Nov 24, 2023 20:02
|
|
|
Mr.Radar posted:james_david 48 minutes ago | prev [–] 
|
|
#
?
Nov 24, 2023 20:40
|
|
|
24. Show HN: An AI-Generated Encyclopedia (mycyclopedia.co) 32 points by mahouk 7 hours ago | flag | hide | 27 comments mahouk 4 hours ago | prev | next [–] Update: Sorry guys, but it seems the server has crashed due to a sudden influx of traffic, and I'm attending a funeral service at the moment so I don't have access to my laptop. Will try to get the site back up asap! reply Edit: visarga 3 hours ago | prev [–] <rant>We need an AI-generated encyclopedia - not for us, but for AI. It should have a trillion articles covering all known entities and concepts, written using RAG over the web. Controversial topics should report the controversy or the distribution of opinions. We can put this big synthetic text corpus in the training set of future models. Why? Because AI needs long form, in-depth texts to train on, and the web doesn't provide it in sufficient quantity and quality. We need chain-of-thought to capture relations between concepts in explicit language. Synthetic data makes it possible to have balanced coverage of topics and combinatorial coverage of skills to improve reasoning. It's also better from a copyright stand point to train models on synthetic data.</> reply jfk13 2 hours ago | parent [–] > the web doesn't provide it in sufficient quantity and quality Do you seriously think that "an AI-generated encyclopedia" would provide a better-quality training set? What would the "AI generator's" articles be derived from? reply visarga 50 minutes ago | root | parent | next [–] So the way I see it, in the first stage the model can take all concepts in Wikipedia and other knowledge bases, and do web search, collect a bunch of references, study and compile a report. That's straight forward search + summarization. The advantage would be that models get to bring together information sitting in separate examples and synthesize or draw conclusions. The second stage would be to generate research questions, then solve them with LLM+web search+code execution+other tools. The results would be compiled in reports. So it's a loop of problem generation, problem solving and validation. You can validate with highly trusted sources, or you can run code or simulations, ensemble multiple attempts, or even leave it to ranking by a preference model. reply 
Mr.Radar fucked around with this message at 20:45 on Nov 24, 2023 |
|
#
?
Nov 24, 2023 20:41
|
|
|
visarga is one of the more deeply insane people on hacker news and that is saying something
|
|
#
?
Nov 24, 2023 21:03
|
|
|
|
| # ? Apr 28, 2024 22:03 |
|
|
dear evan hansen looking motherfucker
|
|
#
?
Nov 24, 2023 22:56
|
|
