Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
BlankSystemDaemon
Mar 13, 2009



PokeJoe posted:

nature has a naturally flared base
So what you're saying is, it's made for sticking up the butt?

Adbot
ADBOT LOVES YOU

BlankSystemDaemon
Mar 13, 2009



I don't think anyone's arguing that it's not okay to watch youtubers you like, but people are always going to be questioning you and your motives if you watch chud-lites or chuds.

Questioning viewing habits isn't a bad thing either, there's no reason that the background input we subject our noggins to shouldn't be subject to criticism.

BlankSystemDaemon
Mar 13, 2009



https://www.youtube.com/watch?v=5nKk_-Lvhzo

BlankSystemDaemon
Mar 13, 2009



I absolutely love this kind of thing, so it's neat to see a video on it.

It's just a shame he gets one little nit wrong, because Japanese has both stress/intonation and pitch accent.

BlankSystemDaemon
Mar 13, 2009



mystes posted:

Huh? AFAIK Japanese doesn't have stress in the sense of stress accent like in English and intonation is not relevant to what is being discussed.
Yeah, I was forgetting that there's a different between pitch accent (which Japanese has) and stress accent where pitch-accent is high/low and stress-accent is strong/weak.

There are commonalities between the two, though - because if someone asks a question in Japanese, there's the same rising intonation at the end as there would for a question asked in English.

BlankSystemDaemon
Mar 13, 2009



Jim Silly-Balls posted:

Whats interesting is what it got wrong and how it got the questions wrong. It often starts out with the right concept and starts doing the math and then just veers off in the wrong direction and cant be convinced that its doing it wrong, in fact insisting upon obviously wrong facts to justify its answers (like insisting that 40 is larger than 63, very basic math stuff that is incorrect)

There is also a pretty interesting theory that things like ChatGPT may eventually become like Wikipedia. In the beginning everyone mostly intrinsically trusted wikipedia because it was this new and gigantic repository of information the likes of which we had never seen. Then we realized that it wasnt always right, and was fallible. Now we treat it as something of a starting point or a jumping off point, but we generally accept that it is not 100% correct 100% of the time and everyone knows that.

He posits that at least in the near term, ChatGPT may go the same way
It's almost as if algorithmically combined and controlled text processing models don't really understand the underlying concepts and have only been trained to seem confident, and that that in turn means it can be confidently wrong.

BlankSystemDaemon
Mar 13, 2009



There's also this.

BlankSystemDaemon
Mar 13, 2009



Your guess is as good as mine.

BlankSystemDaemon
Mar 13, 2009



NoneMoreNegative posted:

I have been watching a bunch of the 'History of the Universe' channel longform videos: https://www.youtube.com/@HistoryoftheUniverse

https://www.youtube.com/watch?v=_IkaetPoBZM

I can't say I have a full understanding of some of the stuff like the most recent 'Theory of Everything' vidoe, but at least I feel I don't understand it better.

https://www.youtube.com/watch?v=E4CsY5B3BoI
Oh great, more of these channels where they'll take vaguely science-related topics and use it as an excuse to speculate on all sorts of things that amount to little more than religion.

The first video has several instances of the argument from analogy fallacy whereby the analogy, for example when they're attempting to use doesn't satisfactorily explain the differences between an oblate spheroid and whether or not space-time is positively, negatively curved, or basically any other time they use any analogy to try and explain something which is, given our current understanding of the universe, impossible to prove.
And if we'd arrived at a theory of everything, I'm pretty loving sure a youtube channel wouldn't be the place for that news to be breaking, because that'd basically be a revolutionary discovery given the limits of our current understanding - so the answer to the second video is "no, period".

I don't understand why people are so loving obsessed about this, when there's so much stuff and things that can be evidentially and experimentally proven that one person can spend their entire life on it and still not make any amount of headway.
And if the point is to point out how little we understand, science already does a great job of this - because everything about the scientific method has shown that the more facts we have, the more questions we can ask.

BlankSystemDaemon fucked around with this message at 20:54 on Feb 16, 2023

BlankSystemDaemon
Mar 13, 2009



Put another way, if you wanna be like "Whoa, dude", you can just put on some music with a Winamp visualization, and eat an edible with an appropriate amount of marijuana extract.

BlankSystemDaemon
Mar 13, 2009



Mr.Radar posted:

https://www.youtube.com/watch?v=Ib6sDW6mRVw

It's a Short but it's pretty informative.
the best way to get the right answer on the internet is to confidently state something wrong

BlankSystemDaemon
Mar 13, 2009



Carbon dioxide posted:

https://www.youtube.com/watch?v=WO2X3oZEJOA

A vid about GPT/language models.


TLDW: language models use tokens. Tokens are either common words or parts of words or individual characters. That way, rare words can be represented by characters, but common words/sentences take up less space. The choice of what words are tokens is simply based on how common those words were in the INITIAL data set.

There are some tokens that are really weird such as PsyNetMessage and SolidGoldMagikarp.
If you use those in a message, GPT-3 completely glitches out and returns nonsensical crap (not in ChatGPT, that has been patched since this was discovered).

Why? Well, what happened is PsyNetMessage is a term that appears very often in some game's debug log that somehow made it into the initial data set, and SolidGoldMagikarp is the name of a reddit user who was extremely active in r/counting... where people just count up numbers. Whatever web scraper they used for the initial dataset picked those up as really common.

Then, when they actually started training the model, the researchers cut out those weird terms and used more sensible data. However, the tokens already existed at that point. Result: the tokens exist but there's basically no information associated to them. The language model has no idea what they "mean" or how to respond to them. This causes the glitchy behaviour.
Along these lines:
https://www.youtube.com/watch?v=wuU-5rGPbyg&t=210s

BlankSystemDaemon
Mar 13, 2009



AtomD posted:

i enjoyed this talk
seeing one of the maddest youtube commenters in existence was also a real treat
Yeah, it's something, alright.
Or something alt-right.

Could go either way.

BlankSystemDaemon
Mar 13, 2009



namlosh posted:

this might be the right thread to ask
to any Physicists out there who understand relativity, space time and gravity:
this video calls out a bunch of other YouTube videos about the cause of gravity. I’m not sure I understand it all but is it correct?

https://youtu.be/PjT85AxTmI0

I’d really like to know
It's "fine" to say that other YouTube videos of understanding of gravity is lacking, because unless there's been a pretty radical change in the scientific community - which we would almost certainly have heard about, given the big impact it'd have had - it's pretty fair to say that nobody understands gravity.

We don't know how/if gravity propagates like the other fundamental forces, and we can't fully integrate it into our existing models.
Scientists ended up building the worlds largest beam cannon just to attempt to find this out at CERN, and so far we're still pretty much in the dark.

So really, while someone can make an rear end out of themselves by claiming other YouTubers don't understand it, that someone also doesn't understand it.

BlankSystemDaemon
Mar 13, 2009



namlosh posted:

thanks for this, I’ll take a look for sure

this is what I thought… I (somewhat) understand the concept of action at a distance and how gravity doesn’t fit into any current unification theory.

I’ve read “In search of Schrödinger’s Cat” and “the God Particle”, lol

recently I watched a couple of those other vids and was surprised they seemed able to explain it. their argument seems to boil down to: Gravity is caused by differing time dilation with respect to the observer

the video I linked says that’s not it, and fair enough.

I just thought maybe our understanding had changed and I hadn’t been notified in accordance with the physics phone tree we all agreed to
I was just pointing out that the YouTuber who posted the video also has no loving idea, despite the implications of the clickbait headline and scrapbook-looking preview image.
That he then goes on to handwave a bunch of explanations, which don't account for propagation and our inability to integrate gravity with the standard model of particle physics as well as how none of what he's talking relates to quantum mechanics, is just about par for the course.

I'm not sure about the best way to go about science education, but I'm sure taking a page out of tabloids is a good solution.

Symbolic Butt posted:

I gotta watch the video until the end but the way I see it is not about questioning youtubers' understanding of gravity, it's about questioning their fluency in applying the general relativity theory.

I think the idea of time difference creating some kinda torque that influences the space dimension is absolutely mind blowing. So wild that I'm inclined to believe it's just popscience people not understanding GR really.

But again, I'm not a physicist. I'll watch his argument more carefully.
If you wanna question fluency in terms of scientific understanding, you first gotta address all the lies to children that fill most peoples imaginations; the notion that atoms somehow look like solar systems, the definitions of acids and bases and how it relies on water being part of the solution (pun fully intended), and things of that nature.

Sure these things help in establishing a basis, but all too many people are stuck with these notions.
Even people who've been piled higher and deeper in a narrow subject, can harbor these lies to children in areas they're not expert in.

Achmed Jones posted:

one of the most frustrating things in the world is the lovely stemlord sophomore who doesn't understand how science works

this is not in direct response to anyone above. the discussion just reminded me of some of those types from the d&d subforum and, less common, when i was teaching. when i was teaching you could set em right pretty easy but the internet ones were _woof_
I'm absolutely for leaving the science education to the people who're qualified for the job; I realized a long time ago that I wasn't cut out for it.

The best thing about science is still that for all the answers we get, we get even more questions.

mystes posted:

Imagine spending 4 years slogging through all of that to discover that it's all 100% pointless/obsolete knowledge
Nah, once the four years are over, you're now one of the worlds foremost experts in a field so narrow, unless you choose to work within that field, it's useless.

If you're lucky, you'll have learned critical thinking and other things along the way, but that's absolutely not a given as it typically isn't part of a curriculum.

BlankSystemDaemon
Mar 13, 2009



mystes posted:

Is this supposed to be some sort of joke that it's no different than grad school?
No?

The "Nah" seems kinda dismissive on a second read, but it wasn't meant that way.

BlankSystemDaemon
Mar 13, 2009



MrQueasy posted:

Lost a bunch of productivity this morning to mild disassociation resulting from this lady's Youtube Shorts about 3d actors influencing a 2d universe and its implications for 4d on 3d space.

https://www.youtube.com/@tibees
Huh, that's a really interesting idea.
I've no clue how you'd go about making a hypothesis for it or even testing it, but it's interesting!

Could make for a fascinating science fiction book, too.

BlankSystemDaemon
Mar 13, 2009



MrQueasy posted:

I don't think she was claiming that 4d entities exist, thankfully. Just a mathematical explanation of how changing the number of dimensions makes everything odd.
Yeah, I get that - it's just that I don't recall having heard of the idea before.

Corla Plankun posted:

There's actually a really easy test! A four dimensional actor can rotate a left-handed glove and cause it to become right-handed in our dimension :2bong:
:hmmno:

BlankSystemDaemon
Mar 13, 2009



Achmed Jones posted:

how much of it is actually even meaningful and how much of it is pure cargo cult a la "unalive" on tiktok and similar?

i really wonder if the dumb thumbnail thing is just "your videos do better when you make a thumbnail. any thumbnail that doesn't suck is fine." but people misunderstood why it worked
I'm sure it's a good mix, but that's part of the problem.

The corporations with the most profit margin of the world are deliberately leaving it unsaid what the exact rules are, because that way they have more leeway in tweaking the algorithm any time the corporations they need to satisfy complain.

BlankSystemDaemon
Mar 13, 2009



It's low-key amusing that a video with a guy talking about saturation has audio that's all over the place.
Some segments are clearly recorded in post and set over B-roll, and it's pretty jarring.

Also, saturation is big part of the loudness war that's been destroying music for decades.

BlankSystemDaemon
Mar 13, 2009



spankmeister posted:

this is why vinyl often sounds better because you can't mix as loud on vinyl
There's a real argument to be made that Phil Spectres use of instrument doubling and echo chambers, which was nominally done to make transistor radios and vinyl sound better, was the start of the loudness war.

At least one paper also makes the observation that it was already happening during the 60s' Motown era, and it also mentions that the term originates from 1979 (a little more than 30 years before it got popularized).

BlankSystemDaemon
Mar 13, 2009



namlosh posted:

anyway, have an educational YouTube I found recently:

https://www.youtube.com/channel/UC8i-4EpLBXMzekETNXwcZjA/videos
Seems like a real neat channel, thanks for the recommendation!

BlankSystemDaemon
Mar 13, 2009



Tom7 is very educational:
https://www.youtube.com/watch?v=Ae9EKCyI1xU

BlankSystemDaemon
Mar 13, 2009



Carbon dioxide posted:

It started out interesting. Then it went on an interesting tangent. Then it went on a tangent that made by head explode several times.

Would highly recommend.
It's becoming a real thing, that when I watch a Tom7 video, I start out having a pretty good clue about what he's talking about, then get to a point where I think maybe I know what he's saying, and by the end I'm just blankly staring at the screen, looking at all the tabs that've just exploded open in Firefox.

BlankSystemDaemon
Mar 13, 2009



Salt Fish posted:

He proves its turing complete and then writes an emulator that plays mario with it.
And it runs at eye-watering speeds.

BlankSystemDaemon
Mar 13, 2009



It seems like what she's tired of is the "lies to children" that permeate all of science? Which is very fair.

Still better at BoI than me, though.

BlankSystemDaemon
Mar 13, 2009



MrQueasy posted:

also, she has a phd, so she's better at physics than me too.
:same:

BlankSystemDaemon
Mar 13, 2009



LanceHunter posted:

On the one hand, this youtuber's fashion sense is the visual equivalent of nails on chalkboard for me. (The number of rings on his fingers and bullshit on his wrists/belt/etc is absolutely comical.) On the other hand, getting a look at Karahan Tepe is pretty cool and it's actual educational content and not conspiracy theory/ancient aliens bullshit.

https://www.youtube.com/watch?v=8EaKFKYPXVk
Let people be in charge of their own presentation, and stop trying to make the world dull and boring.

BlankSystemDaemon
Mar 13, 2009



Here's some α, β and γ:
https://www.youtube.com/watch?v=-qWQaX8_EDY

BlankSystemDaemon
Mar 13, 2009



echinopsis posted:

in the video the dude says it’s to stop him dying
It will do that.

It will also hurt. A lot.

BlankSystemDaemon
Mar 13, 2009



distortion park posted:

In the middle of this when he talks about the level of radiation I would have preferred it if he had talked about actual health limits rather than comparing to radiation levels he had measured elsewhere. I had to go look up this https://www.theguardian.com/news/datablog/2011/mar/15/radiation-exposure-levels-guide to find out what the radiation levels actually were in relation to
Intensity Modulated Radiation Therapy, which is done by heating up wolfram until it gives off electrons and accelerating them then shaping them into a beam, is used for cancer therapies and is supposedly around 2000mSv.
It's typically always listed as 2Gy/day, because the entire point of IMRT is that you can control the dose via beam forming and things like Bragg''s peak to only be maximized in the area where it'll do the most good.

In total, 30 days of IMRT will be 60Gy, which if received in the entire body in an instant, would be fatal within 24 hours.

BlankSystemDaemon
Mar 13, 2009



spankmeister posted:

2000mSv as in milli not micro? as in 2 whole Sievert? :pwn:
Yup.

The way it works is that they have a beam-forming mesh on the end of the accelerator which is then spun around the patient, so that only the place where all the beams cross gets the full dose.
This area is calculated using a mix of PET-CT and MR scans, and are done by a medical professional with a specialty in particle physics using a computer.

During their break, I got to talk a bit with the person who does the work at Aarhus University while I was in cancer therapy. It was absolutely fascinating.

BlankSystemDaemon fucked around with this message at 15:37 on Sep 8, 2023

BlankSystemDaemon
Mar 13, 2009



The idea is to kill the tumor, before it kills you.

Cybernetic Vermin posted:

the spectrum we're working with:

you die ---------- you're cured ---- you die
For a lot of cancers, the best you can hope for is remission, that stays that way.

BlankSystemDaemon
Mar 13, 2009



gently caress! That's the exact aesthetic I want for my HiFi stack.

Currently I have a Pioneer A-333 amp, GR-555 graphic equalizer, PD-M400 multi-CD player, CT-333 casette player, and a F-223L FM/AM radio tuner:

Only registered members can see post attachments!

BlankSystemDaemon
Mar 13, 2009



https://www.youtube.com/watch?v=dFIqNZ8VbRY

BlankSystemDaemon
Mar 13, 2009



Oysters Autobio posted:

I dunno where to pose this question but here but are there any decent tech / programming / data podcasts that aren't just poorly veiled advertising for whatever VC unicorn hunters?
2.5 admins is decent.

Adbot
ADBOT LOVES YOU

BlankSystemDaemon
Mar 13, 2009



i genuinely thought veritasium would've gone belly-up with the exposé that was done on them a while ago

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply