What do you like in science communication?

I'm working on making a channel about AI alignment research; and lainchan seems like a good proxy for the types of people who I'd like to reach with it. So, what style of science communication channel do you like? Questions especially on my mind are; - "One person ranting at a camera a la Tom Scott" vs "Dialogue between a couple of people a la Computerphile" - "Highly supported by visuals a la Artem Kirsanov" vs "Minimalist set-up, just getting to the point" - "Supported by props or physical locations, a la Technology Connections" vs "Simple diagrams that express the point more concisely"

i guess simple, friendly, straight to the point type of communication is good for application (the how) . visuals, ranting and props are great for explaining the why (theory)

to be honest I dont much watch videos like this at all. Id rather have the stuff in a blog with plots and graphs when they're reasonable. maybe as for blogs about this you're going to have a tough sell; too many people blog about this already even if not dedicatedly. IDK about the youtube situation maybe you can find a niche, but if I were to want to watch a video about this, I'd rather one that keeps the flashy stuff (and frankly, the visual stuff) to a minimum so I can just listen to it like a radio programme while I'm doing something else that doesn't require much thought. Can be a dialog or a monologue, but I think the former is mostly useful if you have someone interesting to talk to, and not so much for conveying background. you asked for my opinion so lol there it is

I enjoy when black science guy condescends to me and enjoys the sound of his own voice. But that aside, your question is answered by what you want to do and what audience you are trying to reach. Personally I watch/listen to the whole spectrum. If someone can convey the basics of a complex topic to me that I have little knowledge in that's great! If someone just uses his masters in engineering to construct the dumbest thing then I'm up for that too thus I can not give you any concrete suggestions beyond doing what you chose as good as possible. The greatest factor to success it seems is quality and determination. This is why people watch Tom Scott, Mark Rober who worked on the Mars Rover and Vsauce Here who weirds me out as much as gives the occasional food for thought. Also pick a memorable name and don't go full Thunderfoot.

>So, what style of science communication channel do you like? I really enjoy @JoshsHandle presentations. His animations are wonderful and a great visuals for demonstration purposes, but they way he conveys his ideas and approaches his topics are the more important than the 3D animations. They are friendly to anyone who just happens to enjoy the video itself and the eye candies, while still being informative enough for someone that is interested in the topic to get a basic yet somewhat in-depth look into the subject.

>>11566 I agree with >>11569 in that the best method of communication depends on who you're trying to reach. Anecdotally, something with high production like 3Blue1Brown tends to be popular with good STEM students, while something Numberphile tends to be more popular with curious non-STEM students. I'd imagine this board has plenty in both category. If you're goal is actual understanding of the viewer, then something more clear and concise like 3Blue1Brown would be a good inspiration.

Most media is overproduced and made for stupid people apparently. Like in your screenshot, bottom right. Tom Scott shown with the progression behind him. Wow. Except I don't need that. I already have an internal mental model of the development of technology over time. I don't need these visual cues and mental anchorings. They're invoked automatically, in ways that I designed. Probably because I'm not an NPC. Top right, obfuscating and overproduced. Also misleading. When you stain neurons you have to use special process to really view the finer structure of dendritic arborization and see how mature spines form, stumps form, etc. The actual structure and layout of neurons also differs a lot per region, even in the layers of the hippocampus. So the graphic is useless and overproduced. Top left, computerphile, is somewhat overproduced. Very norm!e, again, because they're like toddlers who really respond to seeing the same face over and over. They have been kept in that state. Remember when everyone was using that dumb phrase "then I do my happy dance" like they're just a puppet in The Sims? That's what I imagine going on in the average viewer's mind with this type of repetitive personality and socially anchored content. Very "we", very reddit. Bottom left I don't recognize. 3blue1brown maybe would be my guess. I don't know, can be good. I do a lot of symbolism, gematria, numerology, etc stuff, but at the core and not really a symbol manipulator type. Nor a "cracker" type either that embraces the rules but then looks for exploits. I just don't get involved or internalize anything to begin with. So that sort of equational thinking doesn't come naturallly. I get the gist of it though. For me personally, the thing with math is that desire overrides it. Like yes, within a sentence or two you can plot out a tree and predict the one or two highly probable branches, and you're almost never wrong. But that makes it pointless, so I try to force others to be something they aren't. Math types don't have this impulse, they passively accept and are gratified solely by the model working and the flexibnility afforded by tweaking its variables (ie the interelation and feedback loops of its parts). I just don't like it. I think it's ugly. Earth is ugly. The human body is ugly. Math to me is just proof of ugliness. Therefore the mind rejects, because I would have to kms and escape the matrix otherwise. What do I like in science? Nothing. Science is ugly and produces ugly things and ugly people, overall. This is the rule. The exception does not disprove the rule. What do I like in science communication? Nothing really. It's ugly. It's infantalized. Watch something like the Freemasonically produced "donald duck in mathemagics land" from the 50's, in fact watch anything fromt he 50's, and see how inferior and watered down everything produced today really is. Many or most kids in the 50's and 60's were given a crystal radio set, and produced one. It was trivial. Something your 8 year old did or whatever. Same with model rockets and such. Today this would be a college course. It's all retarded gibberish made by morons, and it's designed to pollute your mind and artificially slow you down, because they don't want you actually being able todo or understand anything.

of those four screenshots the only one that doesn't actually bother me is the upper left. The graphs just look goophy. Eyes on your pi's ? what why do you even have pi's there ? give me a break. Top right might be OK but its not complete by itself and looks like it could be simpler and more complete with less work if it wasn't done by someone who cares more about graphics than explanation. Bottom right guy looks crazy. I don't want a presenter to be looking right at me through my screen. If you have to show a presenter's face, do it in the lazy way, and don't look right at the camera gosh, have the decency to look at your shoes at least. Top left obviously has no real info in it but at least the still image looks relatively normal and inoffensive. Again though, no need to put presenter's face on screen at all. Ideally just show some guys workshop with him doing his work on your topic. I guess in AI research wankery, showing how the sausage gets maid might get you derezzed from youtube though so idk what to tell you.

>>11591 See? It's even 27 minutes long, the 3rd cube number. 3x3x3. As if it wasn't obvious already, given that animation in general has always been occult and often Masonic. https://en.wikipedia.org/wiki/Donald_in_Mathmagic_Land?useskin=vector https://archive.org/details/donald-in-mathmagic-land-1959

>>11593 Might as well include the rest. ................................................ ................................................. .................................................

>>11594 Forgot to to include. 25 is 5 squared, 5x5, 55. 5 is the number of sensory reality and illusion (can also be "spirit", electricity, or aetheric hidden mover, in elemental systems). So you + I makes 5x5. 5 senses. Two people. 5 and 5. All kinds of stuff like this.

>>11595 Wasn't quite done yet actually. Lot of stuff I just noticed, pretty incredible...... ............. ...........................

>>11595 You sure have some overexcited neurons. As someone who is admittedly into numerology what do you think of the duodecimal/dozental or sexal/hexal numeric system? Like yeah, it's always a ride to see some dude add up a bunch of random things and see cool magic numbers come out at the end but every numerology guy I see schizoing out does so in decimal. But I wonder if it doesn't introduce an unnatural "shift" to the results. Because what adds up to 666 in decimal is not going to have the same digits in an other base. https://www.youtube.com/watch?v=mSJxhQ2RHpQ https://www.youtube.com/watch?v=qID2B4MK7Y0

>>11597 People have tried to determine what base the universe "thinks" in, or whether it simply watches or is a byproduct of our mental activity and therefore imitates / makes use of whatever calendar or numeral system we do. Or whether it's layered in such a way that it has a native language that it uses internally and layers of imitation over it used for specific tasks. There's the dimensional notions, like you have shadowgrams or 2D projections of 3D objects, and likewise with 4D, so some numerical patterns are simply non-linear time / retrocausality ie "4D" protrusions into 3D space. You see this in sacred geometry thought for example. The other is algorithmic or generative, like Yggdrasil, spreading its fractal roots across the Earth. Or with the Hebrews, the tree of life comes up with an idea, drops the fruit into the chamber of guf, and it is manifested into the world. Which is why things come into being under the same numbers or numerical pattern, they grew from the same fruit and are connected by the same roots, even if they themselves have no direct correspondence due to that number alone (like resonance, oscillator sync, the other model for correspondence which basically amounts to numerical sympathetic magic). Kabbalist also look at words as holes in the veil where patterns of light shine through. See eg Stan Tenen's work on that and 4D torus knots or whatever it was. He arrived at a base three coordinate system (geometric) encoding as the actual basis and function of Hebrew, rather than a numerical one. Kabbalists use base 10 because we have 10 (or the phyllanges in some cases, like for mapping genesis onto the hands in base 12) fingers and are made in the image of God therefore God must think in 10s. Rudolf Steiner eg mocked the Masonic 33 saying that it's based on a faulty assumption that the universe thinks in base 10, I recall him saying something like it's probably ternary and that numbers are reversed in the astral. It takes 33 years for the lunar and solar calendars to align. 156 refers to the eclipse cycle across all active saros series, 1 5 or 6 synodic months apart. 156. The person is composed of 4 parts. 33% body, 33% mind, 33% soul, and 1% the I. Or the physical, aetheric, and astral respectively. So 66 or the beast represents the physical world. The earth itself. 6 is the first perfect number. 6, 66, 666, and 616, along with a 345 triangle, basically encodes the entire solar system. Sun, moon, earth, and stars. 666 is "the beast". 33 is therefore the spirit over matter, the hidden mover. It really depends on what system you're looking at. Anything you come up with the universe will grab onto, copy, and start using itself. Or it will undergo some sort of diffusion, as in "morphic resonance" and related ideas. Demonstrable and totally true. Lastly it could be that the universe's thought isn't numerical at all, rather numbers are a byproduct and the way it taps into and reads our body map (ie cerebellum) and thoughts is represented numerically as a byproduct. Or it has to work with it in numbers, despite not being numerical itself.

>>11594 44 is a very key number to evangelion and eclipses. Used as a kill number generally. "kill" = 44, "moon shadow" = 44, "nude" = 44. Arianne Cleopatra Schreiber would have turned 44 on the day of the 4/8 eclipse. She sings for Lilith in Komm Suser Todd, she is therefore "Lilith's mask". All of the characters in Evangelion have birthdays matching their voice actresses, therefore 4/8 would be Lilith's "birthday". The next eclipse over the US is in 2044. The End of Evangelion depicts and eclipse and a moon shadow. The End of Evangleion was shown for the first time in North America, in theaters, on 3/17/2024. 317, 66th prime. Shortly before the 4/8/2024 eclipse. You may not like it. But this is what peak performance looks like.

>>11590 >Anecdotally, something with high production like 3Blue1Brown tends to be popular with good STEM students, while something Numberphile tends to be more popular with curious non-STEM students. This is an interesting take. I always thought of the reverse. Except I can see math students also being interested in 3Blue1Brown. Maybe they lie in a middle ground between these two kinds of people.

>>11590 I know several mathematics students, mostly good students, who got interested in math because of numberphile videos. 3B1B is popular across the board but typically most popular among STE students who are otherwise fearful of mathematics. In general I think high production value is mostly good/necessary when your audience has little knowledge and no confidence. The more knowledgeable/confident your audience is, the less they need a tight script and good visuals to get your point. In other words, a knowledgeable or eager audience wants the most direct communication possible, while an audience with no knowledge or little prior interest needs the flashy visuals and highly crafted metaphors or they'll get intimidated and switch it off. It's a continuum, obviously, but that is the general trend I think.

>>11601 > while an audience with no knowledge or little prior interest needs the flashy visuals and highly crafted metaphors or they'll get intimidated and switch it off. These are useless people. Anything the sponge up will never amount to anything. They're not good transmission vectors. You're just indoctrinating them intoa religion, nothing more. Don't pretend otherwise.

>>11602 If you consider them useless then don't make videos for them. >You're just indoctrinating them intoa religion, nothing more. I'm not doing soykaf. Anyway, just because something is a belief system doesn't mean it is a religion and just because it involves claims of truth doesn't mean it is indoctrination.

>>11603 >Anyway, just because something is a belief system doesn't mean it is a religion and just because it involves claims of truth doesn't mean it is indoctrination. Wrong. And I'm holding you personally accountable.

>>11600 >>11601 >I always thought of the reverse >I know several mathematics students, mostly good students, who got interested in math because of numberphile videos That's interesting. I suppose they also cover very different types of content. 3B1B used to focus more on the type on math you'd learn about in school fairly early on, but he's since moved onto just whatever he's finding interesting at this time. I guess that focus on the type of thing you'd learn in school would make him popular with bad students as well, whereas Numberphile focuses on more 'quirky' math, going over random proofs and things, Mandlebrot, Golden Ratio, etc. which would make him more popular amongst math students in particular. My comment was mostly based on professors who recommended 3B1B and other students I know who mentioned him as well.

>Bad ending! You trusted...... >The Science (tm)

https://www.youtube.com/@ImprobableMatter this is probably the ideal. Decent but not overdone visuals, not assuming the viewer is stupid, clearly speaking and having good focus. >duhhh science is a religion scientism != science

The Real Reason Why You Have Allergies Kurzgesagt – In a Nutshell https://youtu.be/9zCH37330f8?t=30

I like Veritasium's narrative style https://www.youtube.com/watch?v=88bMVbx1dzM CGP Grey is fine too https://www.youtube.com/watch?v=pUF5esTscZI

Is THORIUM the Future of Nuclear Power? Kyle Hill https://www.youtube.com/watch?v=FjHH8Qf3aO4

True Facts: How Camels Conquered the Desert Ze Frank https://www.youtube.com/watch?v=Sc0ZCqLUjSo This narration style is to be fryly funny

The only way I could do that was if you had to do a little more work and I would be happy with it but you

>>11566 So tired of "AI alignment" as a topic. AI is nowhere near as advanced as most people seem to think. All AI alignment "researchers" do is argue that someday AI could become dangerous, therefore give me money to run experiments or something. It's so inane to me. AI researchers should come up with mathematical and falsifiable models first, BEFORE going online to fearmonger and beg for funding. Take a page out of what cryptographers have been doing with advantage.

True Facts: Where Have All The Fireflies Gone! Ze Frank https://www.youtube.com/watch?v=czt_io0h-CA Save the lightning bugs!

>>12368 I don't see what sort of mathematical model you would need to prove something more to you? I presume you mean proof that AI is dangerous, not that AI could be strong? The main thesis if when they break it down is just that reward hacking (which always happens) will be dangerous in any real-life scenario involving an agent-like AI. And yes, there will be people who make agent-like AI- there already are, using RAG and transformers. We don't have a good way to specify morality for an AI, and we are only getting less capable of implanting it the more powerful and complicated AI becomes. The best argument against AI alignment researchers is that they clearly aren't going to fix the problem. OpenAI had just about the worst charity rating ever, because it was obvious from the beginning that their mode of operation was antithetical to their stated goals. Now they've shifted the AI timelines forward- if the threat they spoke about is real, that means they've been the main existential threat to humanity while larping as the opposite. They aren't acting in good faith, I have to consider them psychopaths or idiots by an objective assessment in their own worldview. Even if we had inner and outer alignment problems solved, that would merely elevate us from being slaughtered and tortured by digital psychopaths to the old fashioned biological psychopath. Forget AGI, there isn't any stable long-lived world where over a 100 people have access to world-ending capabilities, like narrow AI in any kind of biowarfare domain. The vulnerable world hypothesis is probably true, and we need a completely different civilization if we want our descendants to colonize the stars.

I want a S0yence Chan that sh1ts on AI, quantum and gematria 24/7, comparing them to New Age and the like. I want it to provide me with complex permutations of fermented dregs, set to British triphop from the 90's.

Is "Kurzgesagt – In a Nutshell" acceptable to you? https://www.youtube.com/watch?v=KNwMiydCYA4

>>12368 They're a smokescreen for the fact that "AI" is an inherently anti-human technology that doesn't need to be self-aware or intelligent or whatever to cause harm. Essentially they are there so that the AI corps can say they have ethics departments.