Opus 23
A new way of understanding AI, and all technologies before
Subscribe for more: https://www.youtube.com/@opus23.1
Website to generate your own concept fission animations! https://concept-fission.com
Follow me:
Instagram: https://www.instagram.com/opus23.1
Twitter: https://x.com/opus23_1
Business inquiries: opus23n1@gmail.com
Chapters:
0:00 Intro
01:42 Part 1: Beyond Tool and Human
04:18 Part 2: Concept Fission
08:07 Part 3: The Social Bundle
17:52 Part 4: Selfless Intelligence
25:09 Part 5: Artificial Intelligence
35:05 End
Musics:
“Die Zauberflöte, K. 620: Dies Bildnis ist bezaubernd schön (Tamino)”
By: Wolfgang Amadeus Mozart
Performed By: John Dickie, Capella Istropolitana, Johannes Wildner
Courtesy of Naxos of America, Inc.
“Je crois entendre encore, caché sous les palmiers”
By: Georges Bizet
Performed By: Yasuharu Nakajima, Annick Massis, Teatro la Fenice Chorus, Teatro la Fenice Orchestra, Marcello Viotti
Courtesy of Naxos of America, Inc.
Source
Stumbled upon a masterpiece
twitter, twitterversity, university
a priori, Kant, a posteriori
Super vidéo 👌 👌 👌 👌 👌
🧙🏿♂️ THIS IS VERY BAD… I'm not convinced that my creativity is artificial…like is intuition artificial?
A star philosopher of modern times
I appreciate, among everything else this visual essay is offering us, SO MUCH how you followed Nakajimas subtle tempo rubato to make the technology split the corresponding two virtues on time. This attention to detail has deeply moved me. It's the slightly delayed inevitability of it all…
And to think this is your first public video…
Many good thoughts. It ironic that this script sounds like it's at least partially written by AI.
Kind of disappointing. But… I guess you attempted to give an answer, which is better than talking at length with no real point or purpose like 99% of youtube video essays or whatever you call these.
In my mind intelligence has two distinct aspects. The expansive aspect, which is the acquisition and application of knowledge, in which AI truly does have intelligence (although artificial). The other form of intelligence is connective intelligence, where previously distinct concepts can be joined together, the opposite of concept fission: concept fusion. For example, AI cannot create it’s own philosophical theory, because it does not have the connective intelligence to connect various concepts together to create something truly new. The same goes for art and many other mediums. Although the AI may have all of these concepts known, it cannot recognize new patterns without at least being taught how to recognize them, while humans can. And that I believe is why I do not consider AI truly intelligent, and also why I believe intelligence is not truly artificial, because while AI can recognize previously taught patterns, we can recognize entirely new ones using our previously learned expansive intelligence.
Reasoning has been disconnected from consciousness, but consciousness has not been disconnected from humans just yet, for that to happen world models like this principal ex Meta researcher talked about would have to come to fruition, and when that happens, consciousness will be "proven" disconnected (although it always was), and then animal rights would arise as a real convo, their rights and the rights of world model AIs
Grate video, but This is should be not a question,
I'm sad that they are naive people that doesn't know how "AI" works and thinks that "A"I has:
consciousness, intelligence or identity. Its only predict stuff based on stats.
I've had this same discussion with someone on discord, I basically said what you said but without the Concept Fission thing, and he tried to say "Trust me bro", on that AI intelligence is not the same as human intelligence and that it only seems that way because it is true, that the way we seemingly become intelligent over time, pretty much equals to how AI does it.
small mistake 29:32 you meant socrates ?
Great video btw
Yippee new video essayist
So the book nerds are the first who disconnected interacting with people and fulfilling the need of human interaction.
Since then we only improve on that with sex robots probably the final form.
Then humans don't need each other anymore, and our population will also completely collapse then.
35:08 Like the concept, but I’m skeptical! imo I can believe your point with these concerns addressed… well moreso two hangups really:
1. “Tech fission” is too much of an oversimplification. As often as technology can split two adjectives from each other, it also ends up fusing two. For example, cars effectively split time and travel but in the same breath fused travel and carbon emission. EVs then came to (try to) split those. These ‘fusions’ are just as critical and can muddy the outcomes of technology. Leaded gas, CFCs, and asbestos come to mind where technology fused very harmful ends to certain means. These harmful ends were so bad these technologies had to be undone. In this manner, if you are right, AI might have fused intelligence with climate change. I understand wanting to keep the concept simple but concerns about AI go beyond “is it what it says it is?” Many concerns are “is it destructive to the biosphere?” if it is, then is it worth pursuing?
2. I don’t believe AI splits humanity and intelligence. Why? Because I believe conviction is critical to intelligence. To explain my point, I want to admit that I believe businesses already split humanity and intelligence! Businesses legally maintain corporate personhood, meaning they are responsible for their own actions. We don’t argue about a business’s sentience, but we take it for granted that a business makes intelligent decisions. If a stupid decision is taken by a business, we begin searching for a secret reasoning. Maybe embezzlement or fraud! Perhaps they’re downsizing or got threatened. Even worse, maybe it was a decision bought from a consulting company (an industry of businesses which literally exist to split responsibility from intelligence lol). It’s never, “ah, more business slop”. From this perspective, what landmarks I believe would prove AI is intelligent is the legal responsibility for actions. Does AI have the conviction, knowledge, and real-world sensory capacity to make real-time decisions that it itself is responsible for? If you put AI at the front desk of a travel agency it might give fake discounts. If you put it in a drive-thru, it might take an order of 69 cups of water without flinching. The legal responsibility of applying AI in the real world is singlehandedly exposing the distrust in AI’s true intelligence. Once these thresholds are crossed, it’s like accepting as a society that AI truly can be intelligent. It just seems like practical application is decades behind in-house testing and benchmarks.
Thesis < YouTube video essay
🔥🔥🔥🔥🔥💯💯💯💯💯
4:09 Oh its the channel picture!
34:22 were those supposed to be "IA?"
Hate ai fear America
Damn first video? You really knocked it out of the park.
What a video to kick off the YouTube journey.
Only 3k subs? See you at a million soon.
Great video, but I disagree with your statement that humans' ability to quickly understand previously unseen systems/tasks is merely an ability given to them from their ancestors' experience of similar systems. We are still able to understand and predict the behaviors of unseen systems regardless of whether they have shown up in our evolutionary history or not. Take videogames, for example. There is a massive variety of games with all sorts of different presentations- from 2D puzzle games, to VR platforming games, and the speed at which we can become proficient at a new game is not predicted by how similar they are to what our ancestors experienced in their past lives. One critical aspect of intelligence which AI still lacks is the ability to predict the behavior of previously unseen systems given a small number of data points. Reinforcement learning is just not nearly as fast as animal learning- this is something that AI is still missing which nature has solved.
so weirdly good. so good i think i have acquired a new taste
I made a quick website where you can create your own concept fission animations! https://concept-fission.com. Very curious to see what you come up with, I'm sure there are so many I missed. Enjoy!
Random accounts with one videos creating a masterful video essay core
ChatGPT pretends to be ChatGPT
~Opus 23
This was a really great video! What was your inspiration for part 3 (the social bundle)? I thought that idea was really interesting idea and I'm really curious about similar ideas.
I'd never really considered that part of the importance so many people put on social interaction could be an effect of the fact that it has historically been so interwoven into otherwise unrelated things. Of course, there is an argument that the social aspect of those things is a positive 'side effect' of those things rather than an unfortunate byproduct. So I really want to take a deeper look into where that thought or similar ones came from!
16:18 literally laughing out loud over here! 👩🍳👌
this video has split me and my incomplete understanding of the deeper problems with AI thank you, may the algorithm carry it far beyond
I love how far you took the whole disrupting bubble animation thingy. It's very visually effective! I also appreciate how the overall presentation isn't too distracting. You obviously take a lot of inspiration from contrapoints and I'm excited to see how your visual identity will evolve from here on out. For what's apparently the channel's first video, the production quality is already crazy good!
Can I just say that language is DEFINITELY NOT only seen in humans and it’s been observed in many many other species
Interesting summary of the issue. I can now recognise my previous position had some cognitive dissonance. Shall "cognite" on this more! As a side note, I am somewhat disappointed that this only focused on LLMs and not the AI tech craze generally (especially stable diffusion and subsequent generation systems) but I am beyond impressed regardless.
Your view on AI is refreshing in this sea of binary views and defensive, small-minded people on both sides. Amazing video, beautifully executed!
Incredible work