Was skeptical when I clicked on the video (there's lots of new ai generated science/ai channels out there these days), but this was a quality video with a unique editing style. Liked, commented and subscribed for maximum engagement. Wish you luck growing your channel bro.
You, whomever you are, discovered this channel through youtube recommendations, I've been subscribed to your channel a day or two now, and I'm proud to be amoung your 1st viewers. Already put out the word you're a channel to keep an eye on. Your content is solid. The illustrations are simple yet dynamic. I think this channel is going to be huge. It's interesting that you've decided to keep your videos to 5-minute digestible lengths and not any longer. I guess it takes a lot of work what you're doing, the video is great quality. I was wondering how you were doing it all alone, but I heard you mention "we" earlier in the video regarding the video you're working on about Steve Jobs, it makes sense, team work makes the dream work.
I agree, this is big. Using Claude or GPT-4 with long prompt causes the LLM to forget or skip instructions. If this long context window can eliminate this issue, it opens a world of opportunities to sort through legal documents, books, research articles, as well as transcripts.
Finding channels like this makes me appreciate the TH-cam algorithm for once. Makes me look forward to when AI can do the same with all kinds of media while being optimized for usefulness.
How do you actually insert say Moby Dick into the context window? Do you have to first acquire a copy of the book in digital form, then right click, copy, paste ALL of the book into the Ai window on your screen? Or do you just type the name of the book, in this case, Moby Dick by Herman Melville, into the window? Any help or insights on this really elementary question is much appreciated. Good video and insights on reading. Maybe I should start reading again like it's a whole new learning experience, see how it effects my thinking. I think AI can make us smarter IF we interplay with it and challenge our own limited knowledge base. Thanks 😊
Thanks for watching! A lot of different way to do it. The primary way would be obtaining a pdf copy of a book and drag it into the chat interface or interact with it through the LLM APIs.
I have tested ChatGPT and Gemini repeatedly on logical reasoning and making connections that are outside of the standard narrative for great works and great historical authors. Both of them fail dramatically, even when given a small localized text. They cannot see what is literally there in the text. I mean LITERALLY THERE in the language of the text. We're not talking about metaphors or similes or figurative language. These models are morons. They have an extended version of programmers' type of intelligence and the deficiencies that go with them, namely all the social elements necessary to being a human and perceiving what humans mean. I'm talking of programmers, and by extension their robots (aka LLM's). Read robot as a derogatory word such as the N-word or "white trash". ROBOT.
Fair opinion. But the ability to synthesize all information and context from multiple books in a single prompt is a pretty substantial breakthrough. Regardless, thanks for watching.
If artist are encouraged to use references why do you think some hate when AI LLM does? Is it fear, human desire to control, or a misunderstanding of what references are used for? Could this be a issue that stemmed from putting worded definitions to everything including emotions, symptoms in the medical field etc? If your doctor can learn to diagnose you from reading a book of symptoms which mostly goes misdiagnosed why can’t a robot? It’s very interesting the jobs we complain about we don’t want to lose to ai.
Really great points. Seeing it as a tool to improve gets lost in the AI fear narrative for sure. We should be smart about its advancement, but open to its benefits.
![[#2024MAMA] G-DRAGON - 무제(Untitled, 2014)+POWER+HOME SWEET HOME+뱅뱅뱅+FANTASTIC BABY | Mnet 241123 방송](http://i.ytimg.com/vi/Ox29z5Nf1Uk/mqdefault.jpg)
"ChatGPT read this book for me. ChatGPT what did I think of this book?"
Is that you, Donald?
This channel has the highest ratio of quality to subscriber count I’ve ever seen. Subscribed. Keep up the great work please!
That’s a ratio we love. Thanks for watching!
Was skeptical when I clicked on the video (there's lots of new ai generated science/ai channels out there these days), but this was a quality video with a unique editing style. Liked, commented and subscribed for maximum engagement. Wish you luck growing your channel bro.
Wow. Thanks! Means a lot and hope we can do things different and better.
This deserves so much coverage. Your channel is so high quality! Great job. This concept is beautiful! Thank you for sharing!
Thanks! Appreciate it.
You can literally have a conversation with one of the characters if you just ask the AI to assume the role of that character...
You, whomever you are, discovered this channel through youtube recommendations, I've been subscribed to your channel a day or two now, and I'm proud to be amoung your 1st viewers. Already put out the word you're a channel to keep an eye on.
Your content is solid. The illustrations are simple yet dynamic. I think this channel is going to be huge.
It's interesting that you've decided to keep your videos to 5-minute digestible lengths and not any longer. I guess it takes a lot of work what you're doing, the video is great quality.
I was wondering how you were doing it all alone, but I heard you mention "we" earlier in the video regarding the video you're working on about Steve Jobs, it makes sense, team work makes the dream work.
Thanks for watching and joining the channel! So much more to come.
I agree, this is big. Using Claude or GPT-4 with long prompt causes the LLM to forget or skip instructions. If this long context window can eliminate this issue, it opens a world of opportunities to sort through legal documents, books, research articles, as well as transcripts.
For sure. Thanks for adding!
Finding channels like this makes me appreciate the TH-cam algorithm for once. Makes me look forward to when AI can do the same with all kinds of media while being optimized for usefulness.
377 sub, and my god is the editing style amazing and your soothing voice is the perfect cherry on top
Wow thanks! Welcome aboard and looking forward to many more videos.
@@Synapse.Official So am I
given your channel is about ai, how much/many ai tools do you use in your video making process?
Good question. Very minimal. Any image of a person gets a charcoal AI effect we made. The rest is good old fashion motion graphics editing.
@@Synapse.Official alright, was just curious about that
@@Synapse.Official I don't mind you used ai to make the entire vid. As long as it's of good quality.
Very interesting style
Thanks for watching
great video. keep them coming. this channel definitely has potential.
Appreciate it! Much more coming.
Interesting... interesting... but what I'd really like to know is:
HOW did you know my Ikea bookshelf is deteriorating?!?
Lucky guess. The odds were in my favor.
How do you actually insert say Moby Dick into the context window? Do you have to first acquire a copy of the book in digital form, then right click, copy, paste ALL of the book into the Ai window on your screen? Or do you just type the name of the book, in this case, Moby Dick by Herman Melville, into the window?
Any help or insights on this really elementary question is much appreciated.
Good video and insights on reading. Maybe I should start reading again like it's a whole new learning experience, see how it effects my thinking.
I think AI can make us smarter IF we interplay with it and challenge our own limited knowledge base.
Thanks 😊
Thanks for watching! A lot of different way to do it. The primary way would be obtaining a pdf copy of a book and drag it into the chat interface or interact with it through the LLM APIs.
Nice
This is amazing! what do you use to animate the visuals in these videos?
Just good old adobe. Thanks for watching!
how do you make these animations sir
This is fantastic!
Thanks for watching
Awesome vid man
Thanks for watching! Much more coming
I have tested ChatGPT and Gemini repeatedly on logical reasoning and making connections that are outside of the standard narrative for great works and great historical authors. Both of them fail dramatically, even when given a small localized text. They cannot see what is literally there in the text. I mean LITERALLY THERE in the language of the text. We're not talking about metaphors or similes or figurative language. These models are morons. They have an extended version of programmers' type of intelligence and the deficiencies that go with them, namely all the social elements necessary to being a human and perceiving what humans mean. I'm talking of programmers, and by extension their robots (aka LLM's). Read robot as a derogatory word such as the N-word or "white trash". ROBOT.
Nice!
How can I contact with you
Gemini 1.0 is so stupid, so I am not sure how well the new model will do, Even Bing is nothing compared to ChatGPT. Opem AI is goat.
Great op btw.
You can't revolutionize reading. At best you can claim this optimizes research. I didn't hear anything of value to the average reader.
Fair opinion. But the ability to synthesize all information and context from multiple books in a single prompt is a pretty substantial breakthrough.
Regardless, thanks for watching.
If artist are encouraged to use references why do you think some hate when AI LLM does? Is it fear, human desire to control, or a misunderstanding of what references are used for? Could this be a issue that stemmed from putting worded definitions to everything including emotions, symptoms in the medical field etc? If your doctor can learn to diagnose you from reading a book of symptoms which mostly goes misdiagnosed why can’t a robot? It’s very interesting the jobs we complain about we don’t want to lose to ai.
Really great points. Seeing it as a tool to improve gets lost in the AI fear narrative for sure. We should be smart about its advancement, but open to its benefits.