Timecodes: 0:00 Make AI think with you, not for you 0:37 About myself and InfraNodus 1:30 Make ChatGPT think with you, not for you 3:50 💡 Sending an AI-generated question back to AI to generate ideas 6:47 Use a knowledge graph to understand the AI’s logic and find the gaps in ideas 8:45 Save AI-generated questions to your InfraNodus graphs 9:40 Using the InfraNodus AI with Google search results 11:30 Finding interesting research questions with Google Scholar and scientific publications 13:30 Exploring the context around a certain topic in science 17:20 💡 “What’s a practical use for it?” 18:15 💡 Generating questions for your Obsidian notes using the InfraNodus graph view plugin 19:56 Try this out yourself - links to the tools Try on infranodus.com
I love your quote, good ideas are at the junction of different fields. A true statement in so many ways, especially in the professional world of cross polenation between “different fields”.
@@railzip glad you like this! Yes, so many times in my own work I’ve found this to be true. It’s like bridging those gaps creates a short eureka moment when you realize something new. So it can be sort of emotional too!
Yes, this is why entrepreneurs make money by solving problems by connecting dots others didn't see could be connected. Why I studied both medicine and business, why the magic happens at the juncture of a man with a woman.... etc etc
Awesome video, really appreciate it. I followed this same thought pattern and wanted to make a set up like this to make sure I get the most out of this AI tech. Great to see another great mind confirm it all! You must be a genius lolz!
thank you! yes, i'd love it to be more popular not only because it would be great to see more people using the app but also because I think we rely too much on getting the answers from AI when questions is actually what it's much better at!
Nice video. Bias sets link to keywords lexicon used and phrasing. OSINT and network graphing the individuals silo and linking groups. Bias sets lead to echo chambers and bias creep or shifting. New to the game is generative I.A. influincers that can be deployed like a botnet, but with agents. Psychological profiling and clasification and sorting for manipulation points. Next to real time feed back for manipulation through agents. Stealth bias shifting. ( the Facebook experiment on manipulation of people. The various rules like 10% tipping point and the spread of information. Bad news traveling faster than good. Ect... ). Psyopps meets A.I. and agents. Social engineered content that injects into the algorithmic pipelines from things like keywords. Agents that have avatars that look realistic. Psychological profiling is a big one and not all systems are equal. A.I. that can see through human delusions and illusions and bias sets. It sounds like your looking into a important area. Thank you for sharing the video. Keep up the good work.
What I am missing here is reasoning about how the AI actually generates candidate questions. Why does it relate the topics together that it relates together? when generating questions
I thought the whole video is about explaining how AI is actually generating the candidate questions and why it relates the topics together. Did I miss something?
Interestingly, I've been working on this since 2011 and the idea of structural gaps and using them to bridge ideas has been there from the beginning. You can check the publications on noduslabs.com/publications and see the first paper from 2011 on this topic. Do you think I could sue Google? :)
I just recommended this to another rapidly growing AI channel, and he replied in the comments that he was excited to give infranodus a closer look. th-cam.com/video/CSuMA_QSEGg/w-d-xo.html Hope that this kind of active sharing helps.
Hey Tony, sure. First of all, no other app provides the statistics about the text network. They don't show anything about betweenness centrality (a crucial influence metric), community structure (modularity algorithms), and general structure of discourse. That makes it impossible for them to provide any actionable insight that is backed by data as to what you can actually do with the graph. The best case scenario is that they just visualize a pretty representation of your ideas, but that's about it. With InfraNodus, you can actually use the graph metrics in an actionable way to see how you can develop the discourse further (or how to read it better), depending on your needs. Hope that helps!
Timecodes:
0:00 Make AI think with you, not for you
0:37 About myself and InfraNodus
1:30 Make ChatGPT think with you, not for you
3:50 💡 Sending an AI-generated question back to AI to generate ideas
6:47 Use a knowledge graph to understand the AI’s logic and find the gaps in ideas
8:45 Save AI-generated questions to your InfraNodus graphs
9:40 Using the InfraNodus AI with Google search results
11:30 Finding interesting research questions with Google Scholar and scientific publications
13:30 Exploring the context around a certain topic in science
17:20 💡 “What’s a practical use for it?”
18:15 💡 Generating questions for your Obsidian notes using the InfraNodus graph view plugin
19:56 Try this out yourself - links to the tools
Try on infranodus.com
I love your quote, good ideas are at the junction of different fields. A true statement in so many ways, especially in the professional world of cross polenation between “different fields”.
@@railzip glad you like this! Yes, so many times in my own work I’ve found this to be true. It’s like bridging those gaps creates a short eureka moment when you realize something new. So it can be sort of emotional too!
nice pun with pollination and different fields 😂
Yes, this is why entrepreneurs make money by solving problems by connecting dots others didn't see could be connected. Why I studied both medicine and business, why the magic happens at the juncture of a man with a woman.... etc etc
Dude I am liking the 3D graph more and more and love it brother and the time line is a big hit! Keep rocking! Will support you every step of the way
Great, thank you so much! Are you using the 3D graph and do you think it's better than the 2D version?
@ yes, I think both 2D and 3D graphs have their place, if the node count is very high the 3D graph is hard to navigate.
This is fascinating. I’ve been thinking about jumping into this platform
Awesome video, really appreciate it. I followed this same thought pattern and wanted to make a set up like this to make sure I get the most out of this AI tech. Great to see another great mind confirm it all! You must be a genius lolz!
Thank you for your work.
This is so cool. More people need to see this !
thank you! yes, i'd love it to be more popular not only because it would be great to see more people using the app but also because I think we rely too much on getting the answers from AI when questions is actually what it's much better at!
Seems like you beat me to this tool 😂 amazed thanks
Nice video. Bias sets link to keywords lexicon used and phrasing. OSINT and network graphing the individuals silo and linking groups. Bias sets lead to echo chambers and bias creep or shifting. New to the game is generative I.A. influincers that can be deployed like a botnet, but with agents. Psychological profiling and clasification and sorting for manipulation points. Next to real time feed back for manipulation through agents. Stealth bias shifting. ( the Facebook experiment on manipulation of people. The various rules like 10% tipping point and the spread of information. Bad news traveling faster than good. Ect... ). Psyopps meets A.I. and agents. Social engineered content that injects into the algorithmic pipelines from things like keywords. Agents that have avatars that look realistic. Psychological profiling is a big one and not all systems are equal. A.I. that can see through human delusions and illusions and bias sets.
It sounds like your looking into a important area. Thank you for sharing the video. Keep up the good work.
@@LaboriousCretin I almost feel like your comment deserves a graph of its own to map out all the different ideas inside :)
What I am missing here is reasoning about how the AI actually generates candidate questions. Why does it relate the topics together that it relates together? when generating questions
I thought the whole video is about explaining how AI is actually generating the candidate questions and why it relates the topics together.
Did I miss something?
Can you add a feature to make the graph rotate in a 360 degrees loop?
There is one in the 3D version where it does exactly that. Looks pretty neat.
Does this work with Typing Mind? Or only Chatgpt?
It works with any web page that has text as it's an extension that extracts text from the web page.
Does this work with Claude?
You could use it on Claude chats as well yes.
So you built this tool as a solution for the information gain Google patent basically
Interestingly, I've been working on this since 2011 and the idea of structural gaps and using them to bridge ideas has been there from the beginning. You can check the publications on noduslabs.com/publications and see the first paper from 2011 on this topic. Do you think I could sue Google? :)
What is this?
I have written a paper on the same topic. Happy to share if interested
Sure, would love to see!
Yes please
I just recommended this to another rapidly growing AI channel, and he replied in the comments that he was excited to give infranodus a closer look.
th-cam.com/video/CSuMA_QSEGg/w-d-xo.html
Hope that this kind of active sharing helps.
@@christopherd.winnan8701 definitely!
imagine if Sun Tzu had this tool while writing "The Art of War"
Absolutely! I think he'd span over so many different fields!
He’d be like “behold, the art of nuclear war.”
@@LeftBehindKitten ...after looking at the gap between science and war that is...
Hi Dimitri, I'd love to know how this differs from other apps which use a graph, such as Obsidian, Capacities, Tana etc?
Hey Tony, sure. First of all, no other app provides the statistics about the text network. They don't show anything about betweenness centrality (a crucial influence metric), community structure (modularity algorithms), and general structure of discourse. That makes it impossible for them to provide any actionable insight that is backed by data as to what you can actually do with the graph. The best case scenario is that they just visualize a pretty representation of your ideas, but that's about it. With InfraNodus, you can actually use the graph metrics in an actionable way to see how you can develop the discourse further (or how to read it better), depending on your needs. Hope that helps!
@@noduslabs The app certainly seems fascinating. I'll have to investigate a bit more 👍