Posts

Showing posts with the label finding

Induction-Deduction-Abduction Framework for Artificial Intelligence

Image
A new framework to think about A(G)I more systematically I feel like we need a new framework to think about A(G)I more systematically. While it is true that there are already some frameworks to consider, such as Juergen Schimidhoober's Godel Machine, Marcus Hutter's AIXI, and Karl Friston's Bayesian Brain (with free energy principle), I tend to feel unsatisfied with all of these existing approaches. Of course, this tells nothing about the correctness and elegance of these frameworks or their disadvantages. My main goal is not to disprove any of these frameworks but rather to build my own that (almost) fully satisfies my (right or wrong) intentions with the framework. The reason that I am not satisfied with the existing frameworks could actually be because I haven't fully read any of them, and therefore, I am probably missing lots of insights and details that are already present in the existing approaches. Maybe I wouldn't even feel the need to come up with my ow...

The CoBra Project

Image
Collective Brain It has been a few weeks since I started thinking about building a website for fun. Basically, I was trying to come up with an idea that would be easy enough to be implemented fast and complex enough to make it interesting for users. I was also thinking about something interactive -- something that would involve the users in an interesting and fun way. So, I came up with the following idea: "Every social media platform makes its users to log in and post something individually. What if they had a platform where they post something collectively?" I don't want to talk about this particular idea and its implementation in this post very much, so, long story short: you can visit the  CoWis (Collective Wisdom)  website and see it for yourselves. Then I kept thinking about something new. This idea of making people do something "meaningful" or "unmeaningful" together seemed very intriguing to me. It has been a few days now that I finally ref...

0, 1, and beyond

Timeline I decided to start writing a book called "Introduction to Mathematics" on 21st of August 2019. This book was supposed to introduce fundamental mathematical concepts to a layman who never really explicitly studied any math. However, of course, it had to explain not any mathematical concepts but my mathematical concepts . Here is the table of contents of the unfinished book: Object Number Operations Expression and Function Set Relativity Ball Experiment Logic and Memory Continuity and Movement Classification I stopped writing it for a while before making some developments again on February and August 2020. To give a brief overview of the "finished" (but also clearly draft) parts of the book, I talked about data (or information) that constitutes every observable object ; then I tried to use the notion of this "objectness" to define numbers ; addition, subtraction, and multiplication/division were of interest in the following chapter of operations...

Adaptive Lossless Compression

Retrospection Adaptive Lossless Compression (ALC) is something that I came up with during the time period between the end of my bachelor and the first year of my master studies. Back then I sometimes joined George Hotz's streams from time to time and once I heard him saying something that was about the idea of compression leading to intelligence being very interesting. Then he said that even in physics or mathematics we try to invent some formulas to represent our sensations or observations. At that moment, the idea of compression caught my attention. I thought about the numbers, functions and sets in mathematics and one question naturally came to my mind: W hy do problems have mathematical solutions?  Another way to ask would be that why do they have solutions that are mathematical? As I kept thinking about this question, I started realizing the true depth of it. I do not have an answer to give today but I developed some intuitions about it. If you want to understand it, you m...