L4 The Mathematical Proof: Why Logic Can Do Machine Learning But Machine Learning Cannot Do Logic

1 year ago

Logic can do machine learning but machine learning cannot do logic and the mathematical proof for that is by using the complexity-theoretic argument that machine learning is under a certain complexity class and logic is under a bigger complexity class and that would be a mathematical proof that indeed logic can do machine learning and machine learning cannot do logic.

For a scientific example, the difference between machine learning and logic is all about how you communicate with the machine, how you communicate with the AI, how do you say what you have to say. In machine learning, your only way is to provide examples. Imagine that the only way you could communicate with others is by giving examples, that's definitely not useful, sometimes you need to say the thing not just give examples and moreover, the receiver when they see only examples what are they going to do with an example that they didn't see it, they can only guess, so you can make an educated guess and that's what machine learning is all about but you will never be certain. In logic, you can just say the thing whatever the thing is, and whatever is implied from that will be understood with absolute certainty.
- Ohad Asor, Founder & CTO of Tau https://tau.net/our-team/ #shorts
#Tau #Tauchain #Agoras

🎥 https://www.youtube.com/watch?v=0shXLPTGBkI&list=PLeBSjZu0y6vW5g7RP8XJL67KPLtrmAiOh&index=4 👈

🧀Test Tau Meta Language.
🔗 tml.tau.net
📄Abstract: Draft for Community Review
🔗https://lnkd.in/gWJZj9sN
😼Published Codes
🔗 github.com/IDNI/TML
🌐Website of Tau
🔗 tau.net/
👨‍🏫Tutorial Videos of TML
🔗https://rb.gy/9ahenx
✈️Telegram community
🔗t.me/tauchain
Facebook: Andrew Onizuka
Twitter: @AndrewOnizuka
Instagram: andrewonizuka
Tiktok: andrewonizuka
Reddit: Andrew_Onizuka

Loading comments...