🌐
Images Videos Blog News About Series πŸ—ΊοΈ
❓
πŸ”‘

Neural Networks: Good enough? πŸ”—
1427127726  

🏷️ news
Interesting study basically coming to the conclusion that Neural networks can be fooled by lies. Hell, so can we! I think this means they're a 'good enough' model for intelligence -- that is, when coupled with a way of experiencing reality (sensors), and using that as the ultimate judge of that which is real and that which is unreal. Worked well enough for us, eh?
25 most recent posts older than 1427127726
Size:
Jump to:
POTZREBIE
POTZREBIE