The definition of intelligence is ability to overcome unknown and unseen scenarios. This definition prevents you from using training and learning to "boost" intelligence. The keyword is unseen and unknown which apriori makes training and learning useless and innate intelligence the quality of this ability.
Maybe someday you'll understand but I doubt.
Also, the MLP perceptron has not changed since more than 30 years so that speaks tons about your ignorance on the topic.
What do you think human toddlers do to become capable of overcoming unknown and unseen scenarios? They observe (training data), imitate (training data), experiment (feedback loop), etc. And what about the innate structure in our brains? Isn't that millions of years of evolutive adaptation? But sure, human intelligence has nothing to do with training
Once again, you conflate learning with intelligence. You don't need learning for intelligence, and likewise you don't need intelligence for learning. it certainly helps but it's not mandatory. That's why we call it machine learning and not machine intelligence. Are you incapable of understanding the simple and basic premise of novelty? Something for which you haven't been taught, nor trained, nor learned. That's intelligence by definition and examples are everywhere throughout history.
Alberto you tried admirably, this guy just has too much intelligence! Didn't you know Magnus Carlsen walked up to a chessboard as a toddler for the very first time and was immediately word-class....innately? I think he was just slacking off his first 23 years until he decided to apply his intelligence in the World Chess Championship in 2013.
Pretty sure Einstein also came out of the womb with a deep understanding of math and physics and didn't need to learn anything aside from language(though he may have had that at birth as well!).
30 years since I wrote my 1st chat bot using neural networks.
Learning and training are not intelligence! Intelligence is innate untrainable ability. This is ridiculous.
30 years since? Maybe that's the problem haha
The definition of intelligence is ability to overcome unknown and unseen scenarios. This definition prevents you from using training and learning to "boost" intelligence. The keyword is unseen and unknown which apriori makes training and learning useless and innate intelligence the quality of this ability.
Maybe someday you'll understand but I doubt.
Also, the MLP perceptron has not changed since more than 30 years so that speaks tons about your ignorance on the topic.
What do you think human toddlers do to become capable of overcoming unknown and unseen scenarios? They observe (training data), imitate (training data), experiment (feedback loop), etc. And what about the innate structure in our brains? Isn't that millions of years of evolutive adaptation? But sure, human intelligence has nothing to do with training
Once again, you conflate learning with intelligence. You don't need learning for intelligence, and likewise you don't need intelligence for learning. it certainly helps but it's not mandatory. That's why we call it machine learning and not machine intelligence. Are you incapable of understanding the simple and basic premise of novelty? Something for which you haven't been taught, nor trained, nor learned. That's intelligence by definition and examples are everywhere throughout history.
Dude you're absolutely obnoxious
Alberto you tried admirably, this guy just has too much intelligence! Didn't you know Magnus Carlsen walked up to a chessboard as a toddler for the very first time and was immediately word-class....innately? I think he was just slacking off his first 23 years until he decided to apply his intelligence in the World Chess Championship in 2013.
Pretty sure Einstein also came out of the womb with a deep understanding of math and physics and didn't need to learn anything aside from language(though he may have had that at birth as well!).