We don’t fully understand how a brain wires together to produce everything we see, but we do understand the basics of how they grow connections and change through reinforcements. This basic knowledge was essential for designing neural networks and various methods of deep learning in the first place, even if it doesn’t produce true human intelligence.
It’s like if you cut out sections of any animal brain and train them to do a specific task. It isn’t exactly the same as how brains work, but it does feature similar shortcomings for similar reasons. It builds connections and reinforces them, associating related “ideas” or clusters of information. The people who built machine learning models took heavy inspiration from how neurons work, while also taking inspiration from evolution when designing how they improve over iterations. There are very simple rules that lead rise to highly complex phenomenon in both.
It takes heavy inspiration from how one mathematician posited that neurons work. Actual neurons are living cells, and have a wider range of behaviors and interactions than the “neural network” model accounts for.
AI is flawed in the same way our brains are, as they process information in a similarly flawed way.
right bc that’s nonsense
centrist, uninformed yet confident
We don’t know how brains work, nobody can authoritatively say that a computer programme and a brain work the same way.
We don’t fully understand how a brain wires together to produce everything we see, but we do understand the basics of how they grow connections and change through reinforcements. This basic knowledge was essential for designing neural networks and various methods of deep learning in the first place, even if it doesn’t produce true human intelligence.
It’s like if you cut out sections of any animal brain and train them to do a specific task. It isn’t exactly the same as how brains work, but it does feature similar shortcomings for similar reasons. It builds connections and reinforces them, associating related “ideas” or clusters of information. The people who built machine learning models took heavy inspiration from how neurons work, while also taking inspiration from evolution when designing how they improve over iterations. There are very simple rules that lead rise to highly complex phenomenon in both.
It takes heavy inspiration from how one mathematician posited that neurons work. Actual neurons are living cells, and have a wider range of behaviors and interactions than the “neural network” model accounts for.