Hi Justus, thank you for reading!
Not exactly. The fact that a neural network can simulate the behavior of the neuron -- or more accurately, predict its behavior -- only reveals that artificial neurons are indeed very simplistic versions of their biological counterparts.
What you say is partially true. But if we look at the bigger picture, taking into consideration the collateral effects of ANNs innefficiency, then it's clearer that bigger networks won't be the answer forever.
For now that's the trend, but even the main players realize it's untenable. There are limits to how big we can make the networks: Physical, mechanical, thermal, and most importantly, social -- which are actually the most restricting limits.
If the AI people following this trend are doing more harm than good by making ANNs bigger without better arguments than not having any other option or any other ideas, then they're doing it wrong. People's wellbeing should go before both financial benefit and technological progress.
But not everyone is doing AI this way. Many of those who aren't, acknowledge what I explain in this article -- although they may not share the same stance about the what's the best solution.
What do you think? :)