Rebuild the brain piece by piece.
There is a question that science has not yet been able to fully answer. How does the human brain really work? and perhaps the most impressive of all is the following, we use it every day, but we still do not fully understand how it creates thoughts, memories, consciousness and even our own perception of reality, but now a group of scientists decided to try something that until recently seemed impossible, simulating a complete human brain.
Researchers at the Jülich research center are working on a project that aims to digitally reproduce billions of neurons functioning at the same time and to do so they are going to use one of the most powerful supercomputers on the planet, Jupiter, a machine capable of processing gigantic amounts of information in parallel, but to understand the size of this challenge it is necessary to look at a smaller example.
Recently scientists managed to completely map the brain of a fruit fly, and although it is tiny, it has about 54 million neural connections, all concentrated in a space the size of a grain of sand, now imagine scaling that to a human brain, we are talking about something around 20,000 million neurons in the Cortex alone, connected by about 100 trillion synapses and even with all the computational power available, today this remains an extreme challenge.
The researchers' idea is to combine several smaller models of brain regions and integrate them into a single simulation, running at the same time, as if they were reconstructing the brain piece by piece and here there is an interesting detail, previous simulations made over the last decade failed to reach this level of complexity, even with enormous investments, but now, with the advancement of supercomputers something has changed.
Scientists have already managed to demonstrate that large-scale neural networks begin to exhibit different behaviors, not only larger, but qualitatively different, as if upon reaching a certain level of complexity, the system began to exhibit properties that did not exist before and this raises an intriguing possibility. Could it be that simulating a full-scale brain can reveal something that we can never observe directly? but there is an important limit. Even if this simulation works, it still won't be a real brain.
According to the researchers themselves, simulating does not mean recreating, because the brain is not just a network of connections, it is influenced by chemistry, environment, body and experiences, and all of this is extremely difficult to reproduce, even so, the path has already begun and points towards a future where perhaps we can observe the functioning of the human mind in a way never before possible and perhaps even answer questions that have accompanied humanity for centuries such as the origin of consciousness, what it really means to think, but understanding the brain is only part of the challenge, because while some scientists try simulate this operation in machines, others are following a much more direct path, trying to create systems that are already born with their own behavior.
