7.3 The Bacterial Alternative
109
Brains can still beat computers in many essential tasks, like image recognition
and motor action. The human brain is 10 million times slower than a run-of-the-mill
electronic computer, and is therefore at a disadvantage in logical operations, but operates with an amazing speed to solve tasks that are really important for survival.
Whereas a well programmed computer beats chess and Go champions, a professional player uses circuits inherited from the ancient hunter and honed by training
to compute in a split second the trajectory of a ball and actions required to intercept
it; a clumsy robot is no match in this skill – so far. An octopus, with its distributed
nervous system placing parts of the brain and photosensors in its tentacles, is unbeatable in the speed of computations allowing it to adjust the shape and color of its
body to a changing environment (Sect. 3.2).
Computational neural networks imitate the high connectivity of live neural networks and, what is most important, their ability to modify connection strengths in
response to activity and experience. Unlike the brain, they are not physical devices
but programs, so they retain electronic speed and precision and are free of chemical
interferences; their electronic dreams would not be induced by psychedelic drugs.
This kind of programming is most successful in machine learning, in pursuit of
Artificial Intelligence (AI). Still, its structure is not as flexible as the brain’s: it is
organized in input, output, and hidden layers, while the brain has an interconnected
architecture of which we still do not know much, and which we may never know in
detail, as it is certainly unique to each individual, and changes in time. The race is
on! AI is developing incomparably faster than our mental abilities (which perhaps
even decline as we become tied to computer screens) – but will it ever cross the line
between computation and intuition? Is the chemistry of our brain the magic force,
the hidden root of what we feel as mind, as soul – or just an obsolete residual of
animal development?
7.3 The Bacterial Alternative
Innovations in the design of multicellular organisms were built upon the basic chemistry and structural organization of single cells. Bacteria don’t have specialized sense
organs but they are able to sense the chemical composition of their surroundings
through receptors on their plasma membrane, and are sensitive to temperature and
illumination. They cannot do without it, because they need to find the way to a food
source or move into a more agreeable neighborhood, just as we do, and they are
suitably equipped for locomotion (Sect. 5.6). They are not permanently tied to other
cells but they are able to communicate by secreting and sensing certain chemicals,
and hence cooperate.
Bacteria form colonies in nutrient solutions, most commonly on surfaces, and
can be more tightly integrated in biofilms, where they become embedded in a slimy
matrix. Interactions between cohabiting bacteria may lead to the formation of clusters similar to the aggregation of particles (Sect. 2.3) or droplets (Sect. 2.5), or
pattern formation involving the same kind of activator–inhibitor combinations as
109
Brains can still beat computers in many essential tasks, like image recognition
and motor action. The human brain is 10 million times slower than a run-of-the-mill
electronic computer, and is therefore at a disadvantage in logical operations, but operates with an amazing speed to solve tasks that are really important for survival.
Whereas a well programmed computer beats chess and Go champions, a professional player uses circuits inherited from the ancient hunter and honed by training
to compute in a split second the trajectory of a ball and actions required to intercept
it; a clumsy robot is no match in this skill – so far. An octopus, with its distributed
nervous system placing parts of the brain and photosensors in its tentacles, is unbeatable in the speed of computations allowing it to adjust the shape and color of its
body to a changing environment (Sect. 3.2).
Computational neural networks imitate the high connectivity of live neural networks and, what is most important, their ability to modify connection strengths in
response to activity and experience. Unlike the brain, they are not physical devices
but programs, so they retain electronic speed and precision and are free of chemical
interferences; their electronic dreams would not be induced by psychedelic drugs.
This kind of programming is most successful in machine learning, in pursuit of
Artificial Intelligence (AI). Still, its structure is not as flexible as the brain’s: it is
organized in input, output, and hidden layers, while the brain has an interconnected
architecture of which we still do not know much, and which we may never know in
detail, as it is certainly unique to each individual, and changes in time. The race is
on! AI is developing incomparably faster than our mental abilities (which perhaps
even decline as we become tied to computer screens) – but will it ever cross the line
between computation and intuition? Is the chemistry of our brain the magic force,
the hidden root of what we feel as mind, as soul – or just an obsolete residual of
animal development?
7.3 The Bacterial Alternative
Innovations in the design of multicellular organisms were built upon the basic chemistry and structural organization of single cells. Bacteria don’t have specialized sense
organs but they are able to sense the chemical composition of their surroundings
through receptors on their plasma membrane, and are sensitive to temperature and
illumination. They cannot do without it, because they need to find the way to a food
source or move into a more agreeable neighborhood, just as we do, and they are
suitably equipped for locomotion (Sect. 5.6). They are not permanently tied to other
cells but they are able to communicate by secreting and sensing certain chemicals,
and hence cooperate.
Bacteria form colonies in nutrient solutions, most commonly on surfaces, and
can be more tightly integrated in biofilms, where they become embedded in a slimy
matrix. Interactions between cohabiting bacteria may lead to the formation of clusters similar to the aggregation of particles (Sect. 2.3) or droplets (Sect. 2.5), or
pattern formation involving the same kind of activator–inhibitor combinations as