⇧ [VIDÉO] You might also like this partner content (after ad)
Bacteria playing tag, that seems rather unlikely. However, scientists have made this possible by modifying bacteria Escherichia Coli. They have thus been “tuned” to behave like electronic components and act like a neural network, a form ofartificial intelligence able to learn.
“ We propose an evolutionary strategy to design genetic circuits capable of autonomously learning decision-making in complex environments. ”, sum up the researchers in one sentence in their pre-published research report on bioRxiv. For the moment, the “decision-making” in question consists in making the right choice in a game of noughts and crosses: which box? Not so strange, since this test is frequently used to test artificial intelligence systems.
It all started in 2019, with the work of another research team. A strain of a particular bacterium, Escherichia Coli, has been genetically modified to be able to sense 12 different chemicals, and react accordingly by altering the activity of certain genes. This strain was nicknamed “Puppet”.
The research team took this work and applied other modifications. They integrated many copies of two circular pieces of DNA, called “plasmids”. Each of these plasmids “codes” a different fluorescent protein. One red, the other green. The ratio of these two plasmids present in bacteria, and therefore their final color, is not determined in advance: it is also influenced by the 12 different chemicals, but also by certain antibiotics. If no such modification is applied, the ratio remains constant. In a way, the composition of DNA therefore forms a sort of “memory”, frozen until the next modification.
When, on the contrary, a new “data” is entered, via chemicals or antibiotics, the ratio (and therefore the color) evolves, but taking into account the previous configuration. This means that some memory is retained, and therefore some form of learning is possible. This behavior is similar to that of an electronic component called a “memristor”. This new biological “component” has therefore been named “memregulons” by scientists.
A memristor, in fact, is a component similar to a transistor, which can not only process, but also store data in its internal memory. It is a component that is used in particular to create what are called “neural networks”. Indeed, in many cases today, what is called “artificial intelligence” consists of a system that is “fed” with a large amount of data to “learn” and extract logical connections in order to a given objective. Recognize faces, texts… OR play Tic Tac Toe… These learning methods are inspired by the functioning of biological neurons. Today, they are closer to statistical methods. We therefore speak of an “artificial neural network”. The data sent circulates in an artificial “grid” of neurons, usually virtual. They are actually points in the network, “linked” together by computer code. This network therefore receives incoming information, the learning data, and transmits outgoing information.
From the moment the bacteria are able to take an incoming data (the chemical product or antibiotic), to transmit a “result” (modification of the DNA and therefore of the color), while retaining the previous data in memory, one can understand the comparison that is made. In the same way, the scientists therefore undertook to create a “neural network” with these bacteria.
This resulted in a system that could adapt by learning from its past behavior, reprogramming itself — like a learning AI, in short. In the example of the crab lice, here is how the scientists proceeded: a bit like to “train” the bacteria, they applied learning by “reinforcement”, that is to say with a system of reward and punishment according to whether the action was strategic enough or not.
These bacteria were cultured in compartments corresponding to a crab grid. Each time the human plays, the bacterium is “informed” by the addition of a chemical product in the compartment, each product corresponding to a location. If, at first, the bacterium “played” its next move randomly, it eventually learned through the reinforcement method, by being “punished”, by the addition of antibiotics when the play action was bad. . Not so simple, however: a game took several days to play.
The demonstration concerns a crab, which is rather harmless. According to the scientists, however, the possibilities of their research are very extensive. ” Altered gene circuits could endow living cells with decision-making capabilities “, they say as well. They could become capable of “reprogramming” themselves in much more complex decision-making contexts.
We want to say thanks to the writer of this write-up for this incredible material
Genetically Modified Bacteria Play Tic Tac Toe With Scientists
Check out our social media profiles as well as other pages related to themhttps://yaroos.com/related-pages/