Scientists are working on a project that applies genetic engineering to bacteria to make them capable of reacting to a stimulus associated with a signal from a linguistic code. The goal is for this population of bacteria to be able to read Morse code, a necessary step to use living organisms in computing. The idea behind the project, which is funded by the United States Office of Naval Research, is to test whether living beings such as bacteria can create neural networks that allow them to support an artificial intelligence system.
A team is working on the project that includes researchers from the Institute for Integrative Systems Biology (I2SysBio), a joint center of the Higher Council for Scientific Research (CSIC) and the University of Valencia (UV), in Spain.
Biological computing studies how to use elements of nature to process and store information. Like any other branch of computing, it is based on the combination of hardware, the equivalent of computer equipment (in this case, living cells), with software, the program to use the information. In the case of the project carried out by the I2SysBio De Novo Synthetic Biology Laboratory, bacteria from a population of the Escherichia coli species are genetically modified to react to a certain signal, thus providing a computer that does not need software.
These bacteria are capable of learning thanks to the fact that a memory has been incorporated into their genes: they have already been capable of learning to play tic-tac-toe against humans and receiving as their only knowledge whether they have won or lost.
“Now we are designing intelligent bacteria that are capable of learning to decode signals,” says the director of the laboratory, the CSIC scientist Alfonso Jaramillo. The principle they apply is based on physics, specifically on the phenomenon known as resonance.
“The particles that make up matter have a characteristic vibration frequency. If you act on them with the same frequency, they will vibrate with the maximum possible amplitude”, explains Jaramillo, who began his research career as a theoretical physicist at the Institute of Corpuscular Physics (IFIC), another mixed CSIC and UV center near the I2SysBio.
What Jaramillo’s team does is modify some bacterial genes so that they oscillate (react) to a certain signal, in this case a chemical pulse with a specific time duration such as Morse code signals (formed by long and short pulses). . The resonance instructions are stored in the memory of the bacteria. Upon receiving the programmed signal, the bacteria generate proteins that cause the bacteria to light up (fluorescence), in a process similar to that of the synapses in our brain.
Alfonso Jaramillo. (Photo: Isidoro Garcia)
Use mushrooms as a supercomputer
“In this way, we obtain a neuromorphic system, a population of bacteria that functions as a super neuron,” describes the CSIC scientist. According to Jaramillo, in the future the sum of the reactions of this population of bacteria would be capable of decoding any letter of the Morse code. At the moment they could only read one letter at a time, but this is the first step to create in living organisms what in computing is known as an artificial neural network, a concept inspired by biology, where a set of units (neurons) are connected to each other. yes to transmit signals.
“If we could use this system in fungi, which have been shown to be capable of conducting electricity and creating networks between trees, we could create something similar to the planet Pandora from the movie Avatar,” says Jaramillo.
The goal of the project is to demonstrate that biological organisms can be used to do computation, a biological computer that, according to Jaramillo, has advantages even over the quantum computer. “A living organism does not consume electricity, it is robust to damage, it can be integrated into other living organisms, it has a low cost and it reproduces itself”, he sums up. (Source: CSIC)