Mar 9, 2021 08:06 AM
(This post was last modified: Mar 9, 2021 08:33 AM by C C.)
Interconnected single atoms could make a ‘quantum brain’
https://physicsworld.com/a/interconnecte...tum-brain/
INTRO: A network of interconnected atoms could be used to construct a “quantum brain” that mimics how a real brain learns. The new system consists of an array of cobalt atoms on a substrate of black phosphorous, and its developers at Radboud University in the Netherlands say that it could have applications in artificial intelligence.
[...] Many of today’s brain-inspired, or neuromorphic, devices use machine learning – the process by which a computer uses software, or algorithms, to train on a given set of examples – to autonomously develop the ability to perform a new task. One such machine-learning model is known as a Boltzmann machine. In physical terms, a Boltzmann machine is an interacting (Ising) system of spins in which randomly fluctuating spins (or magnetic moments) represent neurons.
Magnetic atoms on surfaces are emerging as a platform for realizing such a machine, as it is possible to use them to create tuneable networks of spins that display the necessary random motion. The problem is that magnetic exchange interactions between these atoms usually have a short range, which limits the number of connections to other atoms/neurons that can be formed.
Researchers led by Alexander Khajetoorians and Hilbert Kappen have now created a self-adapting Boltzmann machine by exploiting the orbital dynamics of individually-coupled cobalt atoms placed on black phosphorus. The new work builds on earlier experiments in which they discovered that it is possible to store binary bits of information (0s and 1s) in the electronic state of a single cobalt atom when it is placed on this two-dimensional semiconductor and a voltage is applied to the atom... (MORE)
Cells as computers
https://ethz.ch/en/news-and-events/eth-n...uters.html
RELEASE: Biological cells might one day be equipped with artificial genetic programs that work in much the same way as electronic systems. Such reprogrammed cells could perform medical tasks in our bodies, such as diagnosing diseases or providing treatment. One potential application would be altered immune cells that combat tumour cells. Since tumour cells have different genetic characteristics, the following biochemical program, for example, would have to run in the therapeutic cells: “Destroy a cell if it is type X or Y or Z”.
In mathematics and electronics, such a function is called an OR gate. “These are needed in decision-making processes whenever multiple different things lead to the same outcome, when you have to deal with alternative inputs at the same time”, explains Jiten Doshi, a doctoral student in ETH Professor Yaakov Benenson’s group in the Department of Biosystems Science and Engineering at ETH Zurich in Basel. In collaboration with colleagues, Doshi and Benenson have for the first time developed an OR gate – a molecular switching element that emits a biochemical output signal when it measures one of two or more biochemical input signals – in human cells.
Previous OR gates implemented in biological cells were quite simple, as Benenson explains. When, for example, a cell was supposed to secrete a substance in response to signal X or signal Y, scientists combined two systems: one that secreted the substance in response to signal X and another that released the substance in response to signal Y. The ETH scientists’ new OR gate, in contrast, is a true OR gate – one comprising a single system. As with all biological systems, its design takes the form of a DNA sequence. In the case of the new gate, this sequence is significantly shorter because it is one system rather than two separate ones.
Inspired by nature
To realise the OR gate, the ETH researchers used transcription, the cellular process in which a gene’s information is read and stored in the form of a messenger RNA molecule. This process is initiated by a class of molecules known as transcription factors, which bind in a specific manner to an "activation sequence" (promoter) of a gene. There are also genes with several such activation sequences. One example of this is a gene called CIITA, which has four such sequences in humans.
The ETH researchers turned to this gene for inspiration and developed synthetic constructs with a gene that is responsible for the production of a fluorescent dye and that has three activation sequences. Up to two transcription factors, and one or more small RNA molecules specifically bind to and control each of these sequences. The gene construct produces the dye when transcription is initiated via at least one of the three activation sequences – i.e., via sequence 1 or sequence 2 or sequence 3. The researchers filed a patent application for this new system.
Closing a loop. As Benenson stresses, this research closes a loop. From a historical standpoint, information processing has developed in living creatures over the course of evolution: humans and animals are very good at taking in sensory input with their brains, processing it and responding accordingly. Only in the 19th century did the development of switchable electronic components began: first with the relay; later with vacuum tubes; and finally with transistors, which enabled the construction of modern computers.
In their research, the ETH bioengineers try to bring these mathematical and electronic information processing approaches back into biological systems. “For one thing, this helps us to better understand biology, such as how biochemical decision-making processes take place in cells. For another, we can use these approaches to develop new biological functions,” Benenson says. The researchers benefit from the fact that biological cells offer ideal conditions for this.
More complex forms of diagnostics and treatment
This cellular information processing is expected to be used primarily in medical diagnostics and treatment. “Today’s medical treatments are usually rather simple: we often treat diseases with just a single drug, regardless of how complex the biology and the causes of diseases may be,” Benenson says. This stands in contrast to how an organism deals with external changes. The body’s stress reactions, for example, can be very complex.
“Our biomolecular information processing approach holds out the promise of using artificial genetic networks that can identify and process different signals to one day develop complex cellular diagnostics systems and potentially more effective forms of treatment,” Benenson says. Such forms of treatment would also identify when a normal state has been reached following successful treatment. For example, an ideal cancer treatment fights tumour cells as long as they are present in the body, but does not fight healthy tissue, because doing so would cause damage.
https://physicsworld.com/a/interconnecte...tum-brain/
INTRO: A network of interconnected atoms could be used to construct a “quantum brain” that mimics how a real brain learns. The new system consists of an array of cobalt atoms on a substrate of black phosphorous, and its developers at Radboud University in the Netherlands say that it could have applications in artificial intelligence.
[...] Many of today’s brain-inspired, or neuromorphic, devices use machine learning – the process by which a computer uses software, or algorithms, to train on a given set of examples – to autonomously develop the ability to perform a new task. One such machine-learning model is known as a Boltzmann machine. In physical terms, a Boltzmann machine is an interacting (Ising) system of spins in which randomly fluctuating spins (or magnetic moments) represent neurons.
Magnetic atoms on surfaces are emerging as a platform for realizing such a machine, as it is possible to use them to create tuneable networks of spins that display the necessary random motion. The problem is that magnetic exchange interactions between these atoms usually have a short range, which limits the number of connections to other atoms/neurons that can be formed.
Researchers led by Alexander Khajetoorians and Hilbert Kappen have now created a self-adapting Boltzmann machine by exploiting the orbital dynamics of individually-coupled cobalt atoms placed on black phosphorus. The new work builds on earlier experiments in which they discovered that it is possible to store binary bits of information (0s and 1s) in the electronic state of a single cobalt atom when it is placed on this two-dimensional semiconductor and a voltage is applied to the atom... (MORE)
Cells as computers
https://ethz.ch/en/news-and-events/eth-n...uters.html
RELEASE: Biological cells might one day be equipped with artificial genetic programs that work in much the same way as electronic systems. Such reprogrammed cells could perform medical tasks in our bodies, such as diagnosing diseases or providing treatment. One potential application would be altered immune cells that combat tumour cells. Since tumour cells have different genetic characteristics, the following biochemical program, for example, would have to run in the therapeutic cells: “Destroy a cell if it is type X or Y or Z”.
In mathematics and electronics, such a function is called an OR gate. “These are needed in decision-making processes whenever multiple different things lead to the same outcome, when you have to deal with alternative inputs at the same time”, explains Jiten Doshi, a doctoral student in ETH Professor Yaakov Benenson’s group in the Department of Biosystems Science and Engineering at ETH Zurich in Basel. In collaboration with colleagues, Doshi and Benenson have for the first time developed an OR gate – a molecular switching element that emits a biochemical output signal when it measures one of two or more biochemical input signals – in human cells.
Previous OR gates implemented in biological cells were quite simple, as Benenson explains. When, for example, a cell was supposed to secrete a substance in response to signal X or signal Y, scientists combined two systems: one that secreted the substance in response to signal X and another that released the substance in response to signal Y. The ETH scientists’ new OR gate, in contrast, is a true OR gate – one comprising a single system. As with all biological systems, its design takes the form of a DNA sequence. In the case of the new gate, this sequence is significantly shorter because it is one system rather than two separate ones.
Inspired by nature
To realise the OR gate, the ETH researchers used transcription, the cellular process in which a gene’s information is read and stored in the form of a messenger RNA molecule. This process is initiated by a class of molecules known as transcription factors, which bind in a specific manner to an "activation sequence" (promoter) of a gene. There are also genes with several such activation sequences. One example of this is a gene called CIITA, which has four such sequences in humans.
The ETH researchers turned to this gene for inspiration and developed synthetic constructs with a gene that is responsible for the production of a fluorescent dye and that has three activation sequences. Up to two transcription factors, and one or more small RNA molecules specifically bind to and control each of these sequences. The gene construct produces the dye when transcription is initiated via at least one of the three activation sequences – i.e., via sequence 1 or sequence 2 or sequence 3. The researchers filed a patent application for this new system.
Closing a loop. As Benenson stresses, this research closes a loop. From a historical standpoint, information processing has developed in living creatures over the course of evolution: humans and animals are very good at taking in sensory input with their brains, processing it and responding accordingly. Only in the 19th century did the development of switchable electronic components began: first with the relay; later with vacuum tubes; and finally with transistors, which enabled the construction of modern computers.
In their research, the ETH bioengineers try to bring these mathematical and electronic information processing approaches back into biological systems. “For one thing, this helps us to better understand biology, such as how biochemical decision-making processes take place in cells. For another, we can use these approaches to develop new biological functions,” Benenson says. The researchers benefit from the fact that biological cells offer ideal conditions for this.
More complex forms of diagnostics and treatment
This cellular information processing is expected to be used primarily in medical diagnostics and treatment. “Today’s medical treatments are usually rather simple: we often treat diseases with just a single drug, regardless of how complex the biology and the causes of diseases may be,” Benenson says. This stands in contrast to how an organism deals with external changes. The body’s stress reactions, for example, can be very complex.
“Our biomolecular information processing approach holds out the promise of using artificial genetic networks that can identify and process different signals to one day develop complex cellular diagnostics systems and potentially more effective forms of treatment,” Benenson says. Such forms of treatment would also identify when a normal state has been reached following successful treatment. For example, an ideal cancer treatment fights tumour cells as long as they are present in the body, but does not fight healthy tissue, because doing so would cause damage.
