"Researchers at UC San Diego have created a machine learning algorithm, POLYGON, which enhances early drug discovery by simulating complex chemistry, thus accelerating the development process. This AI tool not only identifies candidate drugs quickly but also focuses on multi-target molecules, potentially reducing side effects seen in traditional combination therapies. Credit: SciTechDaily.com" (ScitechDaily, AI Transforms Drug Discovery With Faster, Safer Cancer Treatments)
"The researchers trained POLYGON on a database of over a million known bioactive molecules containing detailed information about their chemical properties and known interactions with protein targets. By learning from patterns found in the database, POLYGON is able to generate original chemical formulas for new candidate drugs that are likely to have certain properties, such as the ability to inhibit specific proteins." (ScitechDaily, AI Transforms Drug Discovery With Faster, Safer Cancer Treatments)
The AI can make customized medicine production possible. The AI can observe molecular interactions and physical-chemical environments with outstanding accuracy. The system can scan the DNA and search for similarities from multiple sources. That allows researchers to compile the DNA and find the similarities between multiple people, but that thing can also cross the species's border.
When the AI searches for medicine against bacteria or cancer cells, it can find weak points in the other cells' DNA by compiling the chemical code of those cells. The system can use mRNA bites that order those cells to die, or the system can target radiation or acoustic devices into those cells. The acoustic or photo-acoustic systems can break protein fibers.
The acoustic system can create pressure tunnels in cells or the light can vaporize water in the cells. That thing can form hydrostatic pressure waves in that cell. In some models, nano-springs. Along with ion systems can send very fast crystals to targeted cells. And that forms pressure waves that destroy those cells. The nanomachine can break the protein shell of the cell.
But if the system wants to manipulate molecules or other structures, it must see those things. The machine vision must be so effective, that it can follow the movements of the near atom-size particles. The AI and machine vision must also control things like energy levels and energy impacts on the system. In machine vision, the system drives data. That its sensors collect into mass memories.
The thing, that makes machine vision hard to make is: that the AI must know how it reacts when it sees some object. The AI must know the name of the object, it must know what to do. The problem is this, the AI must connect observation with action. In the case of the AI, the computer must recognize objects. That it sees. Then it must find the database that involves the right action.
"Harvard researchers have innovated a compact, single-shot polarization imaging system that simplifies traditional setups and expands applications in medical, AR, and smartphone technologies, enhancing real-time and machine learning-integrated imaging capabilities. Credit: SciTechDaily.com" (ScitechDaily, New Harvard Technology Paves the Way for Advanced Machine Vision)
The machine vision is not like the human eye. The robot can connect itself to outside surveillance cameras. Those cameras give the robot the ability to see itself from outside. That thing gives robots the ability to maneuver and predict situations better than any human can do. However, the system must recognize those situations so that it can react in the right way.
Machine vision faces similar problems all the time. Without depending. Does the system control some molecular actions or robots that walk on streets, it must know how to react in cases when something happens suddenly. If the system extremely long peptides within hours, and something unexpected happens, the AI must react to that thing. Or the entire process is a waste of time.
Same way. If a robot car travels on the streets and something happens, the system must react the right way.
In normal situations, the vehicle must stop in the case of danger. But what if a robot vehicle faces robbery? What if some violent person attempts to stop the robot vehicle? In those special cases, the AI must react otherways than usual.
Those special cases are challenging things for programmers. The programmers must describe situations where, for example, an armed person tries to stop a vehicle. Then the programmer must separate situations, if the person who wants to stop is hostile, or if the policeman wants to stop the vehicle.
https://scitechdaily.com/ai-transforms-drug-discovery-with-faster-safer-cancer-treatments/
https://scitechdaily.com/new-harvard-technology-paves-the-way-for-advanced-machine-vision/
https://scitechdaily.com/pioneering-crispr-gene-editing-trial-79-of-participants-see-improvement/
Comments
Post a Comment