While most people associate AI with robots, there are many other uses for this technology that aren’t as well-known. Here are some unique ways that AI can be used to benefit the world around us.
In the banking industry, AI helps detect fraudulent transactions to minimise losses and thereby protect bank profits. It also allows banks to make smarter lending decisions and generate customized insights for risk management.
Detecting Sepsis
A machine-learning algorithm scours patients’ electronic medical records for factors that increase sepsis risk and suggests treatment protocols. Its results have more than halved the time it takes for doctors to receive information they need to make decisions before a patient goes into septic shock, reducing hospital mortality by 18 percent.
Researchers say the algorithm is one of the most significant breakthroughs in detecting sepsis, a health condition that kills nearly 30% of those who develop it. The tool is the result of an industry venture, which combines research from Johns Hopkins University and spinoff Bayesian Health.
While some people have been critical of AI for predicting disease, others point out that such programs can be useful in a number of settings. With thorough validation and continual assessment, they can become a valuable part of preventing and treating sepsis.
Detecting Missing Persons
Detecting missing persons is a daunting task, especially when it comes to people who have been missing for more than 72 hours. This can be due to a number of reasons including a loved one who has run away or gone missing for a purely accidental reason like getting lost hiking in the woods.
Search and rescue organizations have long relied on data about lost-person behaviors to predict where and how to find them. They also employ a variety of techniques in the search process. This includes identifying distinguishing characteristics such as tattoos, birthmarks, and scars. It also includes using air-scenting search dogs, which are able to track the odor of human hair. The latest technology in this arena is AI-powered facial recognition. The best part is that it can do all this with minimal training.
Identifying Gravitational Waves
In 2015 the first gravitational waves were detected using the advanced Laser Interferometer Gravitational-wave Observatory (LIGO). This was a major breakthrough for science, helping to confirm part of Einstein’s theory of general relativity.
Researchers will use these data to learn more about a wide range of cosmic events, including black hole collisions and neutron star mergers. They will also look for ripples in the cosmological background, which are caused by the Big Bang.
To detect these signals, the detectors at LIGO are designed to pick up a tiny change in spacetime. They do this by sending a laser light up and down two four-kilometre long vacuum pipes in opposite corners of the US.
Identifying Exoplanets
AI is being used to find and confirm exoplanets, which are worlds outside of our solar system. These planets are often spotted by looking for dips in the light from stars that could be caused by planets passing in front of them (called transits).
Scientists use this technique to find the orbits of exoplanet candidates. Using data from telescope missions such as Nasa’s Kepler and TESS, they build up a picture of the dips in light caused by exoplanets.
But it can be a challenge to decipher which of these dips are actually caused by planets and which by random fluctuations in the star’s light. To speed up the process, astronomers are now using a machine learning algorithm developed by researchers at Warwick University. This new method identifies the likelihood that a potential planet is actually a planet, so astronomers can prioritize which ones to follow up on.
Identifying Threats
AI can be used to identify threats in a variety of ways. One way is through predictive intelligence, which uses machine learning to analyze data and spot patterns that security analysts may miss.
This can help cybersecurity experts find out how hackers have changed the way they approach a particular threat and how to counter those changes. This information can also give them insight into what cybercriminals are likely to do in the future.
Another way AI is being used to identify threats is through natural language processing, which can scrape through articles and news about cyber threats. This helps provide security teams with important intelligence on new threats, the types of attacks that are most common, and the defense tactics that cybercriminals use to attack.