Some believe that the proliferation of artificial intelligence and robotics threatens our personal lives, our work and even our safety. More and more tasks go to the implementation of the brain based on silicon. But even the loudest critics can’t acknowledge the obvious benefits that AI and automated systems prepared for mankind. In the framework of the project Grand Challenges BBC brought together experts who presented their vision of the future in the presence of machinery and artificial intelligence.
“We have to consider AI, not as something competing with us, but as something that can strengthen our own ability,” says Takeo Canada, Professor of robotics at Carnegie — Mellon. Because AI has a tolerance for boredom, and can identify patterns much better and faster than people. Automation has already started unraveling the most complex problems of our world, from diseases to violence.
And it can make our lives safer in the 21st century.
The battle with infectious diseases
For billions of people around the world buzzing mosquito near your ear can mean much more than the annoying sting it can be a precursor to illness and even death. One species, Aedes aegypti, especially spread from Africa into nearly all tropical and subtropical regions, carrying Dengue fever, yellow fever, zika and chikungunya (the virus that causes crippling pain in the joints). Only one Dengue infects 390 million people in 128 countries every year.
“This mosquito is a tiny demon,” says Rainier Mallol, a computer engineer from the Dominican Republic, the hotspot distribution Zeke. Along with Desi Raja, a doctor from Malaysia (another country at risk of Contracting the virus), the pair have developed AI algorithms that predict where it is more likely to occur in a flash.
A Premonition of Microsoft Project uses unmanned aerial vehicles to search for pathogens in hot spots distribution Zeke
Their artificial intelligence in medical epidemiologii (Aime) is a system that integrates time and location of each new case of Dengue in the local hospitals with 274 other variable factors such as wind direction, humidity, temperature, population density, type of housing. “It’s all the factors determining the spread of mosquitoes,” explains Malla.
Trials in Malaysia and Brazil showed that they can predict outbreaks with an accuracy of about 88% in three months. The system also helps to determine the epicentre of the outbreak with accuracy up to 400 meters, allowing local doctors time to intervene with insecticides and protection from bites for locals.
Aime is also being developed to predict the outbreak of zika and chikungunya. A huge technology company on its own to develop this idea: for example, a Premonition of Microsoft Project uses Autonomous drones to identify the nidus of mosquitoes, use carbon dioxide and light traps to catch these insects. The DNA of the mosquitoes and animals that they bite, and then analyzed by machine algorithms that detect patterns in huge volumes of data better and better each time and find agents.
Combat arms
Over the past year in the US killed 15,000 people because of the shooting. This country has the highest level associated with weapons of violence all over the developed world. To solve the problem with indiscriminate shooting and gun-related crimes in some cities are turning to technology for help.
Automated system that hears gunshots through a number of sensors that can be used to determine where the shots were fired, and notify security authorities within 45 seconds after it was launched the trigger. ShotSpotter uses acoustic sensors 15-20 per square kilometer to detect a characteristic “cotton” of the shot, determining his place of birth is accurate to 25 meters.
Machine learning technology is used to confirm that the sound was a gunshot, and count the number of shots showing, there will be police to deal with a lone gunman or more criminals, and they use machines or not.
Already 90 cities mostly in the US but also in South Africa and South America – are using ShotSpotter. Small systems have also been deployed in nine campuses in the United States in response to the recent shooting on campus.
Ralph Clark, ShotSpotter’s Executive Director, believes that in the future this system can be used not only for incident response.
“We want to understand how our data can be used for the predictive capability of the police,” he says. “Machine learning can be combined with weather, traffic and others to more accurately inform police patrols”.
The fight against hunger
About 800 million people worldwide rely on the roots of the cassava (cassava) as the main source of carbohydrates. This starchy vegetable similar to yams, I eat like a potato; it can also be crushed into flour for making bread and pastries. It can grow where others can not culture that made cassava the sixth-largest food plant in the world. However, this woody shrub also vulnerable to diseases and pests, which can lead to the destruction of entire fields of vegetables.
Researchers from the Makerere University in Kampala, Uganda, teamed up with experts on plant diseases to develop an automated system aimed at combating diseases of cassava. The project Mcrops allows local farmers to take pictures of your plants to cheap phones and use computer vision to identify characteristics of the four major diseases that lead to the devastation of the crops is cassava.
“Some of these diseases are extremely difficult to detect, and they require different actions,” explains Ernest Mwebaze, researcher of computer technology, who heads the project. “We give farmers pocket expert, so they know to pollinate the crop or destroy it and plant something else.”
This system will diagnose diseases of cassava with 88% accuracy. Usually farmers have to call the government experts to visit the farms to those identified diseases that take days and weeks before the disease spreads.
Mcrops also allows you to upload images to the database, which then diagnosed the epidemic. Mwebaze hopes that this technology will also automatically determine the problems of other types of plants, such as bananas.
The fight against cancer and vision loss
DeepMind by Google can help doctors with treatment of cancer, thanks to machine learning, which will help him to identify healthy areas of the patient’s tissue
Cancer leads to more than 8.8 million deaths worldwide, 14 million people are diagnosed with some form of cancer annually. Early detection of cancer can significantly improve chances of human survival and reduce the risk of relapse. Screening is one of the key ways for early detection of cancer, but to understand the scans and other test results very, very hard and long.
DeepMind and IBM use their AI technology to the problem. DeepMind has teamed up with British doctors of the National health service at the University College London to teach the programme on the basis of the AI to treat cancer, separating healthy tissue from tumors when scanning head and neck. It also works with Moorfields Eye Hospital in London, identifying early signs of loss of vision when scanning the eye.
“Our algorithms are able to interpret visual information while scanning,” says Dominic king, the chief of clinics in DeepMind Health. “The system learns to identify potential problems and recommend the right course of action the doctor. It is too early to comment on results but they are already very encouraging.”
King says that the methods of AI can help doctors to diagnose faster, sifting through the scans and placing priorities on those recommended for immediate consideration.
IBM also recently announced that AI Watson can analyze the image and evaluate patient records, accurately defining the tumor in 96% of cases. Now the system is undergoing medical testing in 55 hospitals worldwide, helping to diagnose cancer of the breast, lung, colorectal cancer, cervical cancer, ovarian, stomach and prostate.
Without turning off the light
Amid the intensifying controversy about how could climate change cause two catastrophic hurricane of historic proportions in the U.S., as could most effectively guide the artificial intelligence study on the use of clean, renewable energy to prevent further damage that leads to climate change?
People around the world increasingly rely on renewable energy to combat climate change and pollution caused by fossil fuels, and the challenge of balancing energy networks with these intermittent sources becomes more difficult. The proliferation of smart meters and digital energy monitors that automatically record the consumption will also provide a lot of data about how and when consumers use energy. In the EU alone plans to install 500 million smart meters in homes by 2020.
“Managing all these assets, it is impossible for man, because the response time is often of the order of several seconds,” says Valentin Robu, associate Professor of intelligent systems at the University of Heriot Watt in Edinburgh. It works with the English company Upside Energy, developing new ways to manage electrical networks.
They create machine learning algorithms to monitor production and energy demand in real time. What does it mean? That energy will continue in the calm hours and then released at peak times, for example in the morning when everyone wants to make some coffee. As soon as electric vehicles and household batteries are becoming more common, technology can be used for energy storage and uniform distribution of renewable energy flows.
Rob also says that AI can be used on an even more basic level, helping to reduce our demand for connected devices. For example, refrigerators AI can be directly controlled to turn on only when electricity demand is the lowest in the network.
As artificial intelligence protects us from cancer and excessive cruelty
Ilya Hel