Artificial intelligence has deeply influenced all aspects of our work and life. In 2016, the White House realized the importance of artificial intelligence in cutting-edge meetings. The concept of a driverless car has become a reality: Uber's unmanned car in Pittsburgh, the latest model of Tesla equipped with fully autonomous hardware. Google's DeepMind platform, AlphaGo, defeated the world champion of Go and achieved the forecast ten years ahead of schedule.
Marie des Jardins, associate dean of computer science at the University of Maryland, believes that "increasing the use of machine learning and knowledge-based modeling methods" is a major trend in 2017. How will this big show be staged?
To predict the trend of artificial intelligence in 2017, TechRepublic invited several experts: Vince Conitzer, Professor of Computer Science at Duke University, CEOFabio Cardenas, Sundown AI, Roman Yampolskiy, Director of Cyber ​​Security Lab, University of Louisville, Professor of Artificial Intelligence, University of New South Wales Toby Walsh.
InfoQ has compiled these experts' predictions for 2017 artificial intelligence into readers.
The influence of artificial intelligence is growing
Conitzer believes that the public is increasingly interested in the social impact of artificial intelligence, especially those that have a greater impact. Many common topics have been identified before: technical unemployment (InfoQ Note: Unemployment due to industrial development; unemployment due to the adoption of new technologies), autonomous weapons (InfoQ Note: means that artificial intelligence can attack targets without human intervention) Weapons, popularly speaking, are killing robots.) Machine learning systems based on bias data, machine learning systems are used to identify and suppress objections, and ethical issues that are constantly generated by artificial intelligence systems—the success or failure of specific new technologies will be Further promote and expand the influence of these technical topics.
Conitzer notes that interest in AI has spread to specific groups outside the industry. For example, lawyers will try to solve the problem of how the law deals with autonomous vehicles; economists study the technical unemployment caused by artificial intelligence; sociologists study more efficiently. The impact of AI-based recommendation systems and personal assistants.
Fabio Cardenas, CEO of Sundown AI, agrees. He goes on to say that artificial intelligence affects specific roles within the organization, such as accounting, finance, supply chain, human resources, or areas where other professionals work. These areas of growth will enable AI to be rolled out across multiple sectors across multiple industries worldwide.
Deviant artificial intelligence
The Terminator plot is a common AI novel in science fiction. But is this situation true? According to Cardenas, deviant artificial intelligence may become a reality in 2017. He imagined, "For some malicious purposes, a handful of computer thieves created artificial intelligence for fraudulent institutions or individuals." He said, "This rogue AI will be able to invade systems that are considered impregnable."
Cardenas noted that fraud has occurred through AI. He said: "If the AI's training library is destroyed, the hacker can insert prejudice or exemption to subvert the AI's predictive ability and benefit from it."
Cardenas said that the possible way is to make other AIs smarter through "AI development." AI can improve by checking the blind spots in the training data. "Or, if we are lucky, write code that can improve other AIs," He said the result is to help optimize the AI ​​system. "But from a super-smart AI, there is still a long way to go. The road is long and the road is long, and I will go up and down."
Roman stated that he also believes that AI "failure" may be the trend of 2017.
He said that the most interesting and important trend in AI is that I am closely tracking: the frequency and severity of AI failures will be proportional to the capabilities of AI. In other words, when we benefit, there is “maliciousâ€. The possibility of AI has also grown.
Ethical issues of artificial intelligence
How do we prevent AI from rebelling? This is also a concern of many AI researchers. In general, experts are concerned about the ethical issues of AI. Because AI makes mistakes, it is obvious: from strengthening prejudice to insulting racial discrimination. In extreme cases, catastrophic accidents cannot be prevented.
See the MIT report: “Ethical Machines†for crowdsourcing decisions about autonomous driving, but it is misled by experts.
The traditional AI community in computer science will address the growing number of social and ethical issues at work. Conitzer said: "AI researchers have become interested in these topics, but we just want to figure out how to make specific technical contributions along these lines in the beginning. This requires relying on more computer scientists to solve them. .
Moshe Vardi said that the trend in 2017 should be: ethical issues related to AI will continue to attract attention around the impact of labor automation, deadly autonomous weapons, fairness and transparency of algorithms.
Walsh claims: "We will see an autonomous car crashing an innocent pedestrian, or a bicycle rider, or other passenger. This will highlight the seriousness of these problems as we develop and standardize AI."
Walsh also believes that value orientation issues – that is, how do we ensure that AI has the same values ​​as humans? "What we are doing today is that it is no longer considered a super-intelligent problem, but an algorithm and machine learning program," he said.
600kw Biogas Generator,Gas Generator,Gas Turbine Generator,600kw Biogas Generator Set
Jinan Guohua Green Power Equipment Co.,Ltd. , https://www.guohuagenerator.com