What is AI and what are its benefits?
Artificial intelligence is a branch of computer science that treats the study and design of intelligent agents or software that can perceive their environments and take actions within them. One way to define AI is an attempt to create machines that think and act like humans, given sufficient data. In other words, AI could be thought of as using software designed to simulate human cognitive tasks such as problem solving, learning tasks or natural language processing. Artificial intelligence is often employed in autonomous systems where control algorithms use predictions derived from machine learning techniques with input from sensors (e.g. vision, hearing, touch, or motion) to accomplish goals set by a human operator. Artificial intelligence has been subsumed as an extension of computer science, mathematics and engineering because it relies on standard concepts from each of these fields.
AI research is deeply linked to computer science. Many leading AI researchers and most major advances in AI were achieved by the application of concepts, methods, and tools from a wide spectrum of areas in computer science, including:
“Artificial Intelligence (AI), sometimes called machine intelligence or “artificial sentience,” is the part of computer science that aims to give computers the ability to simulate intelligent human behavior.” – Wikipedia.
How will AI change the way we live and work in the future?
There are many ways in which AI will have a dramatic and transforming impact on our lives. Arguably, the biggest impact will be from machine learning, in particular, deep learning. This technology is critical for several industries because it is capable of recognizing patterns, identifying anomalies, and making predictions from data. For example, it is expected to improve e-commerce by codifying the browsing and purchasing history of customers and then suggesting products they may be interested in based on their past purchasing history. For example, in the future consumers will be able to find discounts on products they are already planning to buy rather than searching for the best deal. For medical diagnosis, AI will be capable of reading and processing imaging data (e.g. X-rays) that is harder for humans to interpret (Zampiello 2013). The ability to analyze and recognize patterns enables machines to learn new skills and behaviors by trial and error, a process known as reinforcement learning (Oudeyer 2016) and which can be seen in computer programs like AlphaGo. A drawback of this technology is that it may replace human workers for tasks that are repetitive, dull, or require judgement but are necessary for regular operations. This will cause unemployment among the workforce, which is already high in the U.S., especially among the lower half of the wage scale.
What ethical considerations need to be taken into account with AI development and implementation?
The potential for AI to augment or even replace human decision-making is a major concern for many ethicists, policy makers and businesses. On one hand, machine learning can make our lives easier by making predictive decisions on our behalf that are better than we would manage on our own. Instead of having to figure out the best route to work or school, machines can automatically do this for us. On the other hand, this may result in a loss of control over our own lives. For example, if our devices are able to predict what we are likely to buy or say, then they can also suggest products and phrases that may further influence our choices in a way that only benefits the company, not us as individuals. Just as importantly, the effect of AI on people’s lives can be positive or negative. For example, a self-driving car may prevent you from having an accident by avoiding a collision with another vehicle. However, if there is no human driving the car and you are injured in a crash, there may be less legal recourse available to help you recover costs. While it is not yet clear how AI will affect people, a number of ethical challenges need to be addressed to ensure that it is used in an acceptable manner.
How can businesses make sure they are using AI in a responsible way that benefits everyone involved?
The use of AI may lead to disruptive changes in the workplace but it is also important to ensure that these developments allow for ethical development of new work processes and skills. For example, some research has shown that AI can increase the productivity of employees in certain roles by 30% (Stefanov & D’Souza 2013). However, to avoid an inability to meet productivity targets and a loss of equity, managers must assess the risks involved with such developments. They must also ensure that such programs are not the exclusive preserve of the top-performing workers. If this happens, the more senior-level employees would have access to better resources, but this would lead to less equity for all employees.
Stronger AI-related skills will also be valuable for workers who face a greater risk of automation. Examples of where such skills are required might include roles that require higher levels of social interaction or customer service. AI is still in its early stages of development and it is important to ensure that it is used in an ethical way. Research has shown that only 15% of businesses in the EU have policies to promote ethical AI (Brytz et al. 2017).
Brytz, P., Jakubowski, J., Lund, K.-H., & Leirer, M. (2017), ‘Preventing conflicts between employees and robots: Guidance for executives in the era of artificial intelligence’, Managing Service Management, 23(4)
Fitzpatrick, G. (2017) ‘Artificial intelligence: The smart person’s guide’, ComputerworldUK, 15th February 2017.
Frey, C., Osborne, M. A. (2013) “The Future of Employment: How Susceptible Are Jobs to Computerisation?” Oxford Martin School, Working Paper No. 17-042
Are there any potential risks associated with artificial intelligence that we should be aware of?”
There are a number of potential risks with AI. First, there is the risk that a machine may misinterpret its own understanding of the world and make decisions based on those errors. For example, it could be programmed to identify the human form but believe that it is an animal. Second, AI could be flawed by lack of data or training. In this case, its results may be inaccurate or even detrimental to society at large (e.g., if a self-driving car crashes as a result of computer failure). Finally, there is the risk that someone will intentionally use AI to cause harm (Dalai 2016). For example, a criminal might program a chatbot to steal money from individuals or a hacker might create an artificial intelligence that could disrupt computer systems, such as traffic lights.
How much do you know about artificial intelligence (AI)? Test your knowledge with this quiz! Source: ZDNet
Originally Published: 5/10/2016 7:00 PM CDT Updated By: Peter Bright @PeterBright, Comment by Wendy Hui Kyong Chun @whkyong, Comment by Sridak Sahoo @sridak, Comment by John Bennett, Comment by Mark Lantz @marklantz and comment by Aditya Kalyanpur.
Programmers, feel free to comment here. However, please don’t discuss any other technical aspects of this article (including the source code). The comments section is reserved solely for discussion of this particular article and its content.