Curated Tech Content

Whether we are publishing our own original content or helping our partners get the word out about their technologies. We bring you the latest news, market trends and product innovation.

The Rise of the Smart Machines

Smart machines and Artificial Intelligence are already prevalent in our daily lives. This month we take a peak at how these technologies are used amongst us, as ASI chatted to Dr Stephen Gould, AI expert from the Australian National University’s College of Engineering and Computer Science.

Just like the Industrial Revolution brought about a whole new world of harder, better, faster, stronger, Smart Machines or advanced Artificial Intelligence (AI) has introduced a whole new world again. This time it’s not just an academic research area, AI has real applications across many industry sectors from healthcare, to finance, to marketing, to infrastructure.

“These sort of smart algorithms are already entering our daily lives and business and will continue to do so,” says Dr Stephen Gould. “There are enormous benefits, from applications like speech recognition – talking into your phone and having it understand what you are saying or talking to your car’s GPS system while you are driving. Image understanding, robotics, data analysis, environment monitoring – a range of applications and areas can benefit from AI.”

Technology research firm Gartner forecasts that smart machines will have widespread adoption through to 2020. The firm also predicts that by 2020 smart machines will have already radically changed business models across industry sectors, as well as cost structures and competitiveness, before we even need to develop a full-functioning artificial brain.

One facet of smart machines that is seen to be practical is called machine learning, which is feeding algorithms with data to learn how to pick up on patterns or predict certain outcomes. One of the strengths of artificial intelligence is the ability to look through large amounts of data and to draw a hypothesis from it.

In healthcare, for example, computers look through thousands or even millions or health records can pick up patterns that are of interest to GPs. AI can also be used for patient monitoring systems, where machines track a patient’s heart rate with various thresholds so that an alert can be sent to a nurse of doctor when a potential problem is about to occur. The development of those thresholds is part guided by AI and partly guided by expert knowledge.

In finance, AI systems have been developed to act as agents that trade on the stock market. It could also look through financial data and predict if there’s going to be an impending financial crisis, for example.

Recommender systems, where machine learning is used to target specific consumers with specific ads they are likely to be interested in, is another big application of AI. Amazon, for example, looks at users’ previous purchases to recommend new products for them. Recommender systems are also used for suggesting to Twitter and Facebook users who they should follow based on their current contacts and interest groups. All this is based on machine learning and artificial intelligence in the back end.

Traffic optimisation and predicting equipment faults is where AI is most applied in infrastructure businesses. It can help companies find the optimal route for fastest or cheapest delivery of goods, or look at alleviating traffic congestion issues by directing drivers to other routes, for example. Power companies also use AI to predict spikes in demand on their electricity grids or for early detection of equipment failing.

“There’s also been a lot of effort in developing both safety systems and autonomous cars that can drive themselves. It’s still a little way off but even before we get there, there are safety systems with cars now that may pre-empt an accident and apply the brakes before you hit the car in front of you,” Dr Gould adds.

However, there’s still a way to go in developing AI systems, Dr Gould points out. An AI system controlling a car, for example, is still not completely able to deal with the unexpected. “It needs to be able to handle a cyclist pulling in front of the car, or a pedestrian walking in front of the car, or even a policeman directing traffic when normally the car would be directed by a traffic light. There are a lot of these unexpected situations which humans handle quite well, but machines still are a way from being able to handle.”

Verification of these AI systems and being reliable enough to make the right decisions off of the data mined also have a way to go. Despite this, developers of AI systems are taking into account the safety aspects of it, and are getting better at designing interfaces that will allow humans to work with machines safely, he says.

How about you? Have you considered the impact that smart machines and artificial intelligence will have in your life over the next decade? We certainly hope that smart machines bring out a world where humans work less and have more time to enjoy the benefits of a balanced, and maybe safer, lifestyle.

Image courtesy Pixabay