While it goes without saying that there are some things machines just cannot do, there are also many unknown and underutilised functions that are transforming the world we live in.
Here are five things machines can do that you may never have thought possible:
Caught on video
Researchers at MIT in the United States are training computers to predict human interactions before they occur. Using deep-learning algorithms and machine vision, computers can analyse video to predict what will happen next. This capability holds great potential for everything from robots and robotic technology, which will be able to develop action plans based on given scenarios, to security cameras that will automatically call emergency services if they “see” an accident.
Designer Jason Grech worked with IBM’s Watson platform to come up with new ideas, patterns and materials. Watson is a technology platform that uses natural language processing and machine learning to reveal insights from large amounts of unstructured data. For Grech, it “used cold code to make warm fabrics”.
Watson analysed hundreds of thousands of designs – public images posted on social media – aggregating factors such as colours and hem lengths, to predict future trends. The platform enabled Grech to deepen his creativity and come up with unique designs that were featured in Melbourne Spring Fashion Week 2016.
The next hot commodity for big businesses wanting to capitalise on smart computing are data-driven organisations. For instance, IBM’s recent acquisition of The Weather Channel led to the launch of Deep Thunder – a short-term, customised forecasting model to help retailers predict consumer shopping patterns and adjust supply chains accordingly.
Researchers at Stanford have developed an algorithm to identify tissue slides exhibiting a particular type of cancer with far greater accuracy than humans can. This points to a future where subjective aspects of medicine fall to computers that are far better at diagnosis than a human will ever be.
Drawing on insights gathered through its artificial intelligence partner DeepMind, Google is using machine learning to optimise the energy use of its data centres, saving percentage points on its power bill. The same optimisation strategies are useful for almost every aspect of business, from shipments through a port to mining resources, and even in agriculture, where machines can advise on planting times, animal movements and more.