[vc_row][vc_column width=”3/4″][vc_raw_html]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[/vc_raw_html][text_output]Google has announced a new feature for search – Sibyl, a machine learning prediction engine. In ancient Greece, Sibyls were women who were believed to be oracles, and the Sibyl of this day and age is not dissimilar.
So what is machine learning? Machine learning, or deep learning, is an algorithm inspired by how the human brain works. The algorithm, known as the Artificial Neural Network (ANN) has been developed and modeled around the human brain, to learn in the same way we as humans do. There are no theoretical limitations to what it can do; the more data you give the machine, the better it gets.
Machine learning is no new feat. In 1956, Arthur Samuel created the world’s first self-learning program. Arthur taught the machine to play checkers against him, firstly by having it play against itself thousands of times. By 1963, the machine had beaten the Connecticut checkers champion.
In 2009, Stephen Wolfram launched Wolfram|Alpha, an answer engine that uses a series of algorithms to create a highly automated form of machine learning. Wolfram|Alpha is now one of the search engines behind Microsoft’s Bing and Apple’s Siri. At a TedX talk in 2010, Stephen predicted that machines would soon have the ability to innovate at a faster and more efficient rate than humans. Fast forward 5 years and they’re not far off…
Machines can now learn Chinese, are more accurate than humans at image classification and can produce a more accurate medical diagnosis than trained professionals.
For example, self-driving cars can decipher the difference between another car or a dog running across the road, all thanks to machine learning. Such advances have the ability to completely transform our society. Cars may return back home once dropping someone off or circulate and be used by others – will there be a need for car parks anymore? This could change the whole infrastructure of how towns and cities are currently modelled. If people are free to entertain themselves as they wish whilst in transit, could this spell the end of radio? And if you knew the chances being hit by a car when cycling on the road were almost zero, how would this alter your daily routine and health?
Sibyl may hold the potential to revolutionise search in the same way, learning from time-based information in the future such as sales calendars, the weather, channel advertising schedules, the stock market and cyclical seasonal patterns; currently, it generates a +10% improvement in performance. It could develop to become faster and more efficient than humans.
Machine learning grew out of Artificial Intelligence (AI), and technological advances such as these nearly always provoke a debate on the threats they could pose to human employment. Afterall, 80% of employment in the developed world is service based, and include jobs that machines can now perform – driving, diagnostics, preparing food or finding legal proceedings. Sibyl and machine learning systems are an amazing piece of technological work, however it does have one flaw. In my opinion, no machine can compute or predict the most complex of evolutionary systems, that is the human brain. Sibyl will be a powerful tool and aid, but will need that human collaboration and input to make it brilliant in search – at least for now. [/text_output][/vc_column][vc_column width=”1/4″][visibility type=”hidden-desktop”][gap size=”20px”][/visibility][vc_widget_sidebar sidebar_id=”ups-sidebar-2″][/vc_column][/vc_row]