Once the realm of science fiction, artificial intelligence — in the form of deep learning networks and neural processing of big data — is beginning to make a major splash in the world of business. Whatever field your business may be in, if it’s processing any form of data then it’s likely that the rise of the machines is going to be behind a serious upheaval sooner rather than later — though, thankfully, not in a Terminator-esque manner.
Simple artificial intelligences have been around for decades, their ancestry can be traced back to the 1960s with Joseph Weizenbaum’s Eliza natural language processing system. Though simple, Eliza proved adept at luring its operator into an admittedly one-sided conversation. This is much like the more powerful AIs today imbuing voice-activated assistant services like Apple’s Siri, Google’s Now platform, and Microsoft’s Cortana with personality. This is in addition to providing useful information mined from the user’s contextual data.
For companies that believe AI’s use is limited to conversational robots and waves of enemies in popular video games, though, the next few years are likely to come as a shock. AI research is yielding results at an ever-increasing pace. In a very short space of time there is unlikely to be a field not touched by the benefits — and risks — it can bring.
Show Me The Money
For those doubting the potential for artificial intelligence to make a serious impact on the world of business, there are some big names dropping even bigger money in the fight to prove its worth. Computing giant Intel recently announced the acquisition of AI specialist Nervana Systems in a deal valued at $400 million. This has been done with a view to produce processors capable of accelerating artificial intelligence and deep learning workloads by 2H 2017.
Google, meanwhile, has already launched its own AI accelerator hardware. It’s been dubbed the Tensor Processing Unit. Designed for use with Google’s TensorFlow deep learning platform, developed by recently acquired AI-focused subsidiary DeepMind, Google’s TensorFlow hardware has already proven its worth. It’s already helped the company develop a learning the Go champion system capable of easily beating a human. This was a feat previously thought impossible for a computer program to achieve but thanks to Go’s staggering complexity it’s another obstacle overcome. Outside the realm of games, the same technology powering Google’s Go robot is being used in a trial to scan retinal data for signs of macular degeneration before it’s even detectable by human specialists.
Microsoft, for its part in the research, has recently teamed up with graphics hardware specialist Nvidia on an AI-specialised supercomputing platform powered by the company’s DGX-1 hardware. “Every industry has awoken to the potential of AI,” Jen-Hsun Huang, founder and chief executive officer of Nvidia, claimed at the device’s launch, echoing his partner company’s sentiments on the potential for AI to transform the future of business.
The Benefits Of AI
Intel, Google, Microsoft, Nvidia and a host of others wouldn’t be splashing the cash quite so heavily were they not sure of the future of AI. It’s easy to see why. Even in this relatively early stage of what is increasingly looking like a business revolution there have been countless benefits. For example, artificially intelligent deep learning systems have been used to improve the security of data communications. Search for patterns in customer data, boost the quality of machine translations and even to recognize handwriting, speech and even lip-read from silent video footage, all with accuracy and speed unmatched by human experts and previous state-of-the-art software systems.
In the future, AI is likely to have an increasingly strong impact: the success of Amazon’s Echo platform, which provides an always-on microphone in users’ homes with a connection to an AI platform dubbed Alexa, has proven that concerns over privacy are often swept away in the face of user-friendly engineering that can scour a person’s private data in order to answer questions such as “when’s my next flight” or “how are my shares doing?”
Outside user-facing roles, AI will speed data processing and only increase the value of data gathered by businesses. Those who do not analyze their existing data for potentially lucrative correlations are likely to find it easier than ever to do so. While services designed to offer cloud-based data analysis will become increasingly prevalent and profitable.
The Dangers of AI
There are, of course, risks to consider before companies dive head-first into the choppy waters of the AI revolution. At present, there are many competing solutions for both software AI platforms and hardware accelerators to improve their performance. Companies specializing in a single platform may find they have backed the wrong horse should a rival platform romp home to commercial success.
A bigger concern, experts argue, is raised by the ethical considerations both from increasingly powerful artificial intelligences and from our own increasing reliance upon them. Social networking service Facebook recently found itself in hot water when its newsfeed algorithm, a deep-learning system built to replace expensive human editors, promoted fraudulent news stories during the US Presidential elections. While Facebook has claimed the promotion was accidental and inconsequential, it offers a glimpse of a world where information is controlled by machines running algorithms so far removed from anything a human had originally coded as to be incomprehensible. This is something groups like the Partnership on AI, which includes Facebook among its members, are attempting to address.
Whether you’re for or against the encroachment of AI on your business, though, one thing is clear: you ignore it at your peril.
Never miss another post. Sign up for the 100TB weekly newsletter.
Originally published at blog.100tb.com.
