Artificial Intelligence -- friend or foe? Elon Musk thinks it poses the gravest risk to worldwide peace and may spark World War III. President Vladimir Putin believes that whoever leads in AI maturity "...will become the leader of the world." No one knows with any certitude what an AI "endgame" could look like, if the technology is used for evil rather than good.
We can, perhaps, guess who the current AI leaders are and what benefits their leadership is bringing to the average human. But before we get carried away, let's just refresh ourselves on what the definition of AI is, exactly. Too often, AI and adjacent technologies become fused in definitions as one. AI is not in itself deep learning, it is not machine learning and it is not big data.
For a while, it was left to sci-fi movies to define what AI was. In Bladerunner, it was a "replicant" robot that could "think for itself," and exhibit an emotional response to stimuli from memories. In Dark Star, it was a planetary bombing system wherein the bomb became self-aware.
Jessica Groopman, principal analyst at Tractica, sets the record straight: "We define and forecast the AI market through six categorizations: machine learning, deep learning, natural language processing, computer vision, machine reasoning, and strong AI." Deep learning, it turns out, is the biggest AI subset in terms of revenue.
The latest Tractica forecast reveals that everyday consumers are being exposed to AI way ahead of anyone else. Which is a bit surprising given that the military sector has been developing AI applications for a while.
"Although military applications represent AI applications that have been under development for some time, the consumer space holds potential for greater scale -- especially given current mobile penetration, as well as broader use case applications. Also the opportunity for commercialization is higher in consumer," Groopman told SecurityNow.
In fact, the global consumer market for AI software, hardware and services will be worth $2.7 billion by the end of this year -- up from $1.9 billion last year, according to Groopman. And, it's growing very fast indeed. By 2025, it will have exploded to $42.1 billion. Google, Amazon, Facebook, Baidu and Apple are cited as good examples of how today's Internet giants are leveraging AI to interact with consumers, leading deployment. Facebook uses facial recognition to identify and help users tag friends in photos. Google offers smart, predictive autofill of text in Gmail emails. It's easy to see why consumer applications lead the market, but that's going to change soon.
Between now and 2025, the market is expected to fragment, with the development of vertical-specific applications coming to the fore. "Industries like business, advertising, finance, healthcare, investment, media, and entertainment are expected to be leading AI application areas by 2025," said Groopman. "Each of these areas has clear use cases and benefits that revolve around having access to large amounts of data, in which AI algorithms help to optimize processes, lower costs, and boost revenue."
Meantime, given widespread worry about AI-based security threats, is anything creepy going to happen?
"AI presents both new opportunities and new risks to online security. Malicious actors can use AI to develop new methods of penetration; meanwhile there are companies who are using AI to monitor all network activity for patterns and anomalies that help enhance security software and act preemptively."
So, in the end, it might all come down to good robot versus bad robot.
— Simon Marshall, Technology Journalist, special to Security Now