Artificial intelligence remains among the most talked-about topics in modern technology, and for good reason. Although we still haven't scratched the surface of understanding how human intelligence works, AI is already giving machines the ability to "think" like a person does, in some very specific ways. AI is already being applied to solve problems that are essentially invisible, like in the algorithms that automate processes for massive businesses and websites such as Google, Facebook or the financial industry. As the technology continues to evolve, though, AI is also finding its way into more parts of daily life, including unified communications settings at work and toys we use to play.
Not only is AI getting better every day – it's getting cheaper and more accessible to businesses and hobbyists of all kinds. For example, The New York Times reported on several recently developed products and research experiments that are pushing the possibilities of AI into new territory.
New AI trends
One of the latest examples of advanced AI is a small unmanned aircraft, or drone, made by a tech startup called Skydio. The drone, branded as the R1 and still in early development stages, is familiar to anyone who has experience with this new class of gadgets that have recently become popular as stocking stuffers. But instead of being controlled by a remote, the Skydio R1 drone can actually follow its user automatically. Its on-board camera recognizes the user's facial features, body and surroundings to follow you and record video. As it follows your movements, the RI can also fly around obstacles and adjust course right alongside the user. It's billed as a new way to film extreme sports, or essentially as an intelligent selfie-stick, but some have pointed out that its AI technology could also be used in ways that might violate someone's privacy.
Another example of AI advancements becoming more accessible and visible is related to neural networks. It's here where the newest developments in AI could provide some of the most noticeable changes to UC and business communications in practice for a wide array of users.
Neural networks in business communication and more
Neural networks are computer systems modeled after the neurons in a biological brain. Using a basic set of programmed conditions, neural networks are able to solve a complex task by making incremental adjustments. Eventually, they are able to complete a recurring task successfully and progressively more efficiently, a phenomenon sometimes called "machine learning."
Computer systems using neural networks don't exactly mimic human behavior yet, but they can approach certain problems in a similar way compared to a person. Neural networks are already widely used by search engines like Google to identify the content of images, audio or video instantly and on a large scale. Increasingly, these networks are also using that same data to create their own pictures and audio, which can be experienced firsthand in new consumer electronics like the Google Home or Amazon's Alexa voice assistant. Future applications of similar technology could put neural networks on the front lines of tech support call centers. By interpreting a customer's query through speech or textual analysis, the system might be able to diagnose simple, common problems on its own, therefore saving time for the human customer support representatives to focus on the more complicated issues.
AI security risks and challenges
The future of AI holds plenty of promise in terms of communications, business technology and much more. However, more experts in the field are warning that AI technology could be developing faster than we can keep up, presenting risks to many facets of society.
In February 2018, a group of researchers from policy institutions and universities around the world made a strong case for a greater level of caution around AI. While they are quick to point out that AI has already demonstrated many positive applications that could benefit society in numerous ways, the rapid pace of this new technology has left us vulnerable to threats based on AI, too.
One of the most pertinent threats brought about by new developments to AI technology is actually an improvement upon a well-known type of online fraud known as "spearphishing." The term is a play on phishing, which is a kind of fraud that's been well publicized since the early days of the World Wide Web. Spearphishing is different in that it involves an attacker assuming a false identity and targeting a very specific person or group of people. The attacker in this instance needs to conduct a significant amount of work to understand the target's personal information and use that against him or her to steal money or something else of value. One common spearphishing tactic involves the attacker impersonating a close family member or friend, calling the victim and convincing them that the friend or family member is in danger or needs money.
According to the paper, this type of fraud could become easier, and thus more common, if intimate personal details can be extracted faster through automation and AI-enabled processes. Since spearphishing is already difficult to guard against and prevent, a more prevalent form of this attack would present a significant risk to businesses as well as individuals in general. On the other hand, more readily accessible AI tools could improve the ability of organizations to detect and respond to these more complex threats.
While it may seem like a step forward in AI tech requires a step backward in terms of safety and security, it doesn't need to continue this way. Particularly with regard to business communications capabilities, significant opportunities for greater efficiency, productivity and collaboration thanks to AI tools are already presenting themselves. Coworkers should expect to keep finding new and better ways to connect with teams regardless of location, while finding answers to critical questions faster.
To learn more about adopting technology that's at the forefront of business productivity and unified communications, reach out to Teo Technologies.