AI – Replacement or complimentary to humans?

Original Article
Artificial Intelligence (AI) is embedding every aspect of technology, with its potential being tapped for significant enhancements across myriad industries.

That’s why it is one of the megatrends that Gartner has identified, with the research touching on the far-reaching effects it could have in a number of its sessions during this week’s Symposium/ITxpo in Cape Town.

While there is genuine excitement across the ICT sector, some are still hesitant of the disruption it will bring forth. More specifically there is concern about the status of jobs, and whether AI will make the vast number of more basic forms of employment obsolete.

To better understand whether those concerns are unfounded or are a realistic possibility, as well as unpacking some of the additional benefits that AI can deliver, we sat down with research director at Gartner, Magnus Revang.

Augmenting the artificial
Along with several new concepts being bandied about at Symposium/ITxpo 2018, Gartner has been punting the concept of Augmented Intelligence.

The way Gartner frames it, augmented intelligence is where AI is used to support and enhance the decision-making capabilities of a human.

As Revang explains it, AI can be wrong 10 percent of the time, with it usually being the same 10 percent each instance. A human can also be wrong 10 percent of the time, but the things we’re wrong about often differs each instance. If you augment a human with the power of AI, that 10 percent now drops to 3 percent, according to the research director.

For Revang the advantages of augmented intelligence boils down to situational decision-making.

In environments like autonomous vehicles, military drones or healthcare, where the downside of AI being wrong is significant, having a human at the helm empowered by the knowledge of AI but still being able to make more considered decisions makes the most sense.

“In those cases using AI to augment my ability to make decisions, instead of making them directly, is the right thing to do. Using machines to empower my decisions, to be quicker, more precise, is also want you want to do,” adds Revang.

It seems as if Gartner is pinning many of its AI hopes on augmented intelligence, with it viewed as being central to many of the AI-related use cases the research firm has predicted for the near future.

“A lot of the use cases we see for AI will be augmented intelligence instead of outsourcing everything to artificial intelligence,” notes Revang.

Allaying some fears
It is this re-framing of Artificial Intelligence as a tool that can enhance and complement human capacity rather than replace that Revang wishes to hammer home, especially with so much fear mongering around the technology.

“When people talk about AI, a lot of the fear is psychological. Humans only have reference when looking at intelligence, and that’s their own. So when we talk to someone who isn’t familiar with AI technology, there is an innate assumption that it has more capability than it really has,” he says.

Added to this fear mongering is the “marketing” around different AI solutions or services, with the term “it learns by itself” often used. This according to Revang is a lie, with the anthropomorphism that certain channels are trying to imbue into their AI not providing a genuine reflection of the technology behind AI, he adds.

“It’s almost always humans that provide learning for AI,” Revang says.

As such provocative statements like AI learning by itself proves worrisome for the research director, especially with the number showing that employees are afraid that AI (and robots) will take their jobs.

This is simply not the case though, with Gartner’s recent research showing that jobs resulting from Artificial Intelligence will be as high as 2 million by 2025 with it creating the demand for new skills and requirements.

It is worth noting though that the same report states that there will be a percentage of job losses in the middle to lower-end sector. As such there will once again be a need for organisations and employers to ensure that their workforce is being adequately re-skilled and prepared by the coming advent of Artificial Intelligence.

Freeing up capacity
The real value of AI will be its ability to free up capacity for employees, specifically as it pertains to handling the workload for mundane or time-consuming repetitive tasks.

“End users are afraid that Artificial Intelligence will take over their jobs, but the fact is it won’t take over their jobs, it will take over tasks. AI should be viewed as an automation technology, with it taking over tasks that have a sliding scale of quality,” notes Revang.

To illustrate his point the research director looks at a call centre, which could implement a chatbot setup in order to handle traffic. If the chatbot can take 70 percent of calls coming in for example, there is still the leftover 30 percent that needs to be addressed.

The chatbot can’t complete all the functions of a call centre operator either, explains Revang, with the call centre operator now freed up to perform other tasks for the business as less of their time is taken up by handling queries.

“Artificial Intelligence changed the composition of their tasks and what they’re able to do. It did not replace their job. From an aggregate level there is less need for people, but they’re paid more and more employees can be put to use on the maintenance and analytics of the system,” he says.

“Every AI project changes the work of a human’s life. It’s not simply a technology project. It’s a human project as well,” concludes Revang.

Leave a Reply