Why AI will never replace humans

Will artificial intelligence (AI) one day replace humans at work? It’s a question that seems to be asked more often each year, no matter how many times we see evidence or argument on the contrary.

The real effect of AI will be to empower people’s decision making capability, not replace them. Viewed this way, AI can boost human potential in a way not seen since the first Industrial Revolution. According to PwC research, 67% of executives say they believe AI will help humans and machines work together to be more effective. According to the World Economic Forum, 38% of businesses expect to extend their workforce to new productivity-enhancing roles, because AI allows employees to do much more with the time that they have.

In the field of data analytics, I can’t ever see a scenario in which AI will be able to completely replace humans. While machines can crunch numbers, perform certain tasks far faster than humans can, and work tirelessly around the clock, they will never have our intuition. This is where the efficiency of machines and human intuition will always intersect.

We know that any analytics approach that blindly accepts all data will eventually result in bias that can lead to erroneous conclusions. This is why it is essential to pair a human employee with AI to question and reject potentially misleading data points. Only a human has the creative capability to both properly investigate data and decide on the correct course of action from the results.

But there will be considerable positive impacts from the ongoing uptake of AI in analytics. One area that we can already see AI and humans working well together is in the field of recruitment. 

While an AI system could look at a collection of resumes and name the supposed best candidates by analysing the historical data of past hires, eventually the system would simply perpetuate some of the bias that lead to those initial hires. By controlling the data fed to these AI systems, we can ensure that both people and machines aren’t unfairly excluding certain demographics due to either human or systemic bias.

Humans will never be able to process data as fast as computers, but we will always be able to think abstractly and solve problems through broad solutions that don’t get bogged down in irrelevant details. It’s this ability to ignore certain details that enables us to think innovatively and develop ideas seemingly out of nowhere.

Until “General” AI becomes a reality, we are still a long way off an AI that can simply invent something that doesn’t exist. In data analytics, an AI can’t come up with new questions for data on the fly. AI can’t develop a “hunch” that there may be a different answer to their question, and will certainly always lack the curiosity we need to query data.

So, predictions of AI replacing human decision making are still very premature. While machines will increasingly remove the burden of having to perform repetitive tasks, this is no different to the continuous evolution in our relationship with machines since the industrial revolution.

About the author:

As the Founder and Director of Velocity Business Solutions, my team and I work as collaborative consultants to help our clients utilise world-leading data analytics and visualisation tools, while also offering a range of services including deployment, education and ongoing support. If you are currently working with or exploring data analytics, and would like to learn how you can begin transforming your business, please feel free to get in touch with me at [email protected].