He compares automation to transportation.

“If you are in logistics and need to go from point A to point B, you can go on a bicycle, a scooter, a motorcycle or a car,” he said. “Each mode depends on what you can afford and what resources you have. Similarly, for each problem, whether in HR, legal or driving, there will be multiple solutions. Which one you use will depend on your problem, your budget and your available resources.”

His point is that automation comes in degrees, not in a single sweeping change. You see it in Siemens factories where AI predicts equipment failures before they happen, saving both time and cost. At the other end of the spectrum, researchers are exploring systems that could one day manage entire fleets or run global supply networks with limited human involvement. Singh doesn’t see that happening anytime soon.

“I do not believe that in 20 years robots will repair or program other robots and humans will do nothing,” he said. “There will be a lot of change, but not complete automation.”

That view runs counter to the fear that AI will quickly take over most jobs. Singh said the shift will be uneven, with some industries moving faster than others and many jobs changing rather than disappearing. He calls this hybrid growth, where humans and machines work together at different levels of automation.

“Within robotics itself, there will be many new jobs: building robots, maintaining robots, repairing robots,” he said. “There will be teams of technicians, like the corps we have today for mechanical work, who will keep the robots running.”

He is equally concerned about cybersecurity, which he calls the other side of the AI coin.

“When a computer knows where you go, what you eat and what you buy, that information has value, and people will go after it.” Singh told me about large companies hit by ransomware attacks, including a major oil services provider that had to pay several million dollars to unlock its systems after every employee saw the same message: access denied.

Another Houston-based firm, he said, reportedly paid $70 million after losing three days of productivity.

“Think about how much a company that size loses every day when no one can work,” he said.

He likens hackers to pirates off the Horn of Africa, negotiating payments through middlemen. Only now, the attacks can happen anywhere, anytime.

“Today, in the U.S., we have about a million cyberattacks a day. Most are stopped, but some succeed,” he said.

The lesson, he explained, is that while automation makes life easier, it also increases our exposure.

“On the defensive side, you have to be perfect every time,” he said. “One vulnerability is all it takes for them to get in.”

Looking ahead, Singh sees AI developing along several tracks. One is agentic AI, where intelligent systems act on their own to complete a defined task. Think of a software agent that reads purchase requests, checks budget rules, and submits an order when the criteria are met. These agents handle the routine parts, while people step in for exceptions or disputes.

Another is human-in-the-loop AI, which keeps a person in charge while the system does the heavy lifting. In health care, AI flags images that look risky and a clinician makes the call. In engineering, a model proposes options and an experienced manager selects the one that balances cost and safety. The value comes from speed and consistency, but the decision remains with a person who understands context and consequences.

But make no mistake; AI will replace a significant number of human jobs. Employers need to be upskilling their workers now.

“AI transition will take time. Therefore, the upskilled employee will have work in the next 20 years,” Singh says. “The challenge is going to be for the people who are joining the workforce now.”

Leslie Meredith has been writing about technology for more than a decade. As a mom of four, value, usefulness and online safety take priority. Have a question? Email Leslie at asklesliemeredith@gmail.com.