AI and Robotics Made 10 Years of Steady Gains

Machines that think and move
article cover

Francis Scialabba

· 3 min read

Over the past 10 years, we’ve seen dramatic improvements in machine intelligence (AI) and dexterity (robotics). Today, industry competition is red-hot, adoption is increasing, and costs are dropping. So what’s changed?

Artificially intelligent

Algorithms leveled up. In the early 2010s, researchers kept topping each other with breakthroughs in deep learning, a technique marrying machine learning and artificial neural networks.

Deep learning has since been commercially deployed, dramatically improving algorithm performance for how computers “see” (image recognition) and “hear” (voice recognition, natural language processing).

Data: The world’s producing more of it, which helps AI find patterns and spit out predictions. IDC projects that our global datasphere (the sum of the world’s data) will grow to 175 zettabytes (ZB) in 2025 from 33 ZB in 2018.

Hardware: Training bigger algorithms requires better hardware. Initially, this meant researchers repurposing graphics chips. Top-of-the-line computing power was reserved for deep-pocketed university and corporate labs.

But as AI’s computational demands scale, so have data centers and bespoke chips that can handle the heavy lifting. And now, tech companies are developing their own semiconductors for AI training and inference.


Oxford Economics | New industrial robot installations (thousands)

Since 2010, the global stock of industrial robots has more than doubled, per Oxford Economics. Because these machines are getting cheaper and more intelligent, they’re more attractive to employers negotiating a tight job market. Increasingly, these robots don’t need to be explicitly programmed and can work on their own.

Take note: Robot-induced displacement won’t be evenly distributed. Intelligent machines are expected to increase inequality and eliminate 8.5% of all global manufacturing jobs by 2030, according to Oxford Economics.

  • This will hit the low-income, rural regions of the developed world hardest. In the U.S., close to 7 million middle-income manufacturing workers have lost their jobs since 1979 due to automation.
Stay up to date on emerging tech

Drones, automation, AI, and more. The technologies that will shape the future of business, all in one newsletter.

The manufacturing industry accounted for over 86% of the world’s operational industrial robot stock at the end of 2016, but that’s changing. Robots are being deployed in logistics, delivery, agriculture, hospitals, construction, and other settings.

The big question

What does the march of AI and robotics mean for the workforce?

  • What we know: More jobs will become automatable. Reskilling, upskilling, and plugging education gaps is necessary.
  • What we don’t know: The magnitude of automation’s effects. In a viral academic paper (not often do you see those words together), two Oxford economists estimated that about 47% of U.S. jobs are at risk of automation. Other projections have been less alarmist. The OECD, for example, says 9% of global jobs are at high risk of automation.

What can you do to prepare?

A technical STEM degree is a great start. But if that’s not your thing, experts say embrace lifelong learning and creative work. Make sure to sharpen your soft skills.

And learn to cope with AI-augmented knowledge work. As Deloitte puts it, hybrid jobs could become the new norm. These “superjobs” will draw from disparate skill sets and disrupt traditional job profiles.

Stay up to date on emerging tech

Drones, automation, AI, and more. The technologies that will shape the future of business, all in one newsletter.