Why Elon Musk Is Wrong About AI

There’s a growing debate about the impact that artificial intelligence will have on our future, with two tech luminaries themselves -- Tesla CEO Elon Musk and Facebook CEO Mark Zuckerberg -- as figureheads representing glass half empty versus half full perspectives, respectively. Last week Musk commented that AI is an “existential risk for human civilization.” Zuckerberg retorted that comments like this are “pretty irresponsible,” to which Musk tweeted that Zuckerberg’s “understanding of the subject is limited.” While these comments refer to sweeping impacts, many are debating one specific area where we are already seeing the effects of AI: jobs.

As humans, we’re trained to watch for threats to our survival and predict tragedies. Jobs are intrinsically linked to our survival, as they’re the way most of us earn income and are therefore able to provide for our basic needs. However, many are predicting that with the advent of AI, we will see the rise of a “useless class”—people who are not just unemployed, but are unemployable.

This is a chilling and pessimistic view of the future. If the last century of incredible advances in digital technologies leads to the creation of a “useless class” of people who have nothing better to do than play virtual-reality video games all day, that’s a tragedy for civilization. If that’s what happens, we will look back at Musk’s remarks and say they were accurate. But if that happens, shame on us because we are in charge of our destiny, not AI. AI itself is not a thing; it is a series of combined technologies that humans are creating and guiding the impacts of, including impacts on work.

In particular, those of us in the technology industry have an obligation to shape the future of AI and robotics to help create better and more productive jobs. We can leverage AI to ensure that opportunity is more equally distributed around the country and around the world, rather than concentrated in small pockets of urban wealth and opportunity.

Investing energy in the vigilant watch over the future of work is wise because only one thing is sure: Jobs will change. However, buying into doom and gloom is not wise, in my opinion. There is time to shape our future and make it a positive one. Everyone in society has an obligation to ensure that people are educated for a future in which AI touches every aspect of work. But it’s up to those of us who build technology to ensure that it augments human workers, not replaces them.

This is an area where Silicon Valley culture has fallen short, with its obsessive focus on eliminating labor costs. However, there are indications that people in technology are starting to think differently about their obligations toward humanity, and to design their products accordingly.

When it comes to dirty, dangerous, and demeaning work, automation can save lives and increase human dignity. There are already signs that this “fourth industrial revolution” will increase gross domestic product and overall productivity, just as the previous three have done, and it could also increase the flexibility and geographic diversity of work. If this is what we can expect from robots and automation, bring it on.

How AI technology can create opportunity

It’s true that technology has enormous power to eliminate jobs. In 1900, more than 40% of the population worked in agriculture, but by 2000, that was down to 2%, thanks to the efficiencies introduced by farming machines, as economist David Autor points out. Similarly, self-driving vehicle technologies may eventually make millions of truck drivers, taxi drivers, and other driving occupations obsolete. People who do those jobs now will need to find new work.

On the other hand, automation can result in a net increase of jobs. The number of bank tellers in the U.S. has doubled since the introduction of the ATM. And while farm machinery decimated the market for agricultural jobs, overall participation in the U.S. workforce grew steadily throughout the 20th century. In every major transition to date, we’ve wound up with more jobs, not fewer.

There is evidence that this is happening now. Indeed, nonfarm private employment has risen for 87 months in a row and unemployment levels are at record lows, in a sign that Internet technologies have not in fact destroyed jobs. Meanwhile, in the past year, about one-third of U.S. companies have started deploying artificial intelligence. This enormous transition is already beginning.

In the future, AI can help augment people’s work regardless of where they live. For instance, AI-enhanced medical diagnoses may bring the power of supercomputers and the world’s best medical centers into the hands of local family doctors. AI-powered news algorithms can improve our knowledge of world events and help fight fake news. AI can increase the productivity of computer programmers wherever they live, not just in Silicon Valley.

Our obligation to educate ourselves

One reason the last century resulted in so many new jobs is because of the early 20th-century movement to extend mandatory schooling through high school, providing education for people who no longer had farm jobs to look forward to. That decision ensured that we had millions of literate, well-educated people ready to take on the jobs that the second half of the 20th century needed.

We need to do the same now. Only this time, we need to jettison our outdated, 19th-century model of classroom education, and embrace new approaches more suited to our rapidly changing times. Individuals should position themselves for a lifetime of learning since the skills demanded by the workplace are changing more rapidly than ever. Traditional college degrees no longer lead to stable long-term employment opportunities—fresh training on new skills is much more impactful. Companies should also be prepared to retrain people when they replace them with machines. And we need more public-private education partnerships that combine contributions from both business and government.

Yes, we need safety nets to help people through these massive transitions, but instead of merely investing in social safety nets, we need to address the root causes.

Those of us in technology need to guide it to augment humans, not replace them. And companies and society as a whole need to invest in education to ensure we and our children are ready for jobs we can’t even imagine yet.

If we do that, as our ancestors did at the beginning of the 20th century, we can help ensure that AI will usher in an era of opportunity and wealth for all.

This article originally appeared in Fortune. Did you enjoy this article? Click the Follow button below to be notified about my future posts.