Traffic Jam vs. Artificial Intelligence

This post begins with a request—if you would, please assume for the sake of argument that your current job is going to be automated away in the not-too-distant future by some version of artificial intelligence. Then, please also assume that you’re still going to have a good job.

If the pro-AI side is correct, the preceding scenario will mean having a great deal more time available for putting your mental and emotional energy towards vastly more productive and creative things than you might otherwise be stuck doing. For example, instead of wasting four hours of time driving your car in commuter traffic because you live in an area with no viable public transportation, you could be sitting in a self-driving car writing, coding, talking, thinking, dreaming, inventing, what have you.

The freeing power of technology

Udacity President and Co-Founder Sebastian Thrun has made this argument many times, as he did in this excerpt from an Op-Ed he co-authored with Alphabet Chairman Eric Schmidt entitled Let’s Stop Freaking Out About Artificial Intelligence:

The history of technology shows that there’s often initial skepticism and fear-mongering before it ultimately improves human life. The original Kodak camera was seen as destroying art. Electricity was believed to be too dangerous when it was first introduced.

But once these technologies got into the hands of millions of people, and they were developed openly and collaboratively, those fears subsided. Just as the agricultural revolution has freed us from spending our waking hours picking crops by hand in the fields, the A.I. revolution could free us from menial, repetitive, and mindless work. A.I. will do those things we don’t want to — like driving in bumper-to-bumper traffic.

If that sounds like the words of someone sporting rather rose-tinted glasses, rest assured Sebastian is capable of taking a different tone—here he is in a recent piece from Getting Smart fittingly enough entitled: Smart Machines Will Eat Jobs (Except Where Smart People Create Them):

“No office job is safe. Lots of lawyers, accountants, even surgeons will be automated away. Having spent my career watching the long, slow carnage of my own industry, I have some insight into how that will feel, and how to cope.”

Guidelines for the future

In September 2016, Stanford published a study entitled Artificial Intelligence and Life in 2030. At 52 pages it’s a hefty work, but we recommend reading the “Guidelines for the Future” section at the end of the paper, as the language directly addresses the kind of fear-mongering Sebastian and Eric caution against:

“Faced with the profound changes that AI technologies can produce, pressure for ‘more’ and ‘tougher’ regulation is inevitable. Misunderstanding about what AI is and is not, especially against a background of scare-mongering, could fuel opposition to technologies that could benefit everyone. This would be a tragic mistake. Regulation that stifles innovation, or relocates it to other jurisdictions, would be similarly counterproductive.”

The Getting Smart article referenced above quotes extensively from the same study, and pulls some language out that suggests we don’t really know yet what the future holds:

“While AI technologies are likely to have a profound future impact on employment and workplace trends in a typical North American city, it is difficult to accurately assess current impacts, positive or negative.”

The rising importance of skills

An additional quote highlights how this ambiguity significantly amplifies the rapidly growing importance of skills-based learning:

“As AI substitutes for human roles, some jobs will be eliminated and new jobs will be created. The net effect on jobs is ambiguous, but labor markets are unlikely to benefit everyone evenly. The demand for some types of skills or abilities will likely drop significantly, negatively affecting the employment levels and wages of people with those skills.”

While this quote concludes on a rather negative note, we can carry the thought forward to look at the positive side of this narrative—if falling demand for certain skills negatively affects those who have those skills, then rising demand for other skills should positively affect those who have those skills. This is not in fact a theoretical supposition, it’s real—we know demand for certain skills is going up, so our challenge then becomes, how do we train/re-train workers so that they’re equipped with the skills that ARE in demand?

AI’s benefit to humanity

In another recent Forbes article, contributor Tak Lo addresses fear-mongering around AI at both the micro and macro levels. His post is entitled What’s With All The Negative Hype Around AI?, and in it, Lo points out that:

“…many AI companies are being built that benefit humanity. As an example, Mapillary wants to crowdsource maps to make the world more accessible. My own company, Zeroth.ai, has used Sero to diagnose rice crop health issues, predict productivity and increase crop yield. Lingvist is utilizing machine learning to reduce time spent learning a new language. DeepMind, the technology behind AlphaGo that beat the #1 Go player Lee Sedol, also saved Google 15% of its energy costs.”

In addition to these micro-level examples, Tak makes his case that concerns about AI are also “untrue from a macro perspective”:

“In the short-term, self-driving cars alone could save 300,000 lives per year based on the historical precedent dating back to the 1970s. Accenture forecasts that along its three pillars of labor and capital augmentation, intelligence automation and innovation diffusion, AI will double economic growth rates by 2035. And in response to concerns that AI will take away jobs, I posit that it will augment our capabilities to do jobs.”

Which brings us back to our premise from the beginning of this post—that much of what we do for work today will be automated away, but we will still have good jobs.

What IS the future of work?

Where the concern really lies is in understanding the transition from point A to point B—from many of our current jobs disappearing, to a new future where we have a very different, post-AI work life. We’re all capable of understanding how and why the work we do now might be done by machines (point A). And honestly, we’re all able to conceive of a post-AI future in which we have very different—yet viable—jobs (point B). What is very difficult for us to conceive of is, how do we survive the transition without a lot of good, hard-working people suffering or getting lost along the way?

In-demand skills

This returns us to the question of skills, and how to learn them. The ONLY way we all get through the transition safely is to begin identifying right now—to the very best of our abilities—the skills that are going to legitimately be in demand in said post-AI future, and then teaching those skills to as many people as we can. We also have to identify right now—to the very best of our abilities—those who are most at risk of having their current skills go obsolete. Combined with these efforts, we need to shed any traces of bias that might lead us to think there are certain people/cultures/classes/communities that CAN’T learn these in-demand skills.

The freeing power of automation

As another reading recommendation, consider “Automation and Anxiety” from a special report on artificial intelligence by The Economist, from which comes the following:

“AI will not cause mass unemployment, but it will speed up the existing trend of computer-related automation, disrupting labour markets just as technological change has done before, and requiring workers to learn new skills more quickly than in the past.”

The historical perspective offered in this report makes for some very compelling—and ultimately very reassuring—reading, with examples from fields as diverse as textiles and banking provided to demonstrate how, as David Autor, an economist at the Massachusetts Institute of Technology puts it, “Automating a particular task, so that it can be done more quickly or cheaply, increases the demand for human workers to do the other tasks around it that have not been automated.”

My favorite example in this study is about ATM machines. Their arrival was predicted to put bank tellers pretty much completely out of work. But what actually happened is that, while the average number of tellers at any given branch did go down, 43% MORE branches opened between 1988 and 2004, resulting in a net increase of jobs. What those tellers did every day changed, but jobs were not lost. Plus, the changes were very good changes. Bank employees no longer had to spend all their time doing the mundane, routine, repetitive tasks—they were freed up to engage in more “human”-centric activities like sales and customer service.

Unfolding our true creative powers

We close with another quote from Sebastian and Eric’s Op-Ed:

“We can’t wait to see AI free us of mindless, menial work and empower us to unfold our true creative powers.”

At Udacity, we’re engaged in this effort on two fronts. We’re specifically teaching the builders of artificial intelligence how to push this freedom forward, while at the same time teaching the skills of the future to all those who can be freed by AI to pursue new, more rewarding career opportunities. Best of all, we’re already seeing true creative powers unfold. We see it happen every day in our classrooms, where our students are constantly amazing us with their achievements. So to answer the question posed by the title of this post, we come down firmly on the side of empowerment.

~

Artificial Intelligence Nanodegree Program