We Need to Talk About Artificial Intelligence
Ever heard of Liberatus, the poker player? Liberatus and four world-champion poker players sat down and played 120,000 hands, winning $1.77 million in chips. So, who is this poker prodigy? Liberatus is an intelligent robot – that's right, a card-slinging machine.
We live in an incredible age of technology that can learn and make decisions by itself. Artificial minds like Liberatus can even do this while tolerating incomplete information, determining when other sources are lying, and putting on a good poker face. Another artificial mind developed early this year by a team at Northwestern can even outperform humans in visual and analogical reasoning.
Technology's talents aren't limited to poker and analogy. According to a USA Today article
, of all US jobs, about 69% can be at least 25% automated while about 22% could be done by a machine anywhere from 75% to 100% of the time. In an interview
, Andrew Ng
of China's Baidu
guesses that any task a human can do with less than a second of thought will soon be automated. Considering how many tasks we perform with this level of thought helps put that idea into perspective. Now, factor in the other advantages AI has over humans (like higher accuracy, no need for sleep, longer working life, etc.) and its clear that machines are often equally – if not more – functional than humans are.
How does this affect you and me? It affects us as would the arrival of a new coworker who is better trained, makes less mistakes, and can work much more than you. Yeah, better polish up that resume – humans as economic beings are becoming less and less valuable in the face of an automated workforce, and many experts guess that soon, lots of us will be replaced by these machines.
Though it sounds like a doomsday scenario, human economic displacement doesn't have to be entirely bad. Dealing with an automated workforce becomes a question of how society will deal with higher unemployment. Addressing this question is a social responsibility.
Many have thought long and hard about what this responsibility means. Optimistic experts anticipate that instead of raising unemployment, a more automated workforce will create new jobs for displaced workers. This has happened in the past with the arrival of other technological innovations. For example, many people anticipated that the invention of the ATM would eliminate bank tellers. Instead, the ATM reduced the cost of running a bank branch. With reduce costs, banks opened more branches and created demand for even more bank tellers – between 1980 and 2010, the ATM helped raised bank teller employment by 50,000.
The trend of technology creating jobs isn't rare. In fact, there's an economic principle that causes it: the multiplier effect
. Briefly summarized, the multiplier effect says that when restrictions on banks' ability to lend money are loosened (allowing them to lend people more money), the country's money supply will increase; it's basically the idea that money creates money. Economist Enrico Moretti
in his book The New Geography of Jobs extends the multiplier effect to the effects of technology on the workforce. He concludes that for every high-tech job created in the US, five service jobs will be created as well. He cites several examples, one of which is that of Apple, which directly employs 13,000 people while simultaneously spurring 70,000 jobs in the surrounding city of Cupertino. Sounds like a pretty great deal for everyone involved, right?
Even if Moretti is right about job creation, there is still a large problem – the geographic placement of workers. High-tech jobs generally require employees to have one or more college degrees. These college graduates don't have much reason to move to cities where there aren't as many tech jobs, and as a result, they become concentrated in a few cities, opening tons of service jobs in cities that don't need them as much as other cities without the presence of tech companies. Moretti tackles this issue by suggesting governments invest in their economies by helping unemployed workers relocate to cities where there are stronger service economies, a move he claims will result in higher wages for both tech and service workers. Kai-Fu Lee, founder of Chinese venture capital firm Sinovation Ventures with huge influence in Chinese tech, also believes in technology's potential to improve the economy. Lee thinks that artificial technology will take 50% of human jobs, but he also believes that it will help mankind escape their need to do repetitive and unproductive tasks. AI, he says, will help mankind gain riches and end poverty, an even more optimistic outcome than Moretti.
Many others think that the happy ending won't come so easily. They anticipate necessary societal growing pains as we adjust to workforce automation. James Bessen
of Boston University thinks that to help people whose jobs are lost to AI, society needs to do more than just move them – it must redesign the education and economic systems to help train people in skills they will need for new jobs humans will occupy. Bessen advocates for systems that emphasize lifelong learning and the ability to self-educate. Other experts add economic adjustments to Bessen's proposed education fixes. The idea of a universal basic income has many people's attention; this would provide people with a guaranteed income, regardless of their employment status, that would allow them to reeducate themselves and prepare to transition into new jobs without worrying about how they'll eat. Mark Zuckerberg
of Facebook even hypothesizes that a universal basic income could spur innovation and help change society to think of success as more than economic patterns and instead as finding meaningful purpose in life, even if AI encroaches on your job.
The emergence of AI in the workforce has the potential to enrich our world in amazing ways, but it will likely take some careful hard work. So, maybe artificial intelligence isn't what we need to talk about. Instead, we should talk about how society administers education, whether our economic plans are hospitable to technological development, and how we as humans relate to each other, technology, and ourselves.
1. Brokaw, Leslie. "The Multiplier Effect of Innovation Jobs." MIT Sloan Management Review. Massachusetts Institute of Technology, 06 June 2012. sloanreview.mit.edu
2. Casey, Judi. "Does Technology Make Us More Human?" The Huffington Post. TheHuffingtonPost.com, 29 Dec. 2012. huffingtonpost.com
3. Chui, Michael et al. A future that works: automation, employment, and productivity. vols. [San Francisco]: McKinsey Global Institute, 2017. mckinsey.com
4. Dastagir, Alia E. "Robots Stealing Human Jobs Isn't the Problem. This Is." USA Today. Gannett Satellite Information Network, 29 June 2017. usatoday.com
5. "How Artificial Intelligence Will Change Everything." The Wall Street Journal. Dow Jones & Company, 06 Mar. 2017. wsj.com
6. Staff, Investopedia. "Multiplier Effect." Investopedia. Investopedia, LLC, 22 Jan. 2015. investopedia.com
7. Yan, Sophia. "A.I. Will Replace Half of All Jobs in the next Decade, Says Widely Followed Technologist." CNBC. CNBC, 27 Apr. 2017. cnbc.com