I run a customer service outsourcing company that helps consumer technology companies to scale quickly. We provide customer support and back office services to the most technologically advanced firms in the world. It is an interesting position to be in. Most of our clients would love to automate the work we do, and, if anyone is going to automate, our clients , who employ tens of thousands of the most advanced developers in the world among them, will be among the first to do it.
So you can imagine the reaction from our 5,000 teammates to this week's Economist's piece The End of the Line, foretelling doom for the entire business process outsourcing (BPO) industry as the robots rise up to take our jobs.
I've spent the past eight years in the BPO industry, on "the brink of extinction", here are five things that I have learned about automation:
1. Jobs Will Be Lost - Our first client provided voicemail-to-text, a service that would email customers transcripts of any message left for them. We employed over 100 people to listen and transcribe messages. Across the world this client and its competitors employed tens of thousands of people doing this job. Today, almost all of these jobs have been replaced by the voice-to-text algorithms like Apple's Siri, Google Voice and Nuance. Later, we had a client that used us to manually scrape pricing information from across the web. We started with two people and the account grew to over 200 at its peak. Two hundred people doing the work that one, relatively simple, data scraping algorithm could do. That is exactly what happened. The client developed a piece of software and no longer needed the 200 people to scrape data.
2. Education is the Path Forward - Having to stand in front of 200 people and tell them that they have lost their jobs is awful - I know because I did it. After that, we made a decision to move up the value chain - from simple data entry to complex customer service, trust & safety operations and graphic design work - things that would be harder to automate. We worked hard to retrain our teammates, empowering them with marketable skills. This is a model that needs to be deployed globally. Businesses and government must invest in educating their people with skills that are needed in the 21st century. This is one area where the Philippines government is failing miserably. According to the World Bank, Philippines' public schools spend just $138 on each student each year. For comparison, Thailand spends $853 per student, Japan spends over $5,000 and the United States spends $12,401.
3. It Always Takes Longer - I cannot count the number of times new clients have told me, "You know we are going to automate this next quarter right?" Right. Today's techno-optimists say we are just years away from a post-scarcity utopia, where no one will have to work and everyone will have a minimum guaranteed income. There have certainly been massive advances in technology in the past few decades. Some of these advances have put people out of work. But the impact to date has been far less than the techno-optimists would have you believe. In fact productivity growth has slowed in recent years, not sped up, and the number of hours American's have to work has increased, not decreased. If the machines are taking all the jobs, how could this be? I believe it is because things always take longer than we predict. The client's who have promised that they will no longer need our services in 90 days, still need our services three years later. Artificial intelligence is complex stuff, and getting it right is very, very hard, especially when you have a business to worry about building. I am not saying that technology will never replace these jobs, just that it will take longer than people predict.
4. The Machines Need Us - Before the machines take all of the jobs, they will rely heavily on us to educate them. In fact you are already educating them. When your auto-correct learns to type your name or Spotify gets better at predicting what kind of music you'll enjoy, that is machines learning from you. Your actions are educating algorithms. Our smartest clients have used our workforce in similar ways. When we were transcribing voicemail messages, we weren't just pumping out words, we were also educating the voice-to-text learning algorithm which made more and more accurate predictions of what word was represented on the underlying audio file by following the guidance provided by our human transcribers. Today, we are seeing similar technology impact email support. Companies like Gorgias will suggest the best response to a customer email. It does this by using natural language processing to identify the issue type in an incoming email, and through a learning algorithm which refines its suggested responses to customers based on the free written responses and edits of humans. The man-machine collaboration is what work will look like for the foreseeable future.
5. In The End... - If we use machines to respond to emails faster, then we will need less people to respond to emails. Surely, this will put people out of work! cry the doomsday prognosticators. They are right in the short term, but wrong in the long term. As I said, jobs will be lost. But we are not heading toward a future where no one has to work. We are not even heading toward a future with significantly fewer jobs. Two hundred years ago the Luddites fretted about a future without work, as their jobs as seamstresses were lost to the mechanical sewing machine. While these jobs went, other jobs were created during the industrial revolution. In fact, entire new industries emerged that no one previously could have conceived of, and these industries required people with new skillsets. The same is true today. According to the World Economic Forum, 65% of children entering primary school today will ultimately end up working in completely new job types that don't yet exist.
Our industry will transform. Jobs will be lost, temporarily. But businesses and people who invest in re-educating themselves with the skills we need in the coming century, will thrive, business and people who align themselves with the machines, leveraging their power and improving them in the process, will out compete, and businesses and people who best harness the power of human creativity will manifest entire new industries and the jobs that go with them.