Hedging Against Human Translator Obsolescence
Did you know there’s this amazing up-and-coming technology called “artificial intelligence,” apparently abbreviated as “AI,” which is going to take everyone’s job once the algorithms evolve into an artificial general intelligence, and then perhaps upgrade to a superintelligence à la Skynet and eliminate not only our jobs, but our very existence?
Caption: Behold my masterpiece. Let’s see a computer do better than that!
If you’ve gotten more than your fill of hype about AI (not to mention autonomous driving, voice assistants and so on), join the crowd. After all we’ve heard about the technology over the past couple of decades, you’d think we’d be living in a sci-fi movie instead of swearing at Alexa yet again for playing classical music when I clearly asked for some KIϟϟ.
(After that, my final command to Amazon’s voice assistant before disabling her was, “Alexa, go to hell.” She did not comply…)
In addition to the toll on our patience and attention spans, the unrelenting stream of predictions about technological advances being right around the corner has a downside: that we may not be ready when the advances eventually do come along.
As I’ve written before, given the cavernous gap between what a machine can do and what a human can in creative translation, I’m not terribly concerned about my job security. But things can change fast, and even when they move slowly, like the gradual wave of globalization in the post-Cold War world, many people are still too late to adapt. If you follow the news, you know about the growing anger among the members of the working class in developed countries who have lost their jobs to globalization and have little else but fast food, dead-end retail jobs and the gig economy to barely keep them from falling into destitution. It’s a shame that our governments and societies have done far too little to help them acquire the skills that would be marketable for better careers in a changed world.
So the thought that’s always in the back of my mind is that if this could happen to them, then maybe it could happen to white-collar professionals like me. Perhaps the hype is just that for right now, but what if a time does come when machine translation, for example, develops enough to provide some serious non-human competition to the services I offer?
I’ve been a freelance translator, turning Japanese texts into naturally flowing English, for 12 years now, and over that time I have experimented with several ideas to do something on the side that could potentially grow into a sort of backup plan or alternative to translation. As you’ll see below, my record of success in this endeavor has been pretty dismal, but you never know if you don’t try, so I’m glad I made the attempt.
Here are the ideas I’ve played with.
1. Studying Science
I thought that if I gained knowledge in scientific fields that interest me by using college-level textbooks for self-study, then I could diversify my customer base. This idea was more about expanding my range of options as a translator, as getting into technical translation would obviously be a pretty counterintuitive way to maintain relevance in the face of the machine translation threat.
Anyway, I studied a few books, but working while being a dad for an adorable, attention-hungry toddler left me little time to really learn and retain knowledge. Besides, the work in my current areas of specialty was generating plenty of income, so I gave up in a few months.
Although an AI capable of translation could theoretically interpret as well, assuming the voice recognition can be made super accurate, I figured this would be an easy extension of my translation skills. My first interpreting gig at the 2011 Asian Powerlifting Championships in Kobe felt more like a laid-back internship than a full-on interpreting assignment because there wasn’t much to do the day I went, though I did appreciate the opportunity. While that little interpreting stint went fine, it was my next pair of interpreting jobs, also in sports, but for press conferences, that was a disaster, as I detailed earlier here.
That was the last time I interpreted. Not planning to try again anytime soon.
3. More Asian Languages: Korean and Chinese
I’ve always been one of those few people who actually like studying foreign languages, and with economies becoming more interconnected back in the days before Brexit and the Trump presidency, I thought that competency in the three major languages used in Northeast Asia could lead to some interesting opportunities.
Fortunately, there was a foreign language school right here in Kyoto that specializes in Korean and Chinese. Of course they teach in Japanese, but that was perfect, especially for Korean, because the similarities between it and Japanese makes it easier to explain the language’s concepts in Japanese than in English. A Korean or even Chinese course taught in English, on the other hand, would have wasted too much of my time going through basics such as SOV (subject/object/verb) word order in Korean and particles in both Korean and Chinese, which I already know from Japanese.
I decided to start with Korean because of its greater similarity to Japanese than Chinese. For nine months I studied the language. It was quite a joy, especially the Hangul writing system, which I can still read and write because of its logic and simplicity. However, my translation work started to get crazy busy and I didn’t have the mental capacity to prepare for my lessons, so I decided to give up. Never even got started on Chinese.
4. Computer Programming
I don’t remember exactly how I hit upon this idea, but I think I must have read about some online programming courses. My dad works in the tech industry, so I’ve been around computers pretty much my whole life and I enjoy tinkering with software. Similar to the above ideas, I thought that combining my Japanese language skills with some coding chops could lead to some interesting, lucrative work in natural language processing, perhaps even landing me a sweet job back in the States, where I could help design the technology that will make translators obsolete, rather than becoming a victim.
I decided to sign up for a self-paced online coding program. To install Linux, I partitioned the hard drive on my old Windows 7 machine, and then I started learning the basics of SQL and Python. It was a lot of fun! For a side project, I even started making a text-based sci-fi adventure game in the style of Zork, albeit much less sophisticated.
This learning was a blast, until I got to a certain point…lambda functions. I could not make sense of the concept, even though I searched around online to supplement the learning materials provided by the online school. I just hit a brick wall. It was the exact same feeling I had back in high school algebra when we got to factoring. My brain shut down, both in algebra and now with lambda functions.
Revelation: I hate math.
I quit the coding school after six months of study.
5. Travel Writing
This is a work in progress. On occasion my translation customers have offered me jobs writing about sightseeing destinations in Japan. For example, I was involved in revamping the Japan National Tourism Organization’s (JNTO) travel information site with original English to replace the mediocre translations of Japanese on the older version.
Later on, I was contacted by an old customer I hadn’t heard from in ages about conducting research on-site at sightseeing spots around the Kansai region to help produce an English-language pamphlet targeting inbound travelers visiting Japan during the 2019 Rugby World Cup. This is an experience I wrote about in more detail here.
This on-site work was a fantastic assignment! Even though it paid about half what I make translating, I imagine I could make more with experience. But hey, money isn’t everything. In fact, once my son becomes independent a little over a decade from now, perhaps I’ll go into semi-retirement by cutting my translation load in half so I can pursue my hobbies, do more travel writing, along with some…
Yeah, blogging, a big fad like way over a decade ago that is now so out-dated that nobody could care less when you tell them, “I write a blog.” Hell, dogs have blogs. There’s even an awful TV show about just that concept. I doubt this blog, The Kyoto Linguist, will lead to much of anything in the way of business opportunities, but that’s okay because I do it primarily as an outlet to create some original writing instead of rewriting somebody else’s ideas, which is basically what my translation work entails.
As you can see in this anti-climactic end to this post, and which I foreshadowed above, I do not have an answer to fending off future obsolescence wreaked on my profession by the great but as yet unfulfilled potential of AI, machine learning and their ilk. Moving forward, we’ll see what happens and what other ideas I come up with.
Of course what works (and doesn’t work) for me won’t necessarily apply to you. Are you a translator who’s tried to hedge against future obsolescence? What ideas have you come up with? If you don’t mind sharing, contact me on LinkedIn and Twitter let us know!
Click here for the full archive of The Kyoto Linguist. You could also read the randomly selected article below.