I’ve been reading (as is my thing) a lot of science fact and fiction books and internet posts about future technologies and transforming societies lately. Some of the more gung-ho technologists look forward to the time when machines–robots, computers, call them what you will–are able to do many of the things even content creators do now. Technology continues to be a mixed blessing, like all human inventions, giving with one hand and taking with another. Are we really on the verge of being replaced? In a previous post, I said that I doubt it, but it’s still an interesting thought experiment.
What Would an Automated Technical Writer Have to Be Able to Do?
Technical writers are an interesting group of people, often requiring multiple sets of cognitive skills to function well:
- Interacting with subject matter experts and users.
- Defining and articulating user/reader needs.
- Identifying and developing content that is:
- Graphically appealing or intuitive
- Written at a content and language level appropriate to the background and needs of the user/reader
- Easily translatable to other languages or contexts
Can a machine do all that? No. Not yet, I would hasten to add. However, artificial intelligence researchers are continuing to push the envelope on its various tools. And while IBM’s Watson won on Jeopardy a while back, the machines are still writing bad screenplays and making questionable art. What about documentation, though? Training materials? Technical reports?
The Death of Documentation, Revisited
User documentation, based on my experience, could be computer generated, assuming it’s needed at all. I recall the young person at the Apple Genius proudly telling me, upon buying my first iPhone, that “There isn’t a tech manual” or words to that effect. Still, that young person was taught by someone how to use the tool. Someone had to write or develop that training material. Are you telling me the programmers did it all on their own? I’ve talked to programmers–many of them regard tech writers as a necessary evil…and sometimes not even necessary.
Still, let’s assume that eventually several groups of programmers at Apple, Microsoft, Google, etc., have access to technical writers and they teach the computers how to write the stuff. That is, if the programming goes a certain way, the software would know what to say to a user in the event of a problem. That begs a question: would writers willingly contribute to a project designed to eliminate them from the workforce?
A few experiences come to mind to reduce the creeping fear of technology-based unemployment:
- Have you ever tried to get help from Microsoft or Apple, just to name a couple? Say you’ve got a problem like you want your image captions to be numbered to mirror the chapter they’re in. So, for example, if you’re adding a figure to section 1.1, that first figure would be 1.1-1, the second would be 1.1-2, etc. Ah, but now let’s set you’re in the same document at the next level down–section 1.1.1–and you want to do the same thing (with figures now numbered 1.1.1-1, 1.1.1-2, etc.). Can you? If so, what do you type in the help prompt? How many answers do you have to dig through, answering no to “Was this helpful?” until you get the result you want?
[By the way, the answer is, NO, Microsoft Word can’t do that. You can number things at level 2 or level 3, but not both. If one of you knows something I don’t, please pass it on!]
- I don’t know about you, but I get seriously irritated by voice-based telephone menus. First of all, if I had a simple question, I could probably look that up. However, if I have a problem so complex that it requires me to call the bank, the pharmacy, the software company, etc., the odds are better than even that my request won’t even appear in the list of options. And some telephone queues absolutely resist many people’s default option of pressing “0” to speak with a real, live human operator. So then what do you do? You select the option that seems closest, most likely going through that aggravation more than once, then get to a person, or you never find an option that meets your needs, never get the “press 0” option, and end up more frustrated than ever. What’s the lesson here?
- Human programmers can’t think of everything, so they can’t always write answers for every type of question.
- Human customers don’t know how “smart” the computer system is that they’re talking to, so we don’t always know what to ask for when we use our magic keywords to get the help we want.
- This bodes well for both technical communicators and customer service representatives when it comes to job security. As one tech buddy put it to me recently, “Humans will always find a way to screw things up that no one ever thought of before.”
What Do You Do When the Computer Knows Better Than You?
I recall reading an article in 1984 that discussed the fact that computers were now designing their own successors (microprocessor chips) because the devices had become so tiny and the job so complex that humans could no longer keep track of them anymore. That was 34 years ago. Computers and their related hardware and software have only become more complex. But still, human beings are the ones designing the logic and processes that dictate how computers will go about their business. Engineer-to-engineer documentation in that realm definitely requires that you “speak Engineerish” at a very high level. What happens, though, when our various programs, ‘bots, and other artificial thinking machines start doing things that we can’t explain? Does it become a requirement that the computers teach/tell the humans what they did or how they did it?
I might be getting a little ahead of myself. Or maybe I’m behind the times. I am a rocket guy, after all, not a computer guy. Regardless, the more you read, the more you realize how downright spooky-smart our artificial-intelligence-based tools are becoming. What happens when the machines get so smart they start doing things their human designers can’t explain? Will they bother? Will they say the computer equivalent of “You just wouldn’t get it?” Or, if asked, would Watson invoke HAL 9000 and say, “I’m sorry, Dave. I can’t do that?” Good grief, let’s hope not.
The point of this long think piece was just to get you used to the idea that computers are getting smarter, however you care to define the term. And eventually, they will get flexible enough that some of the simpler tasks that have kept us employed as technical writers will no longer require us. Fortunately, as long as humans remain imperfect, our creations will be, too, so there will always be a need for more additions, more modifications, and more edits. We’ll never completely go away, any more than engineers or technicians.
The future will be different.