Before I hit send on this piece, I passed the words on your screen through a program called Grammarly. While it is not a foolproof editor, its algorithm will likely catch grammatical issues immediately that would take me multiple times to spot (if at all). It will also confirm to me that I haven’t stolen the below copy from another writer through its plagiarism engine. It’s just one piece of technology that helped bring this article to life.
There will not be any spelling errors because a little underline appears when I try to do that, alerting me to a missing or extraneous letter or a futile attempt at spelling out a florid* word.
* I previously used the word fancy, but a trip to Thesaurus.com led me to change it to florid.
My SEO plugin has made multiple suggestions on word usage to make sure this article reaches as wide an audience as possible. I used a headline optimizer to come up with the best way to draw you in.
There are countless ways that technology improves upon the written word. Some have been around for so long that we barely appreciate them anymore.
But because it often assists us in the background, we tend to treat mentions of technology as career destroyers, not job enhancers. Especially when we’re discussing bots capable of writing the very stories we write today.
As a reader, you have likely read a bot-written article in a newspaper powered by technology like Heliograf, Narrative Sciences, or others.
If you are a writer, this may potentially keep you up at night. You fear the bots are coming to take your job. And, it’s true, bots might replace a lot of editorial employees. But, if you’ll read on, allow me to propose a more optimistic result of the bot revolution.
//
CODA
Last night, I watched the Pittsburgh Steelers play the Cincinnati Bengals on Monday Night Football. If I hadn’t and wanted to know what happened, I could look at any number of recaps filed last night and this morning. While each story undoubtedly has its unique take on the night, the reality is that all of those stories present a collection of data points.
The key themes of the football game:
Who Won: Pittsburgh, who are still in 2nd place in the conference and lead their division.
How They Won: Comeback – they were down 17 points.
Who They Beat: Cincinnati, who now has a 1% chance to make the playoffs. This is the sixth straight time they’ve lost to the Steelers. Pittsburgh is 62-35 all-time against Cincinnati.
Key Stats: LeVon Bell gained 182 yards from scrimmage and a TD, A.J. Green had 77 yards and 2 TDs, and Antonio Brown had 101 yards and a TD.
Miscellaneous: Multiple players exited the game and did not return. However, one person, Ryan Shazier exited with a traumatic back injury – where he was not immediately able to move his legs, stoking fears of paralysis (he appears to have been very lucky, but still TBD)
#Steelers LB Ryan Shazier has shown promising signs this morning. I’m told he has some movement in his lower extremities after last night’s back injury, but the next 24-48 hours are key for increased improvement.
— Ian Rapoport (@RapSheet) December 5, 2017
These are all distinct, data points that a bot can easily access. Currently, that bot is going to do some things better than a human (access deep, nonobvious trends and stats that a human might not detect or be able to find) and things worse (describe the horror on the field when Shazier went down). A bot might also get tripped up by some confusing narratives (Pittsburgh won yesterday despite having significantly underperformed Cincinnati in 3rd down conversions, which usually implies overall success) Or does it? The point is it would take me a significant amount of time to try that theory out, whereas a bot can have that info on hand.
Point being, bots are ready to take on a lot of sports coverage. And can probably outperform human writers in educating readers on some story templates. This is why the Washington Post is farming out local area sports coverage to bots. And the technologies are only going to get better.
But all is not lost. The ideal template for at least the next couple of years is a writer/editor relationship between bot and human. The bot writes the first draft and the human edits the final draft off of that template (adding color, moving around grafs, removing the less important info). If you think about bots are editorial assistants as opposed to replacement, it seems more empowering and less frightening.
It will also likely free up editorial bandwidth to pursue the narrative stories that bots are much further away from solving.
Such as: what drove Ryan Shazier and others to continue playing football despite the dangers.
The future is uncertain and frightening, but we should keep a grounded perspective. If you think of a bot not as your competition, but as one of those person assistant robots made famous by science fiction movies, it becomes less scary. But maybe don’t think about HAL-9000.