Of all the excitement I expected from this week’s Google I/O keynote, “terror” wasn’t on my short list. Google has spent a couple of decades creating products that make my life easier. This week proved the exact opposite: a way to make communication with other people worse.
Google I/O has focused on artificial intelligence this year, and one of its focus has been on prompt-based text generation, particularly in Google’s office suite. At several points in the keynote, he demonstrated a system called “Help me write,” which rewrites simple instructions in Gmail and other apps as more refined paragraphs of text. But more lucid it doesn’t mean automatically Better.
Good writing can delight us with an intriguing turn of phrase. It can tell us something about how a person thinks and feels. Or it can offer some kind of helpful detail about the world. And there’s no reason AI can’t play a role. I have seen authors use generative tools to create unexpected prose or summarize data legibly. I’m excited to see people explore new uses for it.
But Google’s stage ideas were nearly astonishing in their ability to serve exactly none of the primary purposes of writing. His examples in Help me write calcify the worst tendencies of “professional” communication. They are bland and inflated standards that turn a human suggestion into something that uses more words to say less.
Let’s look at one of the examples. At the start of the keynote speech, a presenter automatically generates a job description for a textile design position which he (in the demo scenario) sends out to the world. Later, another presenter responds, providing the following prompt:
interested! 8 years exp with intl brands. Call on Tuesday.
Help Me Write does its magic and outputs the following:
Thank you so much for contacting us! I am very interested in the Senior Textile Designer position at Cymbal. I have 8 years experience working with international brands and am confident I have the skills and experience you are looking for.
I’m available for a call on Tuesday asap. Please let me know what time it works.
The latter version is obviously how you’re supposed to apply for a job because, written by a human, it suggests that you’re busy enough to type a multi-paragraph response and are capable of hitting certain class and cultural signifiers. However, an AI-generated response serves neither of these purposes. It was created almost instantly based on a predictive text system that requires minimal proficiency in the English language, and so far, that system can only handle fairly formulaic writing; it will not replace human labor in many real-world situations. So all it does is construct a contrived, longer version of the original prompt, which probably only has value until everyone expects it to be written with AI.
And even worse, the AI generation reinforces the idea that over-enthusiastic talk about US affairs is the necessary manner of writing, whether or not it is a necessary skill for the job. I have seen thoughtful stories of people with dyslexia using ChatGPT to produce text that is… as a Washington Post the article puts it – “unfailingly professional and polite”. But there’s an unspoken, simpler alternative: be willing to accept wider variations in how people communicate. I don’t envy anyone using AI writing to meet largely arbitrary standards, but on a society-wide basis, it’s a linguistic arms race towards a duller future.
For another example, try this friendly congratulatory message from Google Android VP of Engineering Dave Burke to VP of Devices and Services Rick Osterloh:
rick congrats on the amazing launch of pixel fold with android dave
Here’s what help me write churns out:
Congratulations on the successful launch of Pixel Fold! I’m really impressed with the device and how well it integrates with Android. The foldable screen is a game changer and I can’t wait to see what you make of it next.
I assume this is ironic because I know the way executives send personal emails and seem totally comfortable with missives without punctuation of a sentence. But then again, it’s worse! It tells me less about what the sender sounds like. It takes longer to read. Makes the sender look like an interchangeable robot.
To understand the pragmatic value of this process right now, you have to look at an email from the very beginning of the keynote. The core of the generation is actually pretty cool: It takes a very simple request for an airline claim and fills it with concrete details pulled from your inbox, like your name and flight date. But after that, he embellishes the facts with a few paragraphs of peevish prose. Here is the original:
ask for a full refund for this canceled flight
And the AI-generated version:
Dear Customer Service,
I am writing to request a full refund for my canceled flight, DF 4542 on February 4, 2023. I understand you are offering a voucher as a goodwill gesture, but I would prefer to be refunded the cost of my ticket.
I have been a loyal customer of your airline for many years and have always been satisfied with your service. However, I am very disappointed with the way my recent flight was handled. I was given no notice of the cancellation and was forced to make last minute alternative arrangements. This caused me a lot of inconvenience and expense.
I believe that a full refund is the only fair way to compensate me for the problems I have encountered. I have attached a copy of my original ticket for your reference.
Thanks for your time and consideration.
Kind regards, Rowan
There is no indication that these furious claims are accurate. He’s a bureaucrat deimatic display: Convincing a hapless customer service rep that you’re crazy enough to type all that text and that it’s probably going to be a real nuisance until you get your money back. I’ve seen this idea best pitched on Tumblr of all places. In a popular post about artificial intelligence, someone described using ChatGPT to bring out some kind of high-tech version of Brazil‘S Form 27B/6 Gambit, generating a threatening false legal letter to a landlord who was breaking house laws. As a fellow longtime tenant, I applaud them.
But this stuff is effective in a short transition period while generative text is not widespread, while readers are likely to assume it is related to a human writer. If you know it’s a machine, the illusion vanishes. You’re left with a world filled with longer, less thoughtful, harder-to-read communications. I’d rather hire someone based on an honest “8 year deadline” than a cover letter filled with empty automated prose.
In contrast, Google’s most useful example of Help Me Write was simply about passing information. In an email about a potluck, its AI was able to look at a document with a list of dishes people had signed up to bring, then summarize that list as one line in an email . It saves writers the step of pasting a bunch of items and readers the inconvenience of clicking to another tab. More importantly, its value doesn’t rest on pretending it was written by a human, and if Google has its way, it’s a gimmick that won’t last long.