Subscribe

Daddy's busy, go bug the AI —

Outsourcing emotion: The horror of Google’s “Dear Sydney” AI ad

Opinion: "Help my daughter write a letter" is not the same as "Help me with boring busywork."

-

Starting point or endpoint?

Google sees the message of the ad differently, of course. "We believe that AI can be a great tool for enhancing human creativity but can never replace it," Google spokesperson Alana Beale told SFGate. "Our goal was to create an authentic story celebrating Team USA. It showcases a real-life track enthusiast and her father and aims to show how the Gemini app can provide a starting point, thought starter, or early draft for someone looking for ideas for their writing."

The idea of AI output as a mere "starting point" for human endeavors has long been key to the marketing balancing act. That Copilot Superbowl ad, for instance, includes outputs with numerous "brainstormed" storyboard images and signage designs for the human prompter to pick from.

Further Reading

New Stable Diffusion 3 release excels at AI-generated body horror
Besides making the AI seem more approachable, the "starting point" framing helps obscure that current AI models usually aren't good enough to operate without significant human intervention. In most professional cases, you still need a human editor or designer to weed out the hallucinated or nonsensical AI-generated options. And even quality AI output often needs editing to avoid coming across as non-human.

But humans are lazy. It's a small leap from "Generate some ideas I can use as a starting point" to "Eh, that starting point looks fine; just submit it as is." That might be fine for busywork situations where the quality of the output isn't paramount. But as NPR's Linda Holmes memorably wrote, "Who wants an AI-written fan letter??"

  • The free version of Google Gemini frames its responses as a draft that you should add to with personal touches.
    Google
  • Gemini Advanced presents a fan letter that's completely ready to go, just fill in your name!
    Google

To Google's credit, putting the "Dear Sydney" ad prompt into Google Gemini returns a response that's clearly labeled as "a draft to get you started," as in the ad. The response even suggests that the prompter "add more personal details to make the letter even more special. For example, your daughter could mention a specific race she watched or a particular quality she admires about Sydney."

But when we put the same prompt into the subscribers-only Gemini Advanced model, that "draft" framing is gone. Instead, you get a short letter that's implicitly ready to send once you fill in the bracketed "[Daughter's name]" sections.

A kid that gets a response like that isn't going to see LLMs as a useful "starting point" tool to avoid writer's block when confronted with a blank page. They're going to see LLMs as a way to generate a complete letter that they can pass off as their own—just fill in your name here, and there's no need to add any more personal touch.

The role of Google Gemini will be played by Joaquin Phoenix.

The whole thing reminds me of the opening scene in Spike Jonze's 2013 movie Her. While Joaquin Phoenix's protagonist at first seems to be dictating a heartfelt letter to a loved one, it quickly becomes clear he is working as one of many faceless drones for "beautifulhandwrittenletters.com," writing a touching message on behalf of a stranger and intended for another stranger.

The scene is a clear statement on the alienating, dehumanizing effects of outsourcing the emotional labor of personal writing. It's also a look at a world where the idea of writing your own personal message to a loved one is apparently so foreign to people that it has become an economically viable business model for outsourced specialists.

As a vision of the future (present?) of letter writing—and creative work in general—Google's "Dear Sydney" ad might be even more grim. I want AI-powered tools to automate the most boring, mundane tasks in my life, giving me more time to spend on creative, life-affirming moments with my family. Google’s ad seems to imply that these life-affirming moments are also something to be avoided—or at least made pleasingly more efficient—through the use of AI.

Kyle Orland Kyle Orland has been the Senior Gaming Editor at Ars Technica since 2012, writing primarily about the business, tech, and culture behind video games. He has journalism and computer science degrees from University of Maryland. He once wrote a whole book about Minesweeper.

Promoted Comments

It's also disturbing because it takes the meaning out of the letter from the little girl. The details in the letter, including that the girl is "working hard" and "dreaming," are made up. Those aren't facts, those are just things that the AI made up because that's what would typically be in a letter like this.

Maybe they are true, maybe not. But as more and more people use AI to draft their letters (and more and more recipients use AI to process and summarize them), the less those details will mean anything. Everyone will know that they are probably made up. Even when the next little girl actually goes through the effort of writing a heartfelt letter, full of true facts, the recipient may well assume it was AI-generated and the facts are made up, to the extent they read it at all.

It's sad.
This is a decent comparison, but with Hallmark cards, there is no pretense that the sender wrote the message -- it's obviously a pre-written, prepackaged message that's sold widely. Getting an AI to generate a similar message (or even a draft) is something kind of different, and akin to that Her scene where you're paying someone to craft a personal message just for you.
The best analogy I’ve run across: People who use AI to write something are not writers, they are like the owner of the New York Times that pays someone to write an article. People who use AI to generate paintings are not painters, they are more like the Pope that paid Michelangelo to paint the ceiling. Using AI does not make you an artist, you are just paying for a product, not creating anything. Even worse you aren’t even paying an artist to put their sweat and effort into creating something. That’s why AI art is soulless.

Channel Ars Technica

Unsolved Mysteries Of Quantum Leap With Donald P. Bellisario

Today "Quantum Leap" series creator Donald P. Bellisario joins Ars Technica to answer once and for all the lingering questions we have about his enduringly popular show. Was Dr. Sam Beckett really leaping between all those time periods and people or did he simply imagine it all? What do people in the waiting room do while Sam is in their bodies? What happens to Sam's loyal ally Al? 30 years following the series finale, answers to these mysteries and more await.

More videos