Last month, Google briefly became the subject of widespread derision after spending millions on an Olympics ad for its AI chatbot Gemini. The spot featured a supposed-to-be-heartwarming tale of a father asking AI to write a letter from his daughter to her running hero.
Olympic watchers were unimpressed. Who wants to receive a fan letter written by AI? Isn’t the purpose of parenthood to teach children how to express themselves, not outsource it? Google pulled the ad before the end of the games.
Goldman Sachs estimates that investment into AI will top $200 billion globally by 2025. Analysts are predicting that tech companies will spend a trillion dollars on AI over the next few years. All of this AI development will likely have a massive negative impact on the environment.
Many of these developments will be embedded into enterprise software, customer service platforms, and data organizing programs for business use. But much of it, like Apple’s new AI products, will be consumer-facing, even if, ironically, AI-labeled products turn some consumers off.
It’s hard to pinpoint exactly how many of the billions being invested in AI are targeting consumer uses. But it’s clear that many companies are putting AI at the front of their value proposition to consumers. LinkedIn has “write with AI” in its post box. X, formerly known as Twitter, recently added an AI image generator directly into its product. This summer, Apple announced its new AI product early, after being accused of falling behind on AI.
I’ve been increasingly curious what all those consumers are actually doing with these products. So I spoke with six Bay Area residents about their use cases, to see how they stack up against the economic, environmental, and existential impacts the technology has already begun to have on our world.
“It can't do the thinking for me”
Caroline, an Oakland resident working in banking who requested to be identified only by her first name to protect her privacy, told me she used the free version of ChatGPT to overcome writer's block, much like the Gemini ad suggests. It wasn’t for a personal expression of admiration, though. Instead, she used it to draft her self-performance review at work.
Corinne Theile, meanwhile, had a quintessential Bay Area experience in July: Getting ghosted by a man she’d dated for seven months. For $12 a month, Theile ran her breakup letter through Grammarly’s AI feature, asking it to rewrite her rambling, emotional letter more “diplomatically,” “compassionately,” and “empathetically.”
“I would copy and paste what sentences I liked out of each one, and put them all together to create what I thought was the best version,” she said “There was one sentence in particular that I was like ‘Yes, that's exactly what I'm trying to say.’”
Some people I spoke with began using AI initially for work, before it bled into their daily lives. Howard Matis, a retired Lawrence Berkeley Lab physicist, works part-time for the Contemporary Physics Education Project, selling physics posters for high school and college students. He uses the free Gemini platform to craft livelier product descriptions for the project’s Amazon listings.
Since signing up, he has used it to write poems for his wife’s birthday and their wedding anniversary. But he insists he tells her they are AI-generated.
Others signed up looking for one thing, and got something else. As the son of an SFO employee, Tony D’Antonio has been compiling lists of places to visit and restaurants to try since childhood. When Google Maps came around, he converted those text files into Google Map lists, which he shared with friends and family. He hoped AI could be the next addition to his trip planning toolkit.
“Google Reviews doesn't allow people to sort by the number of reviews, or even by ratings of places,” he said “Famously, they won't provide that ability. I wondered if [AI] was a backdoor way to get it.” Alas, he plugged the query into Gemini and instead just got the most popular places, not the highest reviewed one. Instead, he used it to create a list of food, drinks, and movies inspired by his next vacation spot.
Oakland’s Andrea Brewster is a brave soul who has entered the arena of using AI as a tool for art. A textiles and paper sculpture artist, she first embraced 3D printing when it was the latest innovation, and in the past two years has begun experimenting with AI to make something that would qualify as fine art.
In May, she had a show at Mercury 20, a cooperative gallery in downtown Oakland. The show, called Field Notes, is composed of AI art she made using a $30-a-month Midjourney subscription, embellished with her own additions to evoke the sense of an artist’s notebook.
“I equate [using AI] to making a spell as a witch,” she said. “You create your spell and you put it out there, but you're never quite sure exactly what it's going to come back with.”
While Brewster's pieces were clearly labeled as being made using AI, not everyone was happy with them; one artist quit over their inclusion.
“My feeling is the art world is taking a wait-and-see attitude,” she said. “They're not going to wholeheartedly embrace it at this point.”
The cost of convenience
Like crypto in 2021, AI is having a moment (and facing backlash from multiple angles, including its energy consumption). I’m not a pro-crypto person — I have 0.06 of a bitcoin in a crypto wallet that I check on once a year when I debate selling it to fund a vacation. But at least crypto made an argument for why it should exist: It promised to decentralize money; make transactions transparent; and make everyone less dependent on banks, financial markets, and government financial policies.
AI has less of a clear argument, even if the companies offering it have invested so much, so quickly, that they need to start monetizing as soon as possible. That need to recoup their investment has seemingly driven them straight into heavy marketing, without a clear story line — or a clear justification for the environmental and social costs.
Most people I spoke to didn’t see AI as life-changing. Not yet, anyway. They saw it as a threat to creative jobs, a way to outsource a boring task, a tool to get around how shitty Google results have become, or just something to play around with.
As the use cases grow (especially as AI systems grow to interact with the outside world, like booking flights), they will likely go from novelty to convenience. But that leaves us, as the consumers, to answer the same question consumers always have to ask: At what point does the societal and planetary harm outweigh making our lives a little easier?
