AI on the Job: Will you be competing with a bot for a gig?

Generative artificial intelligence tools like chatbots are making some workers better at their jobs by augmenting their skills. But as the technology improves, others wonder whether it could automate specific jobs. Like their own jobs, for instance. 

That was a real fear for Kerry Harrison, a U.K.-based copywriter and AI educator, after she learned about an early AI software program that her clients were using to generate email subject lines back in 2017. 

“And I just remember thinking, ‘Oh my goodness, like, what does this mean for me as a writer?’ And ‘Perhaps my days are numbered,'” Harrison said.

That experience motivated her to learn how to use AI in her work. She currently trains and consults with other copywriters and businesses interested in using tools like ChatGPT.

She said it’s helped mostly with basic tasks like creating an outline or getting over writer’s block. But when it comes to work like creating original ad campaigns, it just can’t do that because everything it creates is a regurgitation, as it were,” Harrison added.

But she noted that having AI do the basics is enough for some people.

“I’ve spoken to people who are using these tools to write their blogs and to write their newsletters, and they’re happy with that outcome,” Harrison said.

And it’s already enough for some employers, even deep-pocketed ones like Disney. The studio recently confirmed it relied on AI to generate much of the opening credits for its new Marvel Disney+ show, “Secret Invasion.”

Some copywriters have lost contract gigs or been laid off due to AI outsourcing, according to recent reporting in The Washington Post.

But these tools could make creative workers more productive, said Bill Swartout, a professor of computer science at the University of Southern California. He compares generative AI’s introduction to calculators becoming a cheap, widespread tool.

“It removes the drudgery and allows you to focus on the higher-level concepts that you’re supposed to be learning,” Swartout said. 

Swartout is also co-director of the university’s Center for Generative AI and Society. He said one of the goals of the center, which is not quite running at full steam yet, is to help students learn how to utilize these tools creatively.

“We’re preparing people to work out in the world eventually. And they will be going out into a world where a ChatGPT definitely exists,” Swartout added. 

Holly Willis, professor at USC’s School of Cinematic Arts and co-director of the generative AI center, has sensed mixed feelings about these tools from some of her students.

“[There’s] both an excitement about the possibilities creatively and, you know, ‘Will the skills that I’m building in my studies right now be relevant when I graduate?'” Willis said.

But both Willis and Swartout say the center will also help people critically think about how artificial intelligence is changing society because AI is, in their view, here to stay. 

As for Kerry Harrison, she thinks people should at least try these tools out but keep in mind that something like ChatGPT isn’t a silver bullet for all their problems.

Related links and more insight from “Marketplace Tech”

Harrison’s marketing work includes “AI-inspired” campaigns from before generative AI was all the rage. They include AI-designed recipes for cupcakes, ice cream and cocktails.

And actually, generating cocktail recipes is something these newer chatbots are pretty great at, according to a recent piece in The Atlantic.

In part, that’s because there’s no real “correct” way to make a mixed drink. Unlike, say, with baking, where precision and technique are superimportant, creating a cocktail recipe is about experimenting with what does and doesn’t work.

And a predictive text generator like ChatGPT has probably ingested enough recipes to easily find patterns in which ingredients go well together, thanks of course to all the work human mixologists have already done.

To give you an example, we asked ChatGPT to create a cocktail recipe with a simple prompt: “Give me a recipe for a cocktail inspired by the radio show and podcast ‘Marketplace Tech’ and explain the reason you picked this.”

The chatbot suggested a drink called the tech innovation fizz, a play on the gin fizz, without the egg whites and with special additions like lavender and elderflower liqueur, which it said made for a “refreshing and engaging drink, just like the show itself.”

The recipe and instructions are below, if you want to try it for yourself. But, as the Atlantic piece points out, you may want to proceed with caution regarding its garnish suggestions because it gets a little too creative.

AI-generated recipes, like many AI outputs, probably need some fine-tuning, so if you do try it and have any feedback, you can email us at MPTech@marketplace.org.

The tech innovation fizz

Ingredients: 

• 2 ounces gin (representing the innovative spirit) 
• 1 ounce fresh lemon juice (symbolizing the clarity of information) 
• 0.5 ounce elderflower liqueur (for a touch of modernity) 
• 0.5 ounce simple syrup (to balance and sweeten) 
• 2-3 dashes of lavender bitters (representing the unique and tech-forward perspective) 
• Club soda (for the effervescence) 
• Ice cubes 
• Lemon twist or lavender sprig (for garnish) 

Instructions: 

• In a cocktail shaker filled with ice, combine the gin, fresh lemon juice, elderflower liqueur, simple syrup and lavender bitters. 
• Shake the mixture vigorously for about 15 seconds to ensure it’s well chilled and properly mixed. 
• Strain the cocktail into a highball or Collins glass filled with ice. 
• Top up the glass with club soda to add fizz and a refreshing quality. 
• Garnish the cocktail with a twist of lemon or a sprig of lavender for an elegant touch. 
• Optionally, you can also add a small piece of tech-related garnish, like a flash drive or microchip, as a playful nod to the theme.