Here Are the Jobs Our New AI Overlords Plan to Kill

Could AI text and image generators replace human artists? That’s the million-dollar question. As the world marvels at the advent of tools like OpenAI’s image generator DALL-E, it can’t help but seem possible, if not likely. So far, programs like DALL-E have mostly served as novelties (at Gizmodo, we’ve used them to do a lot of dopey stuff, like insert U.S. Presidents into the world of Game of Thrones or spin up nightmarish modernist paintings of Seinfeld characters). But as the programs get progressively better, a lot of concerns exist about the kind of impact “generative AI” could have—and many onlookers have found themselves wondering whether this may be the end of human art as we know it.

That said, Max Chafkin over at Bloomberg has argued that “artists and designers aren’t about to lose their jobs to AI.” His reasoning? The current generation of image generators don’t seem to work too well...yet.

One of the industries that AI chatbots are most likely to “disrupt” is customer service. Most websites already have chatbots to answer questions, but human operators are still frequently needed to assist customers with more complicated problems. However, as companies begin deploying sophisticated AI-driven companions like ChatGPT, will humans still be necessary? Some have predicted mass automation is coming to the customer service industry. Whether the service will be actually any better is, of course, a hanging question.

Content creators of all stripes may see their heads on the chopping block in the years to come, but one of the professions most vulnerable to automation is copywriting for advertising and websites. Automated solutions have already begun to pop up. Just check out companies like Jounce AI, currently enjoying its private beta launch, which offers “unlimited free AI copywriting.” Now that even more powerful algorithmic platforms like ChatGPT are seeing their debut, many predict a swell in automation for the copywriting industry.

Yes, it’s true: AI-hype men maintain that lawyers could some day be replaced by chatbots. Does it sound insane? Yes. Is it insane? Definitely. Still, that’s what people are saying: Reuters recently reported that robot lawyers may be a thing in the not too distant future, explaining chatbots’ apparent ability to accurately draft legal briefs. Then there’s the legal services company DoNotPay, which has bizarrely offered $1 million to any lawyer willing to let its “Robot Lawyer” algorithm make their legal argument for them in front of the U.S. Supreme Court. DoNotPay’s chatbot is set to argue a traffic ticket in real court on Jan. 13. Suffice it to say, doing legal defense via algorithm sounds like one of the worst ideas I’ve ever heard but such is the way of the future.

Okay, okay—maybe scientists themselves may not be replaced, but science writing could be automated in the near future. That’s according to a recent study from Northwestern University, which claims that even scientists themselves couldn’t often tell the different between scientific abstracts written by humans and those written by ChatGPT. The study showed that humans could only correctly deduce whether a human or robot wrote the abstracts about 68 percent of the time. If this were a science test, that would be a D!

Some onlookers have wondered whether AI chatbots are one the verge of automating large parts of the computer programming workforce. Recent years have seen the emergence of platforms like AlphaCode, which is said to be able “compete with humans at solving simple computer-science problems.” At the same time, chatbots like ChatGPT have also shown promise in being able to automate certain parts of the coding process.

But, again, ChatGPT doesn’t seem to be able to do this job very well yet. The chat app was banned from Stack Overflow for its penchant for spouting off about stuff that sounded correct but was actually garbled nonsense.

Kim Kardashian already looks like a robot, but apparently real robots could make inroads to the fashion industry over the next several years. When I say “robot” I mean AI-fueled computer generated images of realistic looking humans, not Rosey from The Jetsons. Large retail chains have already flirted with the idea of CGI models to sell clothing and, if you Google “generative AI in fashion” you’ll definitely find a lot of nerds champing at the bit to find commercial purposes for this new technology. From a business standpoint, this makes sense: why hire an expensive good looking human when you can just used software to digitally create one?

Given everything we’ve just covered, us journalists would really prefer it if AI chatbots stayed the fuck away from our industry. There are already enough problems and layoffs. Despite this, there has been some amount of banter about whether bots like ChatGPT could ultimately be used by newsrooms to “augment” news-writing operations. Since ChatGPT is good at aggregating information at lightning speed, some have suggested that it could be useful for repackaging content that has already published elsewhere.

That said, we recently tried to get ChatGPT to write one of our articles for us and it sorta shit the bed...so, maybe we won’t be too worried for our job security just yet.