(Credit: NurPhoto / Contributor / NurPhoto via Getty Images)
Existential dread ran deep in the tech world this week. Programmer Matt Shumer's viral column about AI disruption made the rounds on social media, racking up a reported 30 million views. He flips the script on the industry's obsession with coding, arguing that more flexible critical thinking is now the most important skill for the next generation.
Even Elon Musk is reevaluating, admitting that his space travel dreams are a bit too ambitious. At least we're saying goodbye to one of OpenAI's most infamous models today, GPT-4o. That's the one that allegedly fueled AI psychosis, delusions, and suicidal ideation in some users.
Read on for all the headlines you need to know in this week's AI wrap-up.
Oh, the Irony: What Happened to 'Learn to Code—Or Else'?
Shumer is the 26-year-old CEO of HyperWrite, an AI writing assistant. He doesn't have the same name recognition as his fellow AI CEOs, but this week, he took a page out of their playbooks by positioning himself as an industry insider with a big warning for the rest of us. In a confession-style post on X, titled "Something Big Is Happening," Shumer argues that AI has made his programming skills unnecessary, and warns that AI is coming for everyone else's industries, too.
"I am no longer needed for the actual technical work of my job," he writes. "The experience that tech workers have had over the past year, of watching AI go from 'helpful tool' to 'does my job better than I do,' is the experience everyone else is about to have." He lists other industries he suspects will hit the fan in the next five years—law, finance, medicine, accounting, consulting, writing, design, analysis, and customer service.
It's a common refrain we've heard from AI founders since ChatGPT debuted in 2022. Last year, Anthropic's Dario Amodei warned that AI will automate 50% of white-collar jobs in the next five years. At the World Economic Forum in January, Amodei upped that to 100% of all software engineering jobs within 12 months. But I should note that he made a similar prediction last March, when he said it would happen in three to six months.
That's why sometimes it seems like AI companies lean into existential dread as a roundabout way to brag about how powerful their technology is, something Ben Affleck, of all people, called out on a recent episode of The Joe Rogan Experience.
This doomsday talk is incredibly ironic from my perspective as a liberal arts major. Over the last 10 years, I've been hammered with the opposite message: Learn to code or become obsolete. Or at the very least, make less money than those in high-profile tech jobs. But now, they might have to take their own advice and find a new line of work.
What I'm hearing from Shumer is that liberal arts skills are now the most valuable currency in the white-collar world: critical thinking, teamwork, self-direction, and the ability to grasp big-picture concepts. That's what turns a programmer from a drone into someone who can design, iterate, and collaborate on a project—even if their AIs execute it while they babysit.
"The thing that will matter most for the next generation is learning how to work with these tools, and pursuing things they're genuinely passionate about," Shumer says. "Teach your kids to be builders and learners, not to optimize for a career path that might not exist by the time they graduate."
That sounds like a flexible, foundation-setting liberal arts degree, right? Not a skill-specific STEM concentration. At the same time, others say blue-collar work, including skilled trades like construction and HVAC, will become the most sought-after jobs. It depends on who you ask.
Maybe we're all just trying to rationalize not burning ourselves out at an AI startup in the hopes of getting rich. The truth lies somewhere in the middle, as it usually does. I agree with Shumer that it's increasingly important to understand how to use AI to improve your skills and keep evolving. As he puts it, "Start taking AI seriously." Thankfully, working at PCMag keeps me up to date on the latest tech, so I'm constantly trying new things. And while Shumer says "writing" is among the threatened professions (he admits to using AI to help him write his viral post), journalism has also evolved past writing to other formats (video, podcasts, and more).
The question is: What's the best foundation for the next generation to build from? A humanities mindset or a more technical one? It's not a coding bootcamp anymore, that's for sure. It's whichever one compels you to fact-check an AI summary, see through its sycophancy, and stop to notice that the loudest warnings about AI's power can sometimes come from the people who profit from frightening you into using their products more.
Major Headlines You Need to Know:
- Following SpaceX's acquisition of xAI, Elon wants to build a city and put AI data centers on the moon, putting Mars on the backburner for now. No, that's not feasible today. For one, you can't buy land on the moon, per a 1967 treaty. It also raises serious questions about the economics of building in space, latency to beam down to Earth, maintenance/chip upgrades, and so much more. Admittedly, the unlimited solar power and cool atmosphere do sound appealing.
- Turmoil in the ranks at xAI. Musk's AI company saw several high-profile departures this week, including from several co-founders. Musk tweeted that he's re-organizing the company, and is still "aggressively hiring" people who want to put stuff on the moon. (A researcher at Anthropic also made waves for his resignation letter: "I've repeatedly seen how hard it is to truly let our values govern our actions.")
- The White House seeks a solution to rising electricity prices from data centers. Meanwhile, Anthropic claims it will cover its own costs, not put them on taxpayers, as Microsoft announced last month.
- Ads are officially rolling out in ChatGPT. It's the end of an era of interruption-free content. You can avoid them (without paying), but it comes with a trade-off.
- OpenAI kills GPT-4o. The model that became everyone's corny wing person is officially gone as of Feb. 13. CEO Sam Altman admitted the tech was too agreeable, to the point of being dangerous for mental health. Despite pleas to bring it back from those who got hooked, Altman is moving forward with other flagship models, most recently GPT-5.2.
- If you saw the AI.com ad at the Super Bowl, don't sign up for it. I did, and instantly regretted forking over my credit card information for a site that doesn't work, and won't back me up in court if one of its AI agents commits a crime.