Amy Maoz: “Don’t let the word ‘destabilize’ in the headline trick you into thinking this is a fear-mongering piece. Annie Lowery’s nuanced approach and thoughtful choice of interview subjects (something she notes ChatGPT specifically can not do) left me feeling more hopeful than I thought possible.”
It’s tempting to ask ChatGPT to write an introduction to a reading list about ChatGPT, but I went the old-fashioned human route for this one. Not because of journalistic integrity or to avoid a increasingly-tiresome shtick, but because, as OpenAI told me when I tried, ChatGPT is at capacity right now. Too many old-fashioned humans had gotten there before me.
That makes sense: ChatGPT is dominating the headlines, often for its potential to destroy industries, professions, and possibly democracy. In some cases, these concerns are easy to refute, especially as we layer the technology with human innovation to create something new and special (see the NPR piece below for more on this idea). But there’s also a vague sense of dread lurking as some of the tech world’s top thinkers consider what a world filled with generative AI might mean ethically and existentially.
If this leaves you feeling a little uneasy, you’re not alone—and not without resources. We’ve gathered 10 fascinating stories that investigate the different ways ChatGPT has quickly seeped into our lives and economy, and which fears are most (and least) unnerving.
Image by Westend61/Getty Images
AM: “Legendary music journalist Simon Reynolds calls ChatGPT’s attempt to steal his job ‘earnest, plodding, attuned to bland generalities rather than arresting specifics, and irritatingly fair-minded’—harsh words for the machine, but a relief for journalists concerned by the threat of machines taking their jobs.”
AM: “Oops did I just say journalists can feel relieved? I read Mia Sato’s investigation of how a venture capital-backed publisher was relying on predictive text models the way I watch horror movies—with my hands over my eyes, desperate to turn away but too fascinated. I won’t spoil the last line but it certainly nails the issue that underscores most fears of ChatGPT.”
AM: “John Warner is not an algorithm; he notes that in his bio on his Substack ‘Biblioracle’ where he recommends a book based on a person’s last five reads. To answer the question in this Collection’s headline, Warner is not afraid of ChatGPT—in fact he sees it as a jumping off point for reevaluating the systems we live with, the very ones being exploited by the tool in the first place.”
AM: “The Help Desk at The Washington Post suggests using ChatGPT as a tool to help get yourself started at something, rather than outsourcing the entire task, and as a result, knocks the tool down a few pegs from the all-powerful machine some make it out to be. It’s a sentiment best summed up by Post commenter My Name Is Jim: ‘If Chatbox can cure pancreatic cancer, fine. If not, get lost.’”
AM: “Google’s concerns about ChatGPT are different than, say, a third-grade teacher’s: Teachers are not scared that ChatGPT will threaten their $149 billion search business. Google, meanwhile, seems to be all hands on deck in ensuring that their ChatGPT-adjacent releases have the proper guardrails in place to keep them from powering the kind of hate speech and misinformation that thrives on the internet.”
AM: “Now we’re getting into the more pressing reasons to be concerned about ChatGPT, with Alex Kantrowitz giving a quick and sobering overview of the ethical issues generative AI forces us to face. For instance: What’s a chatbot’s stance on war—or the death penalty? As Kantrowitz reminds us, ‘AI’s intelligence may be artificial, but humans encode its values.’ And how those values will be shaped over time is still unknown.”
AM: “Edward Tian went from using ChatGPT to write poems and raps about his friends (relatable) to create and release an app that allows users to identify if their break-up email/homework assignment/cover letter was written by a machine (less relatable.) He calls it GPTZero and hopes for it to be a tool to promote and ‘incentivize originality in human writing.’ While Tian’s achievement is incredible on its own, I look forward to tracking how many other brilliant minds find ways to deactivate some of the threats of generative AI while taking advantage of its usefulness.”
AM: “But hold your horses: Before getting excited about ChatGPT, it’s essential to consider the red flags spotted by Gary Marcus, a leading voice on AI (with more than a few books and machine learning companies to his name.) Here, and in his Substack post early December, Marcus lays out some of the worst case scenarios for how machine-generated text can influence and interfere with human emotions and actions.”
AM: “This piece contains an excellent, plain English explainer on how ChatGPT functions and highlights how we might see the tool as a jumping off point for some of our most creative thinkers.
It’s also important to note that the author, John Naughton, is hardly tuning out the bad stuff—he recently posted about reading Gary Marcus’ work. Does it comfort me that he can consider these dangers but still see a 'mundane' future for this tool? Absolutely. Does it comfort you? I certainly hope so. Or at the very least that you've come away from this with some new ideas about what our ChatGPT-ified future might look like.”