What Medium writers think about AI tools

  • Thread starter Thread starter The Medium Newsletter
  • Start date Start date
T

The Medium Newsletter

Guest

A blind photographer + why writing is thinking (Issue #392)​


ChatGPT was released just over 1,000 days ago. In the time since, the change that AI is driving in the world around us has been a regular theme in this newsletter, from Taylor Swift deepfakes (issue #4) and whether AI can make art (issue #163), to fake academic papers (issue #265) and the meaning of AI selfies (issue #340).

We here at Medium have been wrestling with AI, tooโ€Šโ€”โ€Šfrom how AI-generated stories show up our readersโ€™ feeds to how we can or should incorporate AI tools in Medium itself. Last week, our CEO Tony Stubblebine laid out some principles on how Medium will be using AI moving forward (focus on human stories, protect writer incentives, give readers options), and asked our community of writers a simple but complex question: How can writers use AI to tell human stories?

We asked writers to respond using the #AIforHumans series, and this week, weโ€™re highlighting some of those responses. Reactions varied widely, but with much more nuance than a simple pro/anti AI divide. A few themes emerged. One was disclosure: A number of writers felt that arguing about whether AI writing is โ€œrealโ€ or not is beside the point. Linda Caroll notes that modern LLMs can already write at or above the level of novice authors, making detection unreliableโ€Šโ€”โ€Šfor her, the only meaningful line is disclosure.

Urna Gain is likewise concerned about detection, warning that if Medium leans too heavily on guardrails or flawed detection tools, genuine human work could be mistaken for machine output. โ€œIf I write about my miscarriage,โ€ she says, โ€œto have that kind of story misjudged as AI would not just be insulting, it would be dehumanising.โ€

Exhausted and frustrated, longtime editor Debra G. Harman describes how AI-generated work is flooding publication submissions, wasting editorsโ€™ time, and eroding trust. She sees AI usage as a slippery slope, and takes a hard line: โ€œItโ€™s no longer your own writing. Itโ€™s not okay in my publications.โ€

Another Medium editor, Sam Vaseghi, worries less about the slippery slope or detection, and more about homogenization: when too many writers lean on the same AI tools, their work begins to sound alike. Strong writing, he argues, depends on diversity of expression. If Medium isnโ€™t careful, he warns, it risks becoming a โ€œlinguistic monolith.โ€

Others see the issue more in terms of accessibility. Writer Jim the AI Whisperer explains how for him, AI is an accessibility tool rather than a threat to independent thinking. Living with aphasia, a language disorder that makes word retrieval difficult, he uses AI sparingly to surface words that get stuck. โ€œAI has assisted me <0.2% of the time,โ€ he writes, emphasizing that the ideas and sentences are his own.

Where do you land on the question? When thinking about how AI can improve Medium, what is most valuable to you? Respond here, or add your story to #AIforHumansโ€Šโ€”โ€Šweโ€™ll be reading.

โ€” Anna Dorn

Recommended reading:​

  • Lawrence Lazare explains how he continues to take photos after losing his vision to Stargardtโ€™s Disease. Digital cameras let him rely on muscle memory and autofocus, but he found that inauthentic, so he turned to film. Using manual focus, magnifiers, and even a homemade โ€œdigital assistโ€ system, he built a process that mirrors the reality of working with impaired vision. For Lazare, the goal is storytelling, not perfection.
  • Three hours, a sketchpad, and an AI-prototyping tool: Thatโ€™s all HBO VP of product design Michael McWatters needed to turn a half-formed idea for a streaming platform into a working design. Instead of grinding through traditional software, he wrote AI prompts to test and refine animations, interactions, and transitions in real time. The process, he says, felt less like engineering and more like painting: messy, unpredictable, and alive with possibility.
  • A fatherโ€™s stories can shape a childโ€™s whole sense of self, even when those tales arenโ€™t true. Essayist Ruchama recalls growing up on her dadโ€™s stories about car crashes, endless surgeries, and schoolyard humiliation. Years later, she discovers one of his most vivid stories never happened at all, leaving her to reckon with how the fabrication shaped her identity. Now her own daughter is listening just as closely.
  • Writing online isnโ€™t about racking up followers, argues Pavel Samsonov, itโ€™s about sharpening your own ideas and connecting with readers who care about them. If youโ€™re just chasing metrics or letting AI churn out posts, you end up with empty contentโ€Šโ€”โ€Šgarbage in, garbage out. His advice: Use writing to test thoughts in public, learn from the responses, and develop arguments that actually hold up.

In an era of global crises and polarization, journalism must bridge divides now more than ever​


Semafor is tailored to this reality, offering unparalleled transparency and diverse perspectives. That mission comes to life through Executive Essentials, a collection of four focused briefings on politics, business, technology, and global news. Trusted by heads of state, CEOs, and policy leaders, these briefings cut through the noise to deliver what matters most. Stay connected with the world around youโ€Šโ€”โ€Šsubscribe for free.

Deepen your understanding every day with the Medium Newsletter. Sign up here.

Edited and produced by
Scott Lamb & Carly Rose Gillis

Like what you see in this newsletter but not already a Medium member? Read without limits or ads, fund great writers, and join a community that believes in human storytelling.

stat



What Medium writers think about AI tools was originally published in The Medium Blog on Medium, where people are continuing the conversation by highlighting and responding to this story.

Continue reading...
 


Join ๐•‹๐•„๐•‹ on Telegram
Channel PREVIEW:
Back
Top