Subscribe to Slatest for a daily dose of insightful analysis, critique, and expert advice right in your inbox.
This article was crafted entirely by a human writer. While I didn’t receive assistance from A.I. for outlining or proofreading this piece, I view employing tools for spelling corrections similarly to how I would use A.I. tools—there's no moral dilemma in that for me. At Slate, we are fortunate to have a dedicated editing team, a benefit that not all media outlets can claim in today’s landscape.
It’s possible that I might not have written this story without the influence of A.I. However, it’s not due to the technology performing tasks that only a human could accomplish. Rather, A.I. has streamlined processes that previously caused minor hurdles for my workflow.
My baseline skepticism about A.I. remains, especially since it has been wielded destructively by individuals. I maintain a wary view of figures like Sam Altman and his contemporaries. However, I’ve also made a conscious effort to approach the subject with an open mind; the advancements in technology are hard to ignore. After discussing A.I. with Ed Zitron, a well-known critic, in February 2025, I found resonance in his take that while chatbot applications drew a large user base, they still lacked practical applications that would make them indispensable to those not deeply invested in mastering them. However, spending some time with tools like Claude Code or Codex has shifted my perception. If you let the software interview you about your preferences, it can create a portfolio website based on your responses in a matter of minutes. It’s clear that these technologies don’t need to overhaul daily life to be revolutionary.
If you had asked me a year ago about my use of A.I. in writing, I would have dismissed the idea. Nowadays, though, it plays a role in my daily routine, albeit in ways that diverge from the narratives around job replacement or significantly reducing stress in the journalism field.
I often see influencers on Instagram claiming that A.I. programs like Claude have transformed their lives, enabling them to accumulate thousands of followers in a short span. More often than not, I suspect these assertions are exaggerated or outright false. Similarly, claims that new software releases have made certain professions obsolete appear exaggerated to me. The most beneficial applications of A.I. in my field are subtle but still effective.
Transcription remains one of my most significant challenges in storytelling, and A.I. has made remarkable strides in this area recently. I frequently rely on transcription services like Rev, which allows access to both the A.I.-generated text and the corresponding audio. I’ve also utilized a local model on my computer that adequately handles shorter conversations.
The method of uploading transcripts to Claude, instructing it to identify specific statements regarding an issue—especially for lengthy interviews—is quite advantageous. It’s crucial, however, for anyone utilizing these transcripts to verify the selected sections for credibility. Though Claude and Codex have improved in accuracy, they still aren't infallible. Having an A.I. assist in pinpointing relevant parts of a transcript saves considerable time, ranking among my favorite technological innovations throughout my career.
On a similar note, dictation tools have also advanced significantly this year. Apps like Wispr Flow and Monologue (the latter being my preferred option) leverage A.I. effectively. I find such applications immensely helpful for quickly capturing ideas and outlining while on the go. Now, I can dictate voice memos during a leisurely walk and instruct my device to save them in an accessible format later. Given my track record of clumsiness with typing while walking, this approach is a welcome improvement.
A large portion of my recent articles has originated from a Notion database I call “Idea Dump,” which primarily stems from me verbally sharing thoughts with my phone that later forward to my Notion account. I created a simple A.I. script that organizes my ideas with appropriate tags to facilitate easy access later on.
Despite these advancements, the core aspects of my work that I previously valued remain unchanged. I am not seeking to replace the intensive and creative process of writing, but rather to streamline how I record and organize thoughts when the opportunity arises.
This workflow effectively tackles one of my historical weaknesses: generating exciting ideas only to forget them or jot them down in indecipherable formats on my device. The integration of machine-learning dictation and advanced A.I. has significantly mitigated this struggle, evidenced by the multitude of coherent ideas I’ve saved compared to the few nonsensical notes I came across in the past. In fact, A.I. has even helped me retain some less-than-stellar ideas that seemed promising in the moment.
The technology has also had a practical impact on my daily routine. This year, I’ve embarked on a weight loss journey, losing 17 pounds so far, and adopted a kanban board approach for meal planning. I tend to stick to a system when visuals are laid out clearly. However, I often struggle with shopping efficiently—forgetting or duplicating items. By using Codex, I integrated a shopping list into my project management framework that streamlines recipes into my weekly meal plans. I don’t believe my grocery list access should determine stock market trends, but the 10-minute setup helps avoid kitchen oversights.
While I know I’m not exceptional in navigating this tech without risks, my apprehension comes from potentially losing cognitive sharpness rather than job security. I’ve cautiously employed an A.I. note-taker for select meetings with proper consent, aiming to better engage with conversations instead of focusing on note-taking. Ironically, I’ve found that relying on the A.I. has sometimes reduced my attentiveness, making me perceive the tool as a safety net—drawing parallels to how students may become overly reliant on A.I. for their assignments.
Regrettably, A.I. did not rescue me from the hours I spent thinking through and penning this article. My dictation attempts fell short in capturing my unique writing style, a disappointment since I aim to communicate as I would with a close friend. Perhaps this will change in the future; perhaps it won’t.
Through these months of experimentation, I’m more convinced than ever that automated A.I. will not supplant human writers in renowned publications, no matter how many executives attempt to push those boundaries. Even if machine-generated content passes a test for quality, audiences typically recoil upon discovering they’ve been reading a machine’s output. I share this sentiment because my passion for writing is rooted in the activity itself. The most promising scenario for writers in tandem with A.I. is that we gain more time to grapple with our thoughts, freeing ourselves from laborious tasks like combing through transcripts at slowed playback speeds. While we cannot eliminate the difficulties inherent in writing, we can reduce the more tedious aspects, allowing us to direct our focus and emotions elsewhere.
