Photo by Etienne Boulanger on Unsplash
UNSPLASH.COM
Writing has always felt different from the other kinds of work I do. I started writing because I had ideas I wanted to share about topics like automation and bootstrapping. Years of experience had given me confidence in coding and building a business, but writing was another story. While working on one of my first articles, I discovered the power of a good editor: someone who could turn a lump of clay into a sculptural masterpiece—or at least into something I felt comfortable sharing. The ideas were still mine, but they were communicated more clearly and effectively.
Over the years, I’ve realized that most editors share certain ways of thinking. Recently, while scanning a headline about AI “workslop,” it dawned on me that the “editor mindset” is a great tool for approaching AI use. As LLMs, agents, and autonomous workflows explode, human judgment is becoming more valuable than ever. The most effective leaders won’t be the ones using AI the most, but the ones who know how to curate, refine, and reject its output as needed.
Here’s how to think like an editor when using AI.
After contributing to various publications, Forbes and beyond, I realized that each has a different objective.
“A good piece is a good piece, and a story well told is a story well told, and an investigative scoop is an investigative scoop. Those can work in many places … But each place has a different sensibility and point of emphasis,” notes The Atlantic editor Scott Stossel.
A good editor keeps the outlet’s goals in mind. In the same vein, a strong AI user constantly considers the destination of their work: What function will it serve within the organization’s larger mission? How can I help this AI tool—what information and context can I provide?—better serve that mission?
For example, let’s say I want to use ChatGPT to draft a newsletter announcing an updated version of one of our products. I’d create a prompt that includes the goal of the output (inform potential customers, build interest, increase sales), the company’s broader mission (making users’ lives easier with intuitive tools for tedious everyday tasks like building online forms), and the intended audience (solopreneurs and small- to medium-sized business owners). That context helps the LLM generate a much stronger first draft.
Writers understand the power of thoughtful feedback. It can render great ideas more fluid. It can make good writing even more compelling and poignant. It can convince readers, with ever-shortening attention spans, to make it to the end. Most writers even look forward to feedback, knowing it will improve their craft.
Today’s AI tools are increasingly sophisticated, retaining information from prior interactions and applying it to future tasks. ChatGPT and Claude, for example, can carry context across conversations and, in some configurations, retain preferences or project-specific information over time.
Or consider AI agents: one of the easiest ways to train them is through natural conversation, including pointed feedback. Let’s say a portrait studio owner is training AI agents to field customer questions about planning a photo shoot. The agent replies:
“To plan a photo shoot, start by defining the purpose and theme of the shoot, then create a detailed shot list and schedule. Next, select the right location, gather your equipment, and ensure proper lighting.”
It’s a good first attempt, but now it’s time to put on your editor hat: Make the tone more conversational and friendly. Higher level, this advice is geared toward a professional photographer—not a customer coming to the studio.
The more precise and thoughtful your feedback, the more useful the AI’s output will become.
There’s no doubt that AI tools like ChatGPT and Claude are powerful tools for productivity. With them, you can automate and speed up busywork and leave more time for more meaningful creative work that only you can do—tasks and projects that move the needle.
While their sophistication continues to advance, one fundamental issue persists: their tendency to make stuff up.
“Today’s A.I. bots are based on complex mathematical systems that learn their skills by analyzing enormous amounts of digital data. They do not — and cannot — decide what is true and what is false. Sometimes, they just make stuff up, a phenomenon some A.I. researchers call hallucinations. On one test, the hallucination rates of newer A.I. systems were as high as 79 percent,” reported the New York Times last year.
As any editor will tell you, fact-checking is the bedrock of good journalism and (non-fiction) writing. The New Yorker reportedly has 28 staff employees dedicated to the job.
To think like an editor while using AI tools, it’s indispensable to check all factual information. Follow the links to see if they really do support the research and claims cited. Ensure that any assertions aren’t overstating the case. Test code to make sure that it actually runs, produces the expected output, and handles edge cases rather than merely looking plausible at first glance.
AI tools can generate strong first drafts—text, webpages, chatbots, and more. Treating it like a collaborative tool, with you as the editor, not only improves the quality of the output, but it also can get you thinking more critically about the task at hand. It can make you a better writer, coder, designer, marketer, leader, and beyond, in the process.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。