I never anticipated having to express this, but here I am: my name is Peter, and I am a human being.
This statement may seem obvious, yet it has become necessary due to the increasing misuse of artificial intelligence, which is turning thoughtful opinion pieces into what I would call “slop-inion” that is infiltrating the editorial sections of reputable news organizations.
Recently, Crikey had to retract a series focused on leadership, and the features editor at Capital Brief voiced concerns on LinkedIn, noting that 80-90% of submitted articles seem to be generated by AI.
While plagiarism has long been considered a grave offense in journalism, presenting AI-generated content as one’s own crosses a significant ethical boundary. However, simply running a check for AI-generated phrases in the final draft isn’t sufficient; the nuances are complex. What happens when AI conducts primary research, suggests unique angles, or identifies logical flaws? At what point does the output no longer qualify as human?
As a vocal critic of AI, I have been hesitant to incorporate this technology into my writing. Yet, recognizing the importance of understanding the opposition, I spent the past month exploring Anthropic’s Claude to see how it could potentially enhance my writing process.
To navigate this exploration, I have taken guidance from former Australian Chief Scientist Alan Finkel, who has initiated a global certification program aimed at ensuring creators can “verify” their work as human-generated. I am pleased to announce that I am the first “Proudly Human” columnist, joining a growing community of authors, musicians, and publishers who have also completed this formal accreditation.
The Proudly Human initiative employs a “de minimis” standard to confirm human authorship, a legal principle that defines acceptable inputs that do not compromise a creator’s copyright. This framework delineates what can be utilized to assist a creator’s work while ensuring it remains distinctly theirs. For instance, using AI for grammar checks or brainstorming is acceptable, but it draws the line at generating text or content that significantly contributes to the final piece.
Moreover, Proudly Human advocates for a comprehensive set of prompts to ensure that writers are not automating fundamental aspects of storytelling and creative content. This concept of “provenance” guarantees that it is indeed me communicating with you through these words, similar to efforts to authenticate the creation of First Nations art in the context of exploitation.
Within these parameters, I was curious about how an AI assistant could enhance my writing routine. With two weeks left before submitting my monthly column for the Guardian, I planned to assess the political landscape and develop questions that would resonate with the national sentiment. I would immerse myself in various cultural influences to seek a compelling metaphor.
During the next fortnight, I would begin constructing my argument and glance at preliminary findings to ensure they aligned with my thesis (sometimes, when they don’t, I am compelled to reassess everything, leading to some of my best columns). After drafting, I would share the piece with trusted individuals (including my wife, who knows my work intimately), craft a narrative around the findings, polish it up, and submit it to the editors at Guardian Australia for their review.
By operating within a de minimis framework, I could streamline much of this process, employing the AI assistant to verify my thesis, identify connections in the data, and even suggest cultural references. Once I completed a draft, I could input it into Claude and inquire about any logical gaps, receiving immediate feedback. Thus, one could argue that with the help of an eager research assistant and a supportive checker, my productivity increased.
However, I discovered that the elimination of intellectual friction diminished the quality of the final product. For me, the drafting process is essential; it involves problem-solving, grappling with self-doubt, and the realization that by the fourth version, the piece starts to take shape. Writing transcends mere word generation; it encompasses choices, thoughtful decisions, and the challenge of refining ideas to life.
In a landscape awash with content, it is vital to distinguish human creativity from machine-generated output. A human “certification” not only serves as a testament to my commitment but also acts as a safeguard against the temptation to seek shortcuts when faced with writer’s block. I genuinely do not wish to simplify my work; the challenging aspects are what make it worthwhile. The mistakes I make are just as significant as my successes; taking the easy route feels akin to using a player piano: it may sound pleasant, but I am merely going through the motions.
For those of us concerned about the ramifications of AI, it is not enough to simply express discontent; we must make conscious choices daily. An authenticator like Proudly Human represents not only a pledge from creators but also a proof of authenticity that consumers should demand if they genuinely care about the societal landscape emerging from this so-called revolution.
Cultural content is at the forefront of the AI challenge, as the appropriation of intellectual work threatens to replace creative professionals. If we fail to establish clear boundaries here, I fear there will be no boundaries left at all. Provenance is crucial.
Peter Lewis is an accredited Proudly Human writer and serves as the executive director of Essential, a progressive communications and research firm, as well as the host of the Burning Platforms podcast. No AI was used in the creation of this article.
