How Editors and Employers Should Evaluate the Use of AI in Writing

Artificial intelligence has irrevocably modified human ingenuity.

All forms of creation and work are now either achieved or improved through its assistance. The trajectory of this trend is one of rapid ascendence –there’s no telling what the apex will look like until it is met.

AI in the Writing Economy

Writing is of the most AI-effected trades in the creative economy.

Chat GPT can, in essence (if the measuring stick here is its writing proficiency), outsource most low-paying Upwork-type writing gigs.

These gigs are essentially advertisement pieces for highly material products or convince-providing services.

Hence, there isn’t a lot of substance or complexity involved, such that it would be above Chat-GPT’s pay grade to complete.

The Real Capabilities of Chat-GPT

However, it cannot yet write anything of its own accord that is of higher quality than this, and by “of its own accord,” I mean without profound manual configurations to its outputs (i.e., edits by humans) and well-comprehended inputs.

Indeed, even Upwork writing gigs will still require someone to prompt Chat-GPT.

When Chat-GPT can write sui generis books unprompted, we’ve become conscious digital beings.

Thus, Chat-GPT cannot yet replace a writer –though it is potent enough to be a contributing factor of concern motivating the ongoing writer’s strike in Hollywood (in addition to, and more centrally about demands of higher pay.)

Misconceptions and Media’s Role

Most of the media rabble about artificial intelligence and its stealing of writer jobs is but a one-sided amplification of unfortunate cases where companies have decided to devalue their workers –not that there’s any shortage of devalued workers, just not a surplus of devalued writers who weren’t any less devalued in the first place (writers have always been strange enough!)

Detecting AI Influence: Tools and Ethics

Nevertheless, there is a way to vet writers based on their relationship to Chat-GPT that hitherto hasn’t –to my knowledge — been directly addressed.

Of course, suppose an employer’s writer’s works show up on ZeroGPT as written mainly by artificial intelligence. In that case, that cannot occur in any other way than downright plagiarizing from Chat-GPT’s outputs.

It isn’t unheard of for human-written text to show up as having been partially written by artificial intelligence on ZeroGPT –this is simply because the patterns it is programmed to detect happen to be patterns that come naturally to some writers.

Such partial cases should be carefully considered –i.e., immediate scrutiny should be resisted.

Hence, AI-detecting software is one precise measure to negatively evaluate a writer’s performance and almost certainly a criterion for job termination insofar as a text is detected as having been extensively or written mainly by AI.

AI-detecting software is the most popular among competent solutions to the issue of how editors and employers should evaluate the use of AI in writing.

Due to the limitations of such software, however, other evaluation routes require consideration insofar as it is of value to editors and employers.

Reevaluating Portfolio Assessment

One such route is the track record of the writer in question –namely, whether or not they have published anything before the release of Chat-GPT.

At this point, Chat-GPT can be used to successfully write content at a high level –e.g., a NYTimes editorial — insofar as writers (1) know how to prompt it carefully (i.e., they feed it the material it is to give form to through writing, as opposed to expecting it to have the material already based on a summary idea), and (2) insofar as writers edit its outputs meticulously.

In other words, a mediocre writer can now readily pass themselves off as a much better writer if they can become proficient in this technology.

To bypass this for selection, looking at a portfolio, editors, and employers can check if the writer published any works before 2022 when Chat-GPT was released to the public.

The only other option editors and employers can revert to for verification of their writer’s non-mediocrity is the implementation of surveillance software –something no one can reasonably advocate for without also reasonably taking up the valuation of a “tyrant.”

The Ethics of Improvement and Fairness

Nonetheless, there is a tone of tyranny –by way of undercurrents of paranoia- in checking to see whether a writer has published anything before 2022.

Shouldn’t we praise the fact that a technology exists that can turn a bad writer into a better one?

Shouldn’t this demonstration of self-improvement be rewarded? Or should it be viewed as an unfair advantage?

Ultimately, it is neither –so we’re bound to vehemently (i.e., erroneously) take sides here.

The Dichotomy of Creativity and Commerce

If we are speaking about pure creativity –something not generally produced in widely disbursed media — then Chat-GPT can only be a source of augmentation and, thus, something to be encouraged for use.

It cannot overhaul such creative purity; if one believes otherwise, that is only demonstrative of their lack of sense therein.

In contrast, if we are speaking about creativity in business, capital, and trade, it is absurd to talk of “unfair advantages” because that entire context presupposes unfairness –hence the writer’s strikes!

To put it candidly –no one likes work (we’d all be better off engaging in craft and with friends & family).

Still, if one refuses to work, the response from society is a loss of that individual’s value, such that eating, drinking clean water, and having shelter become increasingly more difficult.

So, if it is deemed unfair, it will not be done by those who are at a disadvantage but by those who run companies, editorials, and the like (i.e., those essentially at the helm of a system that makes the individual value as precarious as just described.)

Power Dynamics in the Age of AI

To nit-pick a portfolio for samples before 2022 might be an effective way to weed out those who almost certainly use Chat-GPT; it is a method that is as questionable as it is effective –those with samples before 2022 also use Chat-GPT!

Its questionability rests in the fact that, from the perspective so outlined above, it rings strongly of a petty grab for power among those with already enough — it’s as if they’d known their impotence in attaining their surplus, as those with true abundance (i.e., that which is earned through hardness, though not unkindness) look to see it in others, not take it away!