An author asks: Can AI Replace Humans in Publishing Sector?

---

 “Robots will be able to do everything better than us”—Elon Musk, 2017

Like fashion, artificial intelligence (AI) is a topic that rides high and low tides. Lately it’s been getting more attention on-line, on-air, and in print:

• Internet giants Google and Microsoft are duking it out to attain AI search engine domination.
• Questions abound regarding dangers of AI facial recognition software (ethics, legal and social policymaking, law enforcement protocols, individual identity and privacy concerns).
• Intellectual property rights are being debated over whether AI-generated writing, music, art, etc., are, legally speaking, stolen fruits of crime.

These AI applications are so pervasive that we take them for granted:

• virtual assistants (Siri, Alexa, Cortana, GPS)
• voicemail, call centers, customer service
• auto-pilot systems, self-driving vehicles, traffic lights
• healthcare billing and documentation, robotic surgery
• image, speech, fingerprint recognition
• dining, dating, social media platforms
• media streaming (movies, video games)
• security (residential, commercial, military, consumer)
• language translation
• autocorrect and autofill (often with unintended results)

How does any of this relate to human content creators?

Whether it’s fiction or nonfiction, writing doesn’t come easily for all writers. Although we’re driven by different incentives and by mercurial muses, generally the process entails these actions:

• create
• edit (again and again)
• rewrite (again and again)
• send/submit

Each step incorporates other actions:

• developing and organizing—Does content make sense?
• proofreading and copyediting—Is content accurate and consistent?
• following submission guidelines—Do word count, formatting, and file type meet requirements?

Putting aside certain types of copyediting and formatting (both mechanical functions), can AI produce the kind of writing a human brain would (a brain, mind you, that’s capable of emotion)? Edit and re-edit so the message remains organic and true to the (human) writer’s intent? The last step, hitting send/submit, is a mechanical action (often fraught with emotion).

What is AI?

Most of us have a general idea despite definitions that often trip over their own words. The most straightforward one I’ve found is from John McCarthy, a computer scientist and one of the AI founders. In the mid-1950s, he coined the term artificial intelligence, defining it as “the science and engineering of making intelligent machines.”

   For me, the notion that machines can replicate human cognition, temperament, experience, memory, and sensibility—core writing components—is perplexing (a humantrait).

   Can AI “writers” identify and address ethics surrounding social and legal policymaking, business hiring and employment practices, cultural diversity? After all, humans still grapple with negative conscious and unconscious biases (gender, religion, food, age, racial, ethnic, body, beauty, name, weight, and height) while writing software programs, news and magazine articles, movie and TV scripts, and textbooks.

   McCarthy’s definition suggests that AI—virtual and mechanical—is programmed to process information, solve problems, and provide services like a human would. If that's the case, who are AI’s creative writing muses? Algorithms, metrics, data inputs, cognitive and behavioral cloning?

Implications for writers

AI is prevalent in B2B communications, sales and marketing copy culled from software templates. Publishers and advertisers draw on it for content, book cover art, and illustrations. Ditto for medical, legal, science and technical writing environments.

   Paralyzed by writer’s block or deadline desperation? AI to the rescue!

Two German companies, Trilogy and Neuroflash, claim to produce AI-generated fiction or nonfiction, as well as clinical and technical content. AI Book Writer is one example. Neuroflash further claims its technology “. . . may even be able to produce a better book than a human author, since it is not limited by the same cognitive biases and limitations.” (Biases, conscious or unconscious, are not inherently negative—but that’s another topic.)

What about translation writers

Most writers know this, but I’ll say it anyway. Translators work with the written word, whereas interpreters work with the spoken word.

   Literary and business translation writers identify and incorporate linguistic, cultural, and dialect nuances because they have legal and ethical implications.

   According to Meta, “We’ve built a single AI model called NLLB [No Language Left Behind]-200, which translates 200 different languages with results far more accurate than what previous technology could accomplish.”

   It’s not clear to me whether or how Meta’s model handles business and legal translations. What’s more, as far as I can tell there’s no universal and comprehensive AI regulation or oversight in the U.S. publishing sector. Meaning no standards guide liability, ethics, or copyright infringement and ownership conflicts—all legal minefields.

Writing software

I often rely on AI-based plagiarism and cliché detectors, name and title generators, sometimes plot generators. I try to avoid depending solely on writing software (you know the names). Call me snooty or vain, but I prefer relying on my brain (and my finicky muse) to generate ideas and turn them into words meant to connect with readers, aka the writing process. I worry that quick fixes will sabotage that process by obstructing efforts to keep my noodle lubed. When fed a diet of algorithms and metrics, can robots

• experience authentic empathy and think critically?
• grasp language subtleties and read between lines?
• interpret organically what emerges from the creative psyche? (Do robots even have psyches or souls?)

Finally, as a writer and self-diagnosed neurotic, I’d be surprised if AI techs can program neuroses as creative writing bytes. I choose not to go down without fighting against Musk’s prediction or knowingly, as somebody said, be the architect of my own demise. Artificial intelligence, after all, is engineered by human intelligence—warts and all.