Generative AI is Coming for Publishing, But It's Not as Clear-Cut as It Looks

Written by brianwallace | Published 2023/03/07
Tech Story Tags: ai | future-of-ai | ai-trends | conversational-ai | generative-ai | publishing | machine-learning | hackernoon-top-story | hackernoon-es | hackernoon-hi | hackernoon-zh | hackernoon-vi | hackernoon-fr | hackernoon-pt | hackernoon-ja

TLDRChatGPT is a chatbot created by OpenAI which is available to the public. It can be used to generate text, tackling anything from simple queries to complex essays. The rise of tools like Grammarly led to a gutting of copywriters and editors. The decision to deploy ChatGPT and other AI generative tools poses practical problems for publishers as well.via the TL;DR App

Generative AI has taken over the news for the past few months. First, in the visual arts, and then in more traditional text-driven publishing with the release of ChatGPT-3. ChatGPT is a chatbot created by OpenAI which is available to the public. It can be used to generate text, tackling anything from simple queries to complex essays. 
For writers and other creatives whose work hinges on consuming a lot of content, interpreting it, and creating original work from it, it has been concerning. The rise of tools like Grammarly led to a gutting of copywriters and editors.
With writers now in the firing line, there’s a lot at stake. At the same time, the decision to deploy ChatGPT and other AI generative tools poses practical problems for publishers as well.

A Miraculous Time Saver  

There's no denying that generative AI can be a cost-saving, time-saving boon for publishers.
Red Venture's media brand, CNET, admitted as much when stealthily deploying AI writing tools as part of a small pilot test for its CNET Money brand.  
"Will this AI engine efficiently assist them in using publicly available facts to create the most helpful content so our audience can make better decisions? Will this enable them to create even more deeply researched stories, analyses, features, testing, and advice work we're known for?" CNET's Editor-in-Chief, Connie Guglielmo, said.
"We're already testing ChatGPT and GPT-3 to support our content and SEO strategies for our publisher clients," another CEO, Paolo Dello Vicario, CEO of ByTek, said. "It's very useful. Since our proprietary AI already tracks popular topics and optimizes headlines, ChatGPT can then supply a ready first draft of content. This saves our publisher clients time and money." 
It's not an idea without merit. For creatives, there is often a lot of busy work in writing that can simply be outsourced. If you're working on content for a food blog, there are only so many ways to boil an egg, to deploy an oversimplification. If you're a tech journalist writing about a phone, the specs remain the same no matter who is writing them. If a film journalist, there are constants like cast and crew. Generative AI like ChatGPT can instantly create content that leverages all the information you'd be working on.
However, all generative AI is, at this point, a tool for regurgitation. The word generative itself is a simplistic misnomer, as is AI. It is not intelligent nor can it generate. It can only truly ever imbibe and regurgitate what other people have created without the ability to understand contextual nuances or create something new of its own. 
Dan Gaul, co-founder, and CTO of the commerce-focused media conglomerate, Digital Trends Media Group, said: “It could be useful for starting rough drafts or fact-finding that would need to be checked. If publishers rely on generative AI to write content completely without humans, then I think they are doing a disservice not only to their audience but to themselves."

Publish in Haste, Apologize at Leisure 

Returning to CNET's experiment with Generative AI, the body seemed to have fallen for that trap.
Several of its CNET Money articles were found to be riddled with factual errors and called out by several news outlets. The company has now added a statement to articles published under the pilot program. It is an embarrassing (and potentially dangerous, considering the subject matter) incident that publishers hoping to deploy AI would be best positioned to learn from. 
“It’s absolutely imperative that the media community and legislators think about how generative AI may negatively impact a community that’s already struggling to overcome a war with fake news and algorithmic bias.
The notion of rapidly generating content that mimics human communication but which may not be factually based should trigger alarm bells for anyone concerned with journalistic integrity,” Tina Mulqueen, founder, and CEO of Kindred PR, a full-service communications company that specializes in data ethics strategies said. 
At the same time, it's instructive to note that AI and generative tools have been deployed in a limited capacity by some publications like the Associated Press for posts around earnings reports and select sports stories with better results than CNET's.
Is there a place for AI tools in writing? Certainly, there is a supplementary position available. However, until AI truly becomes intelligent, until it can understand nuance and context, even miracle-like tools like ChatGPT can only exist as supplements. 
The lead image for this article was generated by HackerNoon's AI Image Generator via the prompt "a robot sitting in front of an easel painting a picture".

Written by brianwallace | Founder @ NowSourcing | Contributor at Hackernoon | Advisor: Google Small Biz, SXSW
Published by HackerNoon on 2023/03/07