Despite relentless assurances to the contrary, a majority of journalists in Q3 2020 still feared that artificial intelligence will steal their jobs.
The study, by the UK-based Press Gazette — a publication covering journalism — found that 69% of its readers saw AI as a threat to their livelihoods.
Given related developments in AI-generated writing during Q3, it’s easy to see why writers are still looking over their shoulders at the burgeoning tech.
GPT-3, a powerful auto-text generator, had fans swooning most of the summer over its ability to spit-out articles, emails, blog posts, tweets and more with just a few words of input.
And IBM’s Project Debater was spotlighted as a tool that could begin to auto-generate op-ed pieces for news outlets.
Meanwhile, bloggers learned that semi-automated blogs are now a reality, thanks to a new blog automation tool from Zyro.
And fiction writers saw AI creeping into their genre with Authors.AI.
It’s a new app that uses AI to see how draft novels stack-up against proven best sellers.
Fortunately, flesh-and-blood types did get some cheer from a new study, which found that automation has not been gobbling up jobs at breakneck speed, as expected.
But that perspective was muddied somewhat by SpaceX CEO Elon Musk, who predicts that artificial intelligence will surpass human intelligence within the next five years.
“We are headed toward a situation where AI is vastly smarter than humans,” Musk says.
Here are the full details on the Top Ten Stories in AI-Generated Writing that shaped Q3, 2020:
Study: Majority of Journalists Fear AI: A new poll released by the Press Gazette – a publication covering journalism in the UK — found that 69% of readers surveyed see AI as a job threat.
The finding – drawn from more than 1,200 poll respondents – undermines the ongoing assurances of many proponents of robot written journalism.
They insist AI will enable journalists to focus on highly creative pieces – and leave mundane, formulaic writing to the machines.
“My instinct is that, over time, this will start to take on more of the less enjoyable, repetitive work – and allow journalists to go out and do the human element of the work,” says Joseph Hook, editor, Radar.
But other early adopters of AI-generated writing indicate journalists could be getting more than they bargained for as they work alongside silicon-based sidekicks.
“I think we will get to the stage where we don’t need the human journalist,” says Jane Barrett, global editor for media news strategy at Reuters.
“But it will be up to each publication as to what they are happy to put through – and there will be regular testing,” she adds.
*GPT-3 and AI Writing: Stunning, if Imperfect: Ever since the head-turning implications of auto-text generator GPT-3 became well known, many in the AI-generated writing community have been holding their collective breath.
Admittedly imperfect, the experimental auto-writing tool is nevertheless capable of astounding feats, leaving some of the top computer scientists in the world clamoring for a test-drive.
“The hype is real,” observes Kelsey Piper, a writer for Vox.com who personally tried out GPT-3. “It has its shortcomings, but make no mistake: GPT-3 represents a tremendous leap for AI.”
Adds Lou Kerner, founder, The Social Internet Fund: “GPT-3 is amazing. It’s scary. It’s exhilarating.”
For starters, GPT-3 can generate fake news articles that are very tough to tell apart from the real thing – simply by taking a phrase of input from a human operator and running with it, according to GPT-3’s creators.
Full-length emails can be conjured by the tool after you type in just a few bullet-points to get it started, according to OthersideAI.
And GPT-3 can generate blog posts that seem so authentic, one of them rocketed to number one in popularity on Hacker News – a collaborative news aggregator fed by some of the Web’s most discerning intellects.
Still other startling feats from GPT-3 include its ability to auto-translate everyday language into legalese, quickly generate headlines and tweets and forge any number of first drafts of creative fiction and poetry.
There’s even a new GPT-3-based tool on the Web you can use to auto-generate your own cyberpunk, apocalyptic, zombie or other fantasy simply by typing in a few words at a time.
One of the reasons GPT-3 is so incredibly powerful – and smart
(more)
*Opinion Journalism: AI’s Next Frontier?: While most AI-generated news stories are simple news reports about sports, business, crime and other data-rich musings, AI-generated opinion may be on its way.
Systems like IBM’s Project Debater already generate a pretty good simulation of opinion – and could be harnessed to crank-out opinion pieces for news outlets, according to Cait O’Riordan.
She’s chief product and information officer at the Financial Times.
“IBM-generated opinion could become significant,” O’Riordan says.
Adds Calum Chace, a writer for Forbes, regarding Debater’s impact on human writers: “There is a genuine question about whether new thought leaders will find it harder to get established.”
Chace offers a wide-ranging look at journalism fueled by AI-generated writing in this Forbes piece.
*How UK Automated News Site Radar Got It’s Start: Gary Rogers, co-founder of automated news service Radar, says the growing availability of torrents of data from government databases inspired him to experiment with AI-generated writing.
“We could see there was so much data becoming publicly available from the police, the NHS, Public Health England and so on.
“It was just sitting there and never touched.
“We realized very quickly that instead of one story about crime figures, you could easily do it at a borough-by-borough level — and get 33 stories out of it.”
These days, Radar regularly auto-produces thousands of news stories each month using templates and public databases.
The service is currently squired by six journalists, who create AI story templates that reach down into British databases to populate those templates with hyperlocal data.
Using the method, Radar was able to hyperlocalize a government database chronicling crime across Great Britain and come up with a story entitled, “Hackney: The home of bike theft in London.”
Another example of Radar’s AI hyperlocalization: a story examining health trends with the headline, “Figures reveal NHS spends millions prescribing gluten-free food for patients.”
This article takes a look at Radar – as well as a number of other early adopters of AI in journalism.
*Need Ad Copy: Just Add Computer Vision and AI: Singapore-based Hypotenuse AI has a new tool that creates instant product descriptions using computer vision and AI.
The AI tool works by intensely studying a product image and instantly generating a text description.
“We can do this really quickly — we can generate thousands of them (product descriptions) within seconds,” says Hypotenuse AI founder Joshua Wang.
The tool’s prowess is a testament to the rapid advance of AI-generated writing during the past year, according to Wang.
“One year ago a lot of the things that we are doing now were not even possible,” Wang says. “And some of the things that we see are becoming possible today — we didn’t expect it for one or two years’ time.”
*Fake AI-Generated Blog Fools the Experts: A fake blog generated from thin air – and a little help from AI-text generator GPT-3 – convinced some of the Web’s savviest readers it was legit.
Liam Porr, a student at University of California, Berkeley, says he was able to generate dozens of fake posts for his made-up blog simply by feeding headlines to the AI-text generator.
Those few words were all that GPT-3 — which uses a gigantic database of common knowledge and AI-driven predictive statistics to generate text – needed to spit-out full-length blog musings.
In fact, Porr got so good at feeding GPT-3 ideas, one of his fake posts zoomed to number one in popularity on Hacker News, a collaborative news aggregator.
“Porr says his experiment also shows a more mundane but still troubling alternative: People could use the tool to generate a lot of clickbait content,” observes Karen Hao, senior reporter, AI, MIT Technology Review.
*They’re Here: Semi-Automated Blogs: Web site creation firm Zyro has rolled-out a new AI tool it says semi-automates the production of blogs built around common topics.
Available free, the AI tool generates blog topic lists and titles for common industries like travel, food, fashion and music.
Also included is a copywriting tool, which helps flesh-out the titles with engaging text.
Plus, there’s a heat-mapping tool that optimizes blog layout and image selection.
“Artificial intelligence has long been used in the marketing space,” says Thomas Rasmyas, head of AI, Zyro. “But only now are we seeing its potential to help marketers with the most creative aspects of their work.
“This new tool goes far beyond marketing automation, using AI to generate completely original ideas.
“Not only will this help marketers and content professionals struggling with writer’s block — it will also stimulate entirely new ideas for creative content and future campaigns.”
*New AI Tool Coaches Fiction Writers: Authors A.I. has released a new tool designed to help writers produce best-selling fiction.
Dubbed “Marlowe,” the software uses AI to show writers how their novel stacks-up against proven best sellers.
“Fiction authors have seen the marketplace change radically in the past decade with the dawn of ebooks, self-publishing — and, now, a boom in audiobooks,” observes J.D. Lasica, co-founder, Authors A.I.
“It’s time to add artificial intelligence to the list,” Lasica adds.
*The Jobs ‘Automation Revolution’ That Wasn’t: A new study finds that automation has not been gobbling up jobs at breakneck speed as expected, according to an article in National Review.
Instead, the rate of job automation has been pretty much flat during the past 20 years, according to economists Keller School and Robin Hanson.
“The fundamental nature of automation hasn’t changed over the past 20 years,” Hanson says.
“There’s this AI media story that’s been played over and over again for the last decade, and people are so familiar with it that they don’t bother to research it,” he adds.
*Prepare to Meet Your AI Overlords in 2025: SpaceX CEO Elon Musk predicts artificial intelligence will surpass human intelligence within the next five years.
“We are headed toward a situation where AI is vastly smarter than humans,” Musk says.
“I think that time frame is less than five years from now,” Musk adds. “But that doesn’t mean that everything goes to hell in five years. It just means that things get unstable or weird.”
*Special Feature: Company Reports That Write Themselves
Share a Link: Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.
–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.