Increasingly, AI will be used to re-package the same news for many, distinct audiences, according to Matthew Kershaw.
He’s a former head of digital at MTV, who offers a deep dive into early adopters who are making AI work in journalism.
“What all these new (AI) applications show is that the future of news won’t just be the stories themselves — but also how they are presented,” Kershaw observes.
Case in point, the BBC: “We published the results of the 2019 general election at a constituency level — a different story for every constituency and in two languages, English and Welsh,” says David Casell.
He’s executive product manager at the BBC’s News Labs.
“In the future, we will also be able to personalize by style,” Casell adds. “Put it into ‘Radio 4 language’ or ‘Five Live’ language — long versions, short versions, depending on the user.”
Kershaw looks at a broad range of AI innovators in journalism with this piece.
In other AI-generated writing news:
*Norwegian Publisher Brings AI to Real Estate News: Norway news outlet Bergens Tidende has begun running AI-generated articles on local real estate sales.
Each automated write-up features the street address of the sale, type of property, price, buyer, seller, geographic coordinates and similar stats.
“The robot also has access to historic data,” observes Cecilia Campbell, chief marketing officer, United Robots.
That enables the system to compare the sales price with the most recent previous sale, calculate the difference and report the result, Campbell adds.
United Robots makes the AI-generated writing tool behind the instant stories.
“By combining specific information about individual sales with general trends in the local region, Bergens Tidende provides readers with a better overview of the market — which means we put them in a better position to make educated choices,” says Jan Stian Vold, a Bergens Tidende spokesperson.
That puts “them in a better position to make educated choices,” he adds.
*AI-Generated Writing as Pitchmaster: Travel goliath Virgin Holidays says it uses Phrasee — an AI-generated writing tool – to goose sales of its travel experiences.
“Phrasee uses AI to automatically write email subject lines,” says Mike Kaput, a senior consultant at PR 20/20.
“These subject lines sound human, but perform better than subject lines humans write,” Kaput adds.
The result? “Suddenly, I had 10 subject lines that were brand-compliant in one minute,” says Saul Lopes, loyalty lead, Virgin Holidays.
Phrasee is able to learn to stay on brand and mimick a company’s voice by analyzing countless examples of previous ad slogans, ad copy and similar text produced by human writers.
*80% of Marketers Gung-Ho on AI Personalization: Most marketers are completely convinced that AI personalization is going to be welcomed by consumers, according to a survey by Econsultancy.
That technique is often used to personalize sales emails and Web site visits with the granular data companies store on each customer.
“The report findings also indicate that customer perception and demand for AI-driven marketing could be a factor in more organizations adopting the technology in the near future,” observes Nikki Gilliland.
Gilliland is a writer at Econsultancy.
*A Look at the BBC’s AI-Generated News: BBC Labs has been using AI-generated writing in election reporting, public health coverage and more.
Working with journalists in Birmingham, for example, the news outlet created a story ‘template’ that could drill down into a health database and produce news stories differentiated by location.
Once armed with the template, it was simply a matter of creating a Web dashboard for the system to make it simple to use.
The dashboard enables an editor to simultaneously trigger the auto-generation of numerous stories from the health database which are reviewed – and then published.
*Smarter AI Chatbots on the Way?: Researchers at Zhejiang University are developing smart AI chatbots using deep reinforcement learning techniques.
The bots – essentially AI-generated writing that responds to human queries – are reportedly better at understanding grammar and meaning than conventional chatbots.
The researcher’s “techniques give AI a good grasp of linguistic structure,” observes Cathryn Virginia, a writer for MIT Technology Review.
*AI Chatbots: Key Features to Look For: AI chatbot maker Kevit has put together its list of what to look for in a killer chatbot.
The company insists that AI chatbots can run circles around human competitors, offer far better efficiency, experience no fatigue — and handle scores of conversation threads at any given time.
“That is the unmatched magic of automation,” Kevit’s blogwriter declares.
*Author: Don’t Write-Off AI-Generated Fiction: While text generators like GPT-3 have a hard time writing long-form fiction, don’t write-off automated novels, according to author Andrew Mayne.
Down the line, Mayne sees AI-text generators eventually mastering the art.
“I’ve seen OpenAI’s API (now in version GPT-3) pick up on patterns and create dialogue where characters have secrets and act out on things planted earlier on,” Mayne observes.
“I’ve also been able to get it to do longer narratives than the sample size by using a few tricks.
“This experience has shown me that as much as I want to believe my ability to write novels is a magical gift that will never be touched by technology, that’s naive.”
Even so, “I don’t think this has to be a creative apocalypse,” Mayne says.
Instead, “an Andrew Mayne novel a decade from now might mean a book where I chose the topic and then supervised an AI that created the book,” he adds.
*Big Picture AI: Where Do We Go From Here?: Four of the brightest minds in media take stock of how AI and related tech are forever changing media in this one-hour video.
At least one – Charlie Beckett, director of the Media Policy Project at the London School of Economics and Political Science – believes the current burgeoning new mix of AI and Big Tech is beyond evolutionary.
“Talk to anybody in pharmaceuticals, security, retail, government – all the arts even – and they can understand that this sort of – the growing power of the platforms, the way that we’re in this data-driven algorithmically powered world — is having a real impact,” Beckett says.
“This is about a fundamental moment when we think not just how we do this stuff — but why we’re doing it and what’s the purpose of it,” Beckett says.
Produced by Bloomberg and NYC Media Lab, the video is part of a Machines + Media series of videos rolling out during the next few months.
Also featured on the panel are Vivian Schiller, executive director, The Aspen Institute; Noelle Silver, former vice president of technology at NPR and Chris Collins, senior executive editor at Bloomberg.
*AI Robot Cast In $70 Million Sci-Fi Film: Add actors to the list of creatives looking over their shoulders at ever-more pervasive AI.
Specifically, an AI robot dreamed-up in a lab by two Japanese scientists has been cast in a big-budget, sci-fi film, according to the Hollywood Reporter.
Dubbed ‘Erica,’ the robot was taught to act by combining the principles of method acting with AI, according to her co-creators, Hiroshi Ishiguro and Kohei Ogawa.
“In other methods of acting, actors involve their own life experiences in the role,” Ogawa says.
“But Erica has no life experiences,” Ogawa says. “She was created from scratch to play the role.
“We had to simulate her motions and emotions through one-on-one sessions — such as controlling the speed of her movements, talking through her feelings and coaching character development and body language.”
*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.