A team of 900+ AI researchers across the globe is attempting to build a powerful auto-text writing engine, whose source code will be accessible to all AI researchers.
The rationale behind the auto-text writing engine, dubbed BigScience, is simple: The more independent, creative minds you have working on the source code for any given auto-text engine, the better that software will be.
BigScience is designed to compete with more established auto-text engines like GPT-3, which offers licenses for its use but prohibits researchers from directly accessing its source code.
Auto text-writing engines differ from auto-text writers in that they provide the underlying programming and computer power needed to host the numerous apps that offer auto-writing to the general public.
GPT-3, for example, is a supercomputer-driven auto text writing engine, which is used by numerous entrepreneurs to offer commercialized auto-text writing apps.
Those auto-text writing apps designed by entrepreneurs merely sit atop GPT-3, mostly providing a user interface that relies on GPT-3 to shoulder the real work (serve as the engine) of auto-generated writing.
Spearheaded by Hugging Face, a Brooklyn, New York-based natural language startup, BigScience runs on the Jean Zay supercomputer in France — one of the world’s most powerful.
Observes Douwe Kiela, head of research, Hugging Face: “Obviously, competing directly with the behemoths is not really feasible.
“But as underdogs, we can leverage some of the things that make Hugging Face unique: The dynamism of a startup allows things to move fast.
“And the focus on open source allows us to work closely together with a strong community of like-minded researchers from academia and elsewhere.”
Another group of open source AI enthusiasts has already built a competitor to GPT-3: Eleuther.
Somewhat less powerful than GPT-3, Eleuther is powered by a network of computers.
In other AI-generated writing news:
*In-Depth Guide: Smartwriter: In yet another spin on the growing specialization of auto-text writers, Smartwriter is designed to hyper-personalize your emails — as well as your posts to LinkedIn, Instagram and similar digital properties.
Observes Patryk Miszczak, a writer who put Smartwriter through its paces: “For me, it’s one of the best cold email outreach software that I’ve ever used.
“However, it’s not the best solution if you want to create long-form content.”
The AI tool’s personalization works by analyzing data discovered about the company you’re emailing it finds on Google reviews, LinkedIn, Instagram, news sites and other digital properties — and then incorporating that data into your message.
From this auto-mined data, Smartwriter is designed to auto-write:
~Personalized email icebreakers
~Personalized LinkedIn messages
~Personalized Google Review Icebreakers
~Personalized Instagram Comments
In addition, Smartwriter also offers standard tools for blog writing.
Bottom line: This is a great guide for prospective users looking for a step-by-step walkthrough on how Smartwriter auto-builds these highly personalized communications.
*Manifest Destiny: Human Writers Will Fall to Machines: AI-generated writing is destined to surpass human writing in terms of productivity and persuasiveness, according to Dennis Francis.
Francis helps run an auto-writing service online, Content Marketing Magic.
At special risk to the growing proliferation of AI-generated writing are copywriters who contribute content to digital properties, according to Francis.
Observes Francis: “Once the algorithm can analyze and understand each recent visitor (to a digital property), it can shape the content specifically for that viewer.
“This will extend to other media, such as videos and audio.”
*Coming Soon: Highly Specific Auto-Writers: OpenAI, makers of one of the world’s most powerful auto-writers — GPT-3 — is now allowing computer scientists to customize the software’s innards.
First released to a cacophony of accolades in Summer 2020, GPT-3 has gone on to serve as the underlying engine for dozens of auto-writing software programs that have popped-up n the market since.
This latest move by OpenAI enables researchers and entrepreneurs to ‘train’ GPT-3 for a specific task by feeding the software genre-specific data.
For example, a researcher developing an app to write in the style of Hemingway will now be able to customize GPT-3 to better write in that style.
(For an in-depth look at GPT-3, check out: “GPT-3 and AI Writing: Stunning, if Imperfect,” by Joe Dysart.)
*Auto Medical Notes System Snares $55 Million in New Funding: The adoption of AI-generated writing throughout healthcare scored another big win recently with major new funding for Suki.
The firm makes an AI-powered digital assistant that auto-captures spoken-word notes from doctors and transcribes them as clinically accurate medical notes.
Doctors are also using the Suki digital assistant to issue a wide range of voice commands to help complete everyday office tasks.
*New Auto-Reporting Tool on Employee Communications Scores $15 Million in Funding: New software that closely monitors and reports on all employee communications in the financial industry has brought in major new funding.
Dubbed ‘Shield,’ the tool monitors each employee’s communications via voice, chat, email, instant messaging, social media and document creation, analyzes what’s going on — and then sends on a written report to appropriate managers for review.
The tool is designed for financial firms looking for safeguards against employee market manipulation — including insider trading, spoofing and front-running, according Shiran Weitzman, CEO, Shield.
*Auto Interview Software Firm Scores $6 Million in New Funding: Metaview has a new infusion of cash for its new software, which auto-transcribes interviews for company recruiters.
The tool is designed to offload the interview process to computerization — as well as ask more questions than typically posed in a face-to-face interview.
Metaview says its software saves companies 28% of the time they usually spend on hiring.
*Survey: AI and Journalism: Research group A14media is seeking media pros to take part in its new survey on AI.
The study focuses on the media’s take on the tech as it relates to AI’s benefits, challenges, risks, ethical considerations and related impacts.
A14media defines ‘media’ widely, seeking perspectives for the survey from pros working in news industry, film, TV, gaming and related genres.
*Just Deserts: Robot Poet Performs: Anyone scolded by a high school English teacher for penning poetry that seemed to robotic will most likely love Ai-Da.
‘She’s’ an AI powered robot that can ‘sort-of’ create and perform her own poetry — and is also similarly talented at painting, drawing and sculpting.
So far, much of Ai-Da’s ‘wit’ is tough-to-follow.
But for some, that’s perfectly fine.
Observes Alyse Stanley, a writer for Gizmodo Australia: “The results are pretty nonsensical.
“But to be fair, so is a lot of poetry.
“Maybe it’s just too high-brow for me.
“Or it could be that I’m too distracted by her soulless eyes to feel anything but creeped-out.”
*AI Big Picture: AI Most Discussed Tech of 2021: Although the AI-gee-whiz-drumbeat seems to have been thundering through countless headlines for years now, turns out AI was still the most discussed tech of 2021.
A study from Walkme.com finds that more than 175,000 articles were published on AI last year.
Moreover, Walkme.com also found that only slightly less popular in the public consciousness was the concept of machine learning — a branch of artificial intelligence.
Says Rafael Sweary, president, Walkme.com: “With the pervasiveness of AI, it’s no wonder it’s one of the most discussed technologies today.”
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.