Microsoft AI

Microsoft’s Big Coup: AI Apps Powered by GPT-3

In a major coup, Microsoft has snagged exclusive rights to create and license apps powered by auto-text generator GPT-3.

Many AI users have been swooning over GPT-3’s prowess since it was released this summer by OpenAI.

Microsoft’s plan is to embed the technology in its own products –as well as license the AI tool to others.

“We see this as an incredible opportunity to expand our Azure-powered AI platform in a way that democratizes AI technology, enables new products, services and experiences — and increases the positive impact of AI at scale,” observes Kevin Scott.

Scott is chief technology officer at Microsoft.

GPT-3 began turning heads early summer after users realized it could 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.

Moreover, 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 first draft works of creative fiction and poetry.

Microsoft invested $1 billion in OpenAI – the makers of GPT-3 – last year.

As of May, GPT-3 has been running on one of the world’s most powerful supercomputers , courtesy Microsoft.

For an in-depth look at GPT-3, check out: “GPT-3 and AI Writing: Stunning, if Imperfect,” by Joe Dysart.

In other AI-generated writing news:

*Microsoft / GPT-3 Union Makes Some Nervous: Microsoft’s sweet new deal with OpenAI to be the exclusive licensor of OpenAI’s GPT-3 has some people on edge.

Initially founded as a nonprofit with aims to advance AI for the ‘betterment of humanity,’ OpenAI has increasingly been drawn in Microsoft’s for-profit orbit.

For example: Microsoft invested $1 billion in Open AI last year.

And the start-up has become increasingly reticent about releasing GPT-3’s code to the general public since then.

The underlying problem, according to detractors?

Commercialization of GPT-3 could leave its AI advance in the hands of a few elites, according to Karen Hao, senior reporter, AI at MIT Technology Review.

Observes Hao: “Over the past few years, there has been growing concern over the way AI concentrates power.

“OpenAI seemed to offer an alternative solution that would rely on neither corporate nor government dollars,” Hao adds.

“But that no longer seems to be the case,” Hao observes.

*Microsoft Upgrades Business Intelligence Software With AI-Generated Writing: The latest release of Power BI – business intelligence software from Microsoft – now includes text descriptions of business data.

Observes ZDNet writer Andrew Brust: Microsoft’s ‘Smart Narratives’ “provides plain-English summarizations of the data in a report — either on a per-visualization or page-wide basis.

“The narratives automatically update when data is filtered or drilled down upon.”

The drive to make business intelligence software more accessible by translating business data into easy-to-understand bursts of short text has been a key offering of AI-generated writing firms for a number of years.

The add-in tools to existing business intelligence software have been especially popular among business users who use the AI to add text descriptions to the charts, graphs and other visualizations that are produced by their business intelligence software.

For an in-depth look at these auto-text generator add-ins, check out “Business Reports That Write Themselves,” by Joe Dysart.

*GPT-3 Spawns Another Auto-Email Writing Service: OpenAI’s powerful auto-text generator, GPT-3, is powering another auto-email writing service.

The in-development tool from OthersideAI auto-creates full-blown emails from just a few bullet-points, according to company founder Matt Shumer.

“It sounds as if you wrote it yourself,” Schumer says.

Granted, not every email is perfect.

But in many cases a few bullet-points are turned “into something that sounds fantastic,” Schumer adds.

*Opinion: GPT-3: Writing Without Thinking: While many in the AI-generated writing community are understandably wowed by the power of auto-text generator GPT-3, Guillaume Thierry is unimpressed.

Thierry is a professor of cognitive neuroscience at Bangor University.

Thierry’s point: Auto-generated writing is one thing.

Real thinking is something entirely different.

Observes Thierry: “Language can be easily generated without a living soul.

“All it takes is the digestion of a database of human-produced language by a computer program — AI-based or not.

“GPT-3 is a language smoke machine — entirely hollow of any actual human trait or psyche.

“It is just an algorithm. And there is no reason to expect that it could ever deliver any kind of reasoning.”

*Opinion: Even with GPT-3, Writers’ Jobs Seem Safe: Despite GPT-3’s rave reviews, blogger Robert J. Woodhead thinks highly creative writers need not fear losing their jobs to AI.

Observes Woodhead: “There is a place for AI writing. It may become a crutch that weak writers depend on to get them started.

“It may replace boring, silly, and unoriginal text that is now written by humans — like the ones that are always bugging me to write articles for this blog as a link-farming exercise.

“But if you have any creativity in your approach to writing, I think your job is safe for a while.”

*The Age of Widespread AI-Generated Opinion Is Near: The fear that AI-generated writing will quickly swarm the Web with machine-generated opinion is well-founded, according to Renee Diresta.

Diresta is technical research manager at the Stanford Internet Observatory.

Diresta predicts the rise of artificially created ‘generative text,’ which will deepen “suspicions and change the information environment in other ways.

“In the media, editors will find themselves exercising extra vigilance to avoid publishing synthesized op-eds.

“Major Internet companies will work to make detection of generated content as fast and effective as possible.

“Still, as the detection technology grows in sophistication, so too will tools that generate (fake) images, videos, and text.”

*Automated Writing: Inevitable as an Everyday Tool: While writing purists may bemoan the advent of auto-text generators, they’re here-to-stay, according to Manu Chatterjee.

He’s an engineer and a contributing writer at Medium.

Observes Chatterjee: “The automatic bots are here, whether we want them or not.

“This is the proverbial tip of the iceberg and we need to start having a broader discussion of how and when this technology should be deployed.

“It’s not a question if — only a question of where and how much.”

AI-Generated Literature: With a Little Help From My Human Friend: While AI has yet to independently generate an engaging novel or other work of long fiction, AI/human collaborations are beginning to yield interesting results.

Poet David “Jhave” Johnston, for example, has penned a number of poems by relying on AI to generate the first draft of his works.

Observes David Wright: “Each month, he produced a book of poetry using AI.

“Neural network code adapted from machine-learning libraries was ‘fed’ with ‘human literature’ — that is, a contemporary poetry corpus.

“With neural net augmentation, a ‘block’ of artificial intelligence text was generated.

“Jhave compares these ‘blocks’ of text to stone.

“Extending this metaphor, he describes his cursor as a ‘chisel.’ The incomprehensible blocks of AI-generated text were ‘human-edited’ or ‘carved’.

“The resultant poetry is co-written by human and machine.

#ad

*AI Text Summarizers: A Look Under-the-Hood: Satyanarayan Bhanja, a machine learning engineer, offers a step-by-step look at how AI tools are used to summarize text in this paper.

Included is a look at Huggingface Transformers and BERT.

*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.

Grammarly
#ad

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.

Never Miss An Issue
Join our newsletter to be instantly updated when the latest issue of Robot Writers AI publishes
We respect your privacy. Unsubscribe at any time -- we abhor spam as much as you do.
Subscribe
Please notify me when new comments are posted about this article.
guest

0 Comments
Inline Feedbacks
View all comments