Eventually, AI Agents Will Put Even Authors Out of Work

Computers play better chess than humans. They can be instructed on how to do detailed manufacturing, beyond the abilities of humans. They shrewdly buy and sell stock. They can read medical literature and aid in the treatment of disease. It won’t be long before even the last bastion of the human mind, creative writing, will be replaced by AI agents.

One of my personal observations that’s probably not unique to me is the following. “If your job can be done by a robot, it will be.” That’s because, of course, computers are far less expensive to operate and maintain than human beings.

That replacement process gets slowed down in many areas because of human nature. That is, while we wouldn’t mind being served a burger by a descendent of Pepper, a robot doctor performing an intimate physical examination is quite another matter. Human factors in society will naturally, selectively decide which jobs an Artificial Intelligence (AI) agent, perhaps in robot or android form, can assume in the workplace.

One of the jobs that has always been considered one of the last to fall to AI agents is the art of creative writing. That’s because the human mind has unique abilities when it comes to relating to other human beings. And yet. One can imagine a sufficiently advanced AI agent on a massive supercomputer being able to pass the Turing Test and able to generate both fiction and non-fiction that’s, in fact, pleasing to read.

One precursor to all this is IBM’s Watson. Watson is an AI agent that has been put to use reading, cataloging and organizing information about oncology by reading medical journals. From IBM:

By combining attributes from the patient’s file with clinical expertise, external research, and data, Watson for Oncology identifies potential treatment plans for a patient… Watson for Oncology draws from an impressive corpus of information, including MSK curated literature and rationales, as well as over 290 medical journals, over 200 textbooks, and 12 million pages of text.

Other research efforts have led to computers that have written experimental novels. One is called “True Love.” Here’s a story from 2014. “Computers Are Writing Novels: Read A Few Samples Here.”

This week, I learned of a homegrown effort by the Washington Post to automatically produce news stories about the Rio Olympics.The Post’s Jeremy Gilbert who leads the project said, “We’re not trying to replace reporters. We’re trying to free them up.” (For now!) You can read more about the project here in my pick of the week: “The Washington Post will use robots to write stories about the Rio Olympics.”

This emerging initiative is a natural thing to do. The amount of text, news and data being created by humans (and computers) is growing fast. Only computers can deal with all that information and, likely, properly analyze the total body of work.

A good example of human failure is when a particular author has an axe to grind about Apple. A myopic understanding of all that Apple does can easily lead to an article that looks to be born of expertise but is really quite self-serving. What if a descendent of Watson were to analyze all the facts about Apple over the years? What kind of analysis would it produce? How would i be received?

In any case, the simple task of writing news looks to soon be the realm of computers. The race will soon be on to see who can write better analysis, tips and editorials: humans or AI agents.

Next page: The Tech News Debris for the Week of August 1st. More on the ‘Macintosh Problem.’

Page 2 – The Tech News Debris for the Week of August 1st

More on the ‘Macintosh Problem.’

A few weeks ago I delved into the tardiness of Apple’s new Macs. “A Deeper Look at Apple’s ‘Macintosh Problem’.” Apple’s neglect of the Macintosh product line isn’t just something I noticed. It’s all over the Mac web. Here’s another very good assessment of the problem at The Verge. “First Click: Apple should stop selling four-year-old computers.” It doesn’t have my hopeful tone, but it’s also pleasantly thorough and blunt. It’s a good read.

Image credit: Apple

Apple is experiencing a double whammy here. When Macs were the best computer money could buy and were refreshed on a regular basis, it was easy to both visualize and construct a personal upgrade path. But with all the Macs getting long of tooth and expensive, the value proposition no longer works so well. As one reader wrote us: “In all honesty I see the Apple products as being a bit dated and drastically overpriced.”

Another potential problem is that if Apple were to drop certain Macs from the lineup, many customers who need to upgrade could find themselves forced into an alternative Mac with configurations and prices that don’t meet their needs fully. That would be frustrating.

Perhaps, as a result of Apple’s neglect, some Apple customers are seeking their own way out. See: “Is it worth it to build a Hackintosh?”

Once again, I must take the approach that when we see Apple’s new Macs this fall, all will be forgiven. (Here’s a glimmer.) However, if we’re not all surprised and delighted, expect plenty of fur to fly. I’ll be on the forefront of the fur tossing.

____________________________

Apple’s growth in revenue has slowed a bit. But its spending on R&D is still strong. In fact, the rate of R&D spending growth remains positive, according to this chart by Business Insider. That’s as it should be. Apple has to plan for future products even if one of the periodic global slowdowns happens. This article from May explains. “Apple R&D Reveals a Pivot Is Coming.” It’s exciting to think about what Apple may be up to because we know they focus their research of products that can be successfully brought to market.

This past week I wrote about a critical analysis of Apple Pay published at Pymnts.com. While there have been reader claims that the author I cited was biased against Apple, the original analysis both agreed with what I had previously read and am currently experiencing with Apple Pay in my neck of the woods.

However, there’s a broad spectrum of user experiences and approaches to analyzing Apple Pay. While this article by Joanna Stern at the WSJ seems mildly orthogonal to my original discourse, it’s also valuable additional reading. See: “Chip Card Nightmares? Help Is on the Way.”

For those who may suspect that the declining sales of the iPad means eventual doom, I highly recommend this analysis by Neil Cybart. “The iPad’s Dark Days Are Over.”

I’ve written before about the “Ultra HD Premium” certification created by the UHD Alliance. If you’re in the market for a new 4K/UHD TV, and you’re not quite sure what that sticker on the box means, I recommend this non-technical, readable summary. “Ultra HD Premium explained: everything you need to know.”

Finally, how are you feeling about iOS apps these days? Apple’s payments to developers suggest that the market is healthy. “Tim Cook Tweets: Record July for App Store; Developers Have Earned $50 Billion.” And yet, it’s remarkably hard to break into the market place for new developers. With two million apps, the road to true creativity and excellence is a tough one. And there’s absolutely no room for poor to mediocre apps. Michael Gartenberg expounds on this. “It’s the end of the app as we know it — and I feel fine.” So does Walt Mossberg: “I just deleted half my iPhone apps — you should too.” He writes: “We’ve reached peak app.”

My own feeling is that AI agents will eventually replace most apps. Perhaps all those out of work developers will turn to blogging.

Oh, wait.

____________________________

Teaser image via Shutterstock.

Particle Debris is a generally a mix of John Martellaro’s observations and opinions about a standout event or article of the week (preamble on page one) followed on page two by a discussion of articles that didn’t make the TMO headlines, the technical news debris. The column is published most every Friday except for holidays.

Share this post