Let's Take A Closer Look

Explaining complicated subject matter simply since 1986

Cavemen sharpened the points of wooden poles and used them to kill game and fight their enemies. When pointed steel tips were added to the business end of these long poles, spears became lances. When lances were adopted by cavalrymen (and yes, they were men) on horseback, the weapon became even more dangerous. Wikipedia says at the Battle of Waterloo, French lances were heavy, 10-foot-long wooden poles with steel points on the end.

In medieval times, lancers were mounted jousters

A free lancer was just that; bound to no one, a fighter for hire, a lone-wolf medieval mercenary. The Knights of the Round Table were King Arthur’s employees; Robin Hood was a freelancer.

Most of us are aware that hundreds of thousands of permanent jobs have been scrapped

Many businesses have moved at least part of their work away from a permanent workforce to the on-demand and short-term hiring of freelancers and independent contractors. 

When work is outsourced, independent contractors are paid only for what they produce, what factories once called piecework

Using contractors-for-hire is a great business model if you’re focused purely on costs. It comes in third, ahead of bloated payrolls and behind slavery (#1) and indentured servitude (#2). These last two were old-time landowners’ favorite types of labor force because neither was paid anything. Freelancers are not paid for chatting in the break room or sitting around pretending they’re working (like more than a few of the co-workers you’ve known, I’ll wager).

Companies save thousands on benefits, too

Freelancers and independent contractors don’t get paid vacations, sick days, personal days, or overtime. There are no payroll taxes for employers to pay, no workers comp payments, no pension plan matching, no bereavement or pregnancy leaves, and no pay for sitting in classrooms enduring another one of HR’s training snoozefests.

What I like about being a freelancer

  • Using my own private office. It has a great ergonomic chair, an ideal desk layout with three large hi-resolution monitors, a custom swivel keyboard tray, and an Eastern exposure where I watch the sun come up over a small pond. 
  • Choosing my assignments. My first choices are things that interest me. I do like to write about things I know, but I also like writing about things I don’t know, because I get to do more research and learn more.
  • Working to my calendar and clock. I start before dawn and put in a full day before lunch. My busiest days are Friday, Saturday, Sunday, and Monday because there are fewer interruptions. 
  • Working as much as I like. I don’t have a 9-to-5 mindset. I like to work until I’ve accomplished something, take a day off, and start on the next thing.
  • The friendly dress code. Whatever is comfortable. In summer, that’s shorts and a t-shirt.

What I don’t like about being a freelancer

The work disappears with little or no notice, and looking for work is an unpaid part-time job in itself.

My non-freelancing job

Here at Let’s Take a Closer Look, I write only about things that interest me enough to want to see what’s behind the curtain. I find interesting nuggets everywhere, whether the topic is Inkblots, the History of Drive-Thru Windows, Popeye, Margaret Mead, the Seven Dwarfs, Goldilocks, the Three Little Pigs, the Tooth Fairy, Colonel Mustard, Aunt Jemima and Betty Crocker, the Frito Bandito, Amarillo Slim, Amos ’n’ Andy, the Okay Corral or the Müller-Lyer Illusion.

When individuals and organizations ask for a portfolio, I send them a link to the Home page. There they can scroll until an article catches their eye.

Freelancers look for jobs through digital websites and apps 

Job sites and apps are intermediaries hoping it’s their ad you responded to so they can earn their fee from the job poster. LinkedIn, Upwork, Freelancer, Fiverr, Flexjobs, and many others post freelance writing jobs. They’re all aggregators, so many jobs are posted on more than one site.

Job brokers see themselves as marketplaces but the setup is more like the one at the Home Depot, where transient workers line up outside each morning hoping to get a full day’s work.

I look for interesting assignments

Innate curiosity is one of the reasons I became a researcher. I like learning of all kinds, figuring out things, and seeing how they fit together. Researchers don’t carry guns and badges but they do have more in common with detectives than most people think: researchers gather evidence, interview witnesses, uncover clues, figure out what happened, and build strong cases. 

I’ve worked briefly for several companies, enjoying some and parting ways with others when I discovered things were not what they appeared to be

Then I got hired as a freelance product reviewer. Thirty years as a researcher taught me how to collect evidence, summarize, draw comparisons, and so on, efficiently, effectively, and thoroughly. All of my four hundred or so reviews are available online. None of them are fakes and I never used A.I. to cut corners.

A.I. vs. G.A.I.

Artificial Intelligence is in the news everywhere. What is now called conventional A.I. is computer programs using algorithms to perform specific tasks. Every time you Google something, conventional A.I. takes over. YouTube, Netflix, Pandora, and a zillion others use it to tee up a new video, make suggestions based on your past choices, or direct you where someone wants you to go next.

Generative Artificial Intelligence

This is a horse of a very different color. G.A.I. trains a machine to generate new content, not to point you to something already existing, the way plain old A.I. does it. You’re probably aware of the machine writing tool ChatGPT, but you may not know that it started as a chatbot. Its other ancestors include spell checkers, auto-fill, and auto-suggest. 

ChatGPT is the most publicized language processing tool but there are many alternatives

Harry Guinness, writing on zapier.com, says the reason that dozens of G.A.I. tools have so many similar features is that they’re all based on the same models. This is the same way all spreadsheet software uses rows, columns, and formulas to produce the same outcomes, with only a few superficial differences in appearance.

No matter the brand name, all these tools scrape the same information from the InterWeb

They do so without the permission of the authors whose work they use to train their tools – and without payment. Some well-known authors have said this is the exact definition of plagiarism because these tools don’t provide accurate source information. They paraphrase the words but not the ideas. Some say it’s out-and-out copyright theft. Call it what you will, but the tools scrape the work of human writers, take their ideas, paraphrase their words, and hang them on a prefabricated outline, much as you hang ornaments on a Christmas tree.

Generative Artificial Intelligence (G.A.I.) builds layers on top of layers

This is the same way 3-D printers work. G.A.I. is also like computer coding, where no one starts over, they just pin one more tail on the pre-existing donkey.

More than a few people say G.A.I. is mass plagiarizing

G.A.I., like all algorithms, seeks the most common answers. When G.A.I. takes over, the result is no longer a hearty, flavorful stew or chowder, with fresh, local-grown ingredients, but more the like the goopy, bland strained peas you feed babies in high chairs.

Most of the things written by G.A.I. will look like the rest

There won’t be many new slants, new angles, or challenging assumptions because assumptions are what G.A.I. is built on. What we read will be the fast-food version of writing.

What does it matter?

Not much. Reading for entertainment and enlightenment is becoming a lost pastime. If mine is the kind of reading you enjoy, please keep it going by sending my articles to friends and colleagues who enjoy reading things written for inquisitive and informed adults.

Brits call it Rubbish In, Rubbish Out

We haven’t heard the term GIGO used much lately, but maybe we should. Garbage In, Garbage Out describes the concept that the quality of any output is limited by the input. Feed the machine raw garbage and you’ll get processed garbage out the other end. In statistics, we used to say you cannot make an insights silk purse out of a data sow’s ear. The same is true with algorithms, the formulas that tell machines what to do. Algorithms learn only from what they’re fed and do only what they’re told.

Generative Artificial Intelligence’s machine-written words, phrases, sentences, and paragraphs are meant to resemble human-created content

Because it analyzes patterns and relationships to understand the underlying rules, G.A.I. is restricted by the humans that are telling the algorithms what to do. Tech Republic posted an excellent article by Owen Hughes, who says G.A.I. obeys its algorithms without question and uses only the data it was trained on, so shouldn’t we be interested in who’s conducting the training? If it’s a chore assigned to middle managers who are outsourcing it to people who work on the cheap, it means folks who have lower standards than you are willingly cutting corners and skipping important steps.

Most mechanized programs can write stuff simple enough for 8th-graders to read

Automated writing programs are also only extensions of the auto-correct, auto-suggest, and auto-fill functions on your computerized devices. Tools such as Grammarly, Hemingway, and Slick Write inspect sentences and offer suggestions for improvement. The suggestions do not come from Strunk & White’s classic handbook on English usage, The Elements of Style, though. They come from previous readers who clicked on Accept or Reject. When the users of the tool get it wrong, their incorrect usage is what’s teaching the writing program: GIGO.

Dumbing down

I often see writers using it’s for its, and its for it’s. More than once, while following a company’s template, I typed its and the program responded by underlining my word in red and suggesting it’s instead. A few pages later, I typed it’s and — that’s right — the program recommended its. These programs help every now and then, especially when I typed the same word word twice in a row row. 


Companies buy language processing tools so they can save money by getting rid of human writers altogether. But the trend in writing jobs these days is hiring freelance writers to read what the machine wrote and clean up the mistakes. 

It’s a job that calls for very little writing, so companies can hire less-skilled freelance “writers” to catch the glaring mistakes. For years, writers were told to write at the level of an 11th-grader, but lately, the new directive is to write for a 8th-graders, a clear concession to the decline in reading skills among Americans.


There are some excellent G.A.I. programs friends of mine use, but only a few. Most do run-of-the-mill work, such as writing product reviews. G.A.I. programs can be helpful for some things, but never forget that the algorithms are being written by people who may well have lesser language, vocabulary, and grammar skills than you do. Most of the machine-written reviews we see in English read awkwardly because they’re programmed by people for whom English is a second language. 

Those who have been reading some of these articles written by machines agree they’re generic, boilerplate, with frequently incorrect spelling, usage, and tense. If you’re an 8th-grader, you probably won’t notice.

Enter your email address to subscribe to this blog and receive notifications of new posts by email.