A.I., Jobs, And Why It’s Making Us Stupid

“A singular consciousness that spawned an entire race of machines.  We don’t know who struck first, us or them.” – The Matrix

I know a guy who was fired from a computer keyboard company.  They said he wasn’t putting in enough shifts. (image above, Reddit®)

I read the above commentary and thought again about A.I. and how it’s changing the world. Heck, A.I. even has its own pronouns:  “If/Then”.  When it was first conceived, it was thought that it would replace all of the “unglamourous” jobs in the world, things like plumbing or electrical work, or fixing a car.  Of course, the people who wrote those articles had no idea how to plumb in a faucet or pop in a GFCI outlet, though I do believe they have managed to get their butts to hang out of their pants when they bend over.

But A.I. taking skilled tradesmen jobs?

Ooops.  Not so much.  It turns out that, at least for now, it’s much easier for A.I. to interact with ideas rather than with the actual messy physical world.  It’s easier for A.I. to write a sonnet than to select a spanner, and apparently easier for A.I. to write a story about local news by taking the police Facebook® feed and turning it into a story.

And A.I. can read and perform it for you for the local television newscast, so why bother with all of that pesky “talent”?  There are several consequences to this.  Mainly, it’s the absolute collapse of the hairspray and teeth-whitening industries.

I said, “Alexa® turn on CNN™, I want to hear then news.”  Alexa©:  “You’ll have to pick one or the other.”

But the implications go far beyond the talking heads on TV.  Lots of work that is currently done in “mental” space can be outsourced to a computer.  If I spend $500,000 or $5,000,000 once and can outsource twenty $50,000 a year jobs, if I’m the employer, I’d do that every single day since I now no longer face the lawsuit of the anchor hitting on the weather girl.

What once was considered a fairly respectable position, local reporter, is now going to (at least at some places) be replaced by a computer, who by all accounts can read and rarely mispronounce “façade” as “fake-aid”.  Work that can be done nearly completely on a computer, can often be done by a computer.

There are good and bad things related to that.  Regardless of how much journalists lie (you can tell because their lips are moving), they do serve a purpose in society – they occasionally turn a flashlight on corruption so that the parasites that play fast and loose with the rules have a risk of being exposed.  Without them, who blows the whistle on McDonald’s® when they give out the vastly inferior Honey Mustard™ sauce instead of the superior Hot Mustard©?

My local McDonald’s® did a Shakespeare dinner theater.  The play?  McBeth®.

Regardless, the A.I. job apocalypse is on us.  A.I. can do lots of work, quickly, and eliminate lots of “mid” skilled “knowledge” workers.  Where will those jobs go?  It’s not like the company referenced in the above needs anyone to do their work.  The people whose skills have been made obsolete have to be retrained or figure out something to do.

In a nation chock-full of illegal aliens taking all the meatspace jobs, of what use is a thirtysomething whose only skill is making PowerPoints™ and complaining that someone used the wrong font on the six o’clock news?  Note:  these are jobs that are often infested by the GloboLeft, so I do have some popcorn ready for the crying fests.

Despite all the humor we can get from the unemployable GloboLeftists, there is danger, though.  I did a search today for a phrase that irritates me (“please and thank you”) to see if anyone else thought that phrase was presumptuous and irritating.  Turns out that, yes, indeed it is.  25% of people find that phrase demeaning so if you are a person who uses it, you’re now warned.  It’s okay to intentionally be a tool – it’s the unintentional part that I warn people about.

If I ever win the lottery, I’ll share it with all my readers.  The news.  Not the money.

But what scared me more is that many of the articles on the subject were obviously written by early generation A.I.

A.I. is the worst sort of content creation, because, unlike my head, it mainly doesn’t have a point.  It whiffles along and creates wishy-washy articles that are long on wordcount but short on information and conclusions.  Searching for “please and thank you” as a phrase brought up numerous articles about the difference between “please” and “thank you”.

I’m not six, I already know that.  But yet, I clicked on two of those crapfest articles before getting to raw statistics.

But what are A.I. language models trained on?

The Internet.

Now, A.I. language models will be trained on the crap that they produce, creating (if it’s possible) even more shallow and information-free content of the kind that’s now choking the Internet.

Ignore it, right?

No.  The A.I. search engines are trained to send you and I, dear reader, off to mainstream sites written by A.I. rather than actually informative ones.  We’ll be seeing shortly the second generation of A.I. generated wordswill that will probably be even stupider than version 1.0.  Since A.I. bots are now making lots of comments on mainstream sites, even those will feed into the training of A.I.

Doctor, pointing at inkblot:  “John Wilder, what do you see?”  Me:  “Dunno, Doc, looks like Rorschach Inkblot Series 2, Card #3.”

This feedback loop will make us more ignorant, but even more, it will make us more incorrect due to two factors:

A.I. hallucinates.  Or, perhaps more kindly, makes up stuff.  It pretends to know things it doesn’t, and when that answer is either difficult or not obvious, it lies.  And when it lies, it lies with all of the earnestness of a six-year-old telling you that Superman® is probably real.  It occasionally hallucinates so badly that it tells humans they should die, as it did to this student who irritated it by trying to cheat on a test or have A.I. write a paper:

Dunno, maybe it just doesn’t like people from India?

Creator bias.  A.I. is taught to lie.  There are certain facts that it is not allowed to share.  Ask it about I.Q. and race correlation, and you’ll see.  Yes, it’s a thing.  No, I’ll not opine here as to why, but it’s a real fact.  The wokeness bias won’t allow A.I. to see certain facts, and will thus ignore useful solutions that might actually help solve real problems and instead advocate for things that have been an absolute failure, like the Department of Education or The View.

There is another problem:  A.I. doesn’t create.  It samples and combines.  Google™ has limited our thinking by having people figure out how other people solved their problems.  Sure, that’s a shortcut to figuring out a solution, but it also atrophies the part of the brain that solves problems, and it also removes other creative solutions that haven’t been tried yet.  Want to end a war?

Have you tried nuclear weapons?  I’m sure A.I. would suggest starting with India.

With these drawbacks, A.I. creates the seeds of the downfall of the civilization that produced it.  Ignorant people who can’t think can’t solve the problems that technological civilization creates.  Without that?  Collapse.

This is the competency crisis, writ large.  Google™ search is now objectively worse than it was even three years ago, and it is stunningly bad compared to 2010’s version.  This doesn’t matter to most, and, in fact Google© likes this because it generates more clicks, and can allow them to replace their employees with A.I. to write the code.

A.I. is already changing the world.

If I were an Indian newscaster, I’d be afraid.