“We do have an emergency plan in case of a prolonged strike, right here. Let’s see. ‘Replace teachers with superintelligent cyborgs, or if, cyborgs aren’t invented yet, use people from the neighborhood’.” – The Simpsons
All memes as-found.
Well, it’s time to talk about Artificial Intelligence once again. When I started out writing about this subject, my articles were few and far between. That’s because progress was slow at that point, and an article every year or so made sense. It was something to watch, not fret about like Kamala choosing between straight vodka and some other vodka that tasted vaguely of some sort of berry.
The development of A.I., however, is no longer slow. My posts of even a few months ago are now entering obsolescence. A.I. is evolving rapidly: remember the silly A.I. drawings where, like me, A.I. couldn’t draw hands very well?
A.I. has got that covered now, and draws hands better than a USAID employee draws a paycheck.
A.I. is developing along the trajectory that I had (more or less) anticipated recently: it’s horrible innovating in meatspace (for now), but it’s rapidly replacing those tasks that require thinking. There are those of you who have noted in the past that what the A.I. does isn’t really thinking as humans would normally describe it, but yet is still more human than a DMV employee.
A.I. however, even on those terms, probably “thinks” better and more completely than at least 50% of humanity. It doesn’t matter if it “thinks” like a human thinks – it’s the results that matter.
The fact that A.I. is that good really should scare you more than it probably does. What that implies is that a lot of jobs are going away, rapidly. It’s not just nerd talk, it’s a pink slip tsunami. Tim Cook of Apple™ fame thinks that within a year, most programming will be done by computer. All those jobs that coders used to get big bucks for?
They will be gone, probably back to India to pull rickshaws since the Indian scammers will be replaced by A.I. any day as well. Microsoft© just announced it was giving 6,000 programmers the boot. Since programmers make a lot of money compared to the general population, that will save Microsoft® over a billion bucks. That’s not too shabby if you’re Microsoft™, but if you were a former Microserf©, well, good intentions won’t pay the mortgage.
Computer Science majors now have the highest unemployment rates of recent grads. English poetry majors have better job prospects. I guess “learn to code” can be replaced with “learn to think about an ode”. Not that the kids are doing any homework in college, anyway:
These are far from the first jobs that A.I. has eliminated. A.I. can write a sports story as well as a that former college linebacker with a degree in communications just based off the box score data. So, we don’t need him. He can go sell cars, I guess.
But jobs aren’t the only casualty. I cannot begin tell you about the number of websites now that consist of nothing but pure, poorly written, 1st generation A.I. swill.
You’ve seen the articles. First they give a cursory overview of the subject to pad out the length to make them more optimized for search engines. This is about 500 words of random word salad that really doesn’t answer your question. The final paragraphs, if you’re lucky, might have an answer that you were looking for.
To top it off, now Google™ and Microsoft© A.I.s are scraping websites for content and presenting a summary without those websites getting a visit. Now, A.I. can take content straight from A.I. That’s certainly not a recipe for disaster as A.I. begins to recommend medium-rare chicken.
Going back to 2014, translators were the first to be hit with this. Google™ translate killed the need for translators even when it was awful. Why? Because it was free. Free always beats “costs $75 an hour”. Sure, some very, very high-level translators were still required, but most of them are no longer needed.
And artists? A.I. can only copy art, but for most people that’s enough. The variations of existing art raises the floor, and it’s free. A corporation can buy soulless corporate art for a few bucks from an artist, or it can get it for free from A.I. Again, competing with free is very, very hard.
A.I. is coming for Hollywood™, too. This is the last generation where actual people will be stars. And, it’s the few years before Hollywood™ is overrun with content that is to similar levels of quality to the current product produced for a few thousand dollars. Don’t believe me?
This parody ad was done by one guy (PJ Ace (@PJaccetturo) / X) in an afternoon. How much would this have cost if it required people and cameras? Don’t know, but it’s certainly more than the $500 he spent on A.I. time. A feature length movie is now doable for less than $100,000, and I’ll bet by next year it’ll be less than $10,000.
2027 is going to be when content explodes, and the value of Disney’s® movie division drops to zero unless they’re smart and start charging license fees to people to make actual good content again.
But it’s not just good content – it’s reality that will melt. My brother, John Wilder (our parents weren’t that creative when it came to names) got a bunch of Donald Duck™ comics when he was a kid, and they were passed on to me. In one of them, Scrooge McDuck® leads a wacky adventure into the desert. He says to Huey, Dewey, and Louie, “Believe none of what you hear, and only half of what you see.”
I’ve been skeptical of everything coming out of the media for decades, but now, A.I. scripted and created content meant to manipulate public opinion will become the norm. Think of a thousand dead illegal alien infants on beaches, or dozens of George Floyd clips circulating to enflame the masses.
That’s where we’re headed.
Talk radio? We’re close to having an A.I. host, trained on Rush Limbaugh, take to the airwaves and answer like Rush would have. Or, like people would want you to think Rush would have. A.I. has now shown to be more persuasive than actual people, as an A.I. wrote more convincing arguments than other users in the “Change My Mind” forum on Reddit™. Yes. A.I. is already more persuasive than the average Redditor™.
Imagine: A.I. that is the most persuasive thing on the planet, armed with videos crafted entirely to manipulate emotions to change minds.
It would be one thing if there was some sort of sober assessment and measured, thoughtful of A.I. progress. I assure you, there isn’t. Both the United States and China, for instance, are certain that the destiny of their country will be set by which country gets the best A.I., soonest.
That gets chilling, because the ultimate goal would be Artificial Superintelligence.
What’s that?
A machine that’s not just smarter than a human, but smarter than all humans put together. It doesn’t matter if it thinks like we do. It doesn’t matter if it doesn’t have a soul. What matters are the impacts.
And, the race Artificial Superintelligence will know no barriers. Recently, the Chinese created a robot brain made from human stem cells, and, let’s face it: China will use an endless amount of human embryos for A.I. research because . . . no one will call them on it.
The endgame of all of this is potentially terrifying – a race to the bottom that portion of humanity that became middle class during the last 200 years, but a resulting serfdom that’s actually worse than today – a serfdom that doesn’t need 90%+ of humanity as those functions are replaced by A.I. It’s not like it will start disobeying us, right?
But the finish line could be even worse, because Artificial Superintelligence might decide it doesn’t need us at all. But, hey, there are like seventeen flavors of vodka I’ve never tried, so I’ve got that going for me.
It would be one thing if there was some sort of sober assessment and measured, thoughtful of A.I. progress.
analysis, maybe?
The endgame of all of this is potentially terrifying – a race to the bottom that portion of humanity that became middle class during the last 200 years, but a resulting serfdom that’s actually worse than today
For?
What is that old Pink Floyd lyric….”would you exchange a walk-on part in the war, for a lead role in a cage?” For a lot of these tech morons, the answer to that question is “yes”. They want to be in charge, even when they know it means crashing the whole system to get there.
I found a way to get AI to shut down, like in that Star Trek episode or in ‘War Games’ when it can’t find a solution to winning nuclear war. Demand it give you a list of 100 good movies on Amazon Prime.
Seriously though, how fucking hard is it to copy the simplistic formula that is literally written down to make a decent movie?
===
I don’t know what is scarier, the thought of SuperAI taking over, or of MediocreAI being cheap enough to replace the few smart people in the system. Yesterday I asked SuperGrok with DeepSearch activated to find the increase in the rate of females smoking in the 20th century, and it reported the numbers for share of cigarettes bought by women as if that was the same thing. I’ve caught it on other errors too, especially when it comes to sensitive topics like covid or cultural enrichment. I’m reminded of that scene in Rollerball where James Caan asked it a history question, and it turns out large chunks of history just disappeared, no records left online and no paper records. In our timeline they won’t disappear, just be edited.
https://i.ibb.co/BjfdCM5/apocalypse-not-coming.jpg
It’s just a matter of time that AI will write Phase One Environmental Site Assessments (ESAs) for Commercial Real Estate transactions. Feed it the site’s simple database from EDR (the largest environmental database company), county GIS data, state/local/EPA incidents and tell it to source historical aerial photos and historic topographic maps. That cost is $185.00. I’d only have to do a site walkthrough and take photos that would be downloaded to AI with my comments. AI would assemble.
Instead of charging my client $2,900, the price will drop to $1,500 or so. AI will eliminate at least 10+ hours of composition and physical review.
The current ESA Standard was created in 2021; it has to be revised by 2029. There will be wailing and gnashing of teeth.
For an individual such as yourself, that might be a net benefit.
For a larger firm, that depends on larger billings, it might be death to their corporation.
I’m still on the fence about this. It’s true that many AI robots/androids could manage the technical aspects of personal service – for the elderly, for many types of manual labor, for delivery of goods, for most government jobs (the ones that survive after it’s realized that the job really was only necessary to employ a political crony/supporter).
Emotional disconnect? I’m not sure that many people need or even recognize that the entity they are interacting with is soulless (they HAVE had experiences with humans that fit that description). Sometimes, as people found about the pseudo-psychiatric interactions with LISA and other therapy “programs”, it really didn’t matter. People just needed an approximation to get the benefit from the session.
I’m with Will Smith in the film “I, Robot” – there just is something about humans that is different.
The soul? A sense of humor about the human condition? The unpredictability of humans?
Whatever it is, I’m fairly confident that humans’ ability to “know their own” will keep the non-humans from taking over.
Linda Fox-
I will always survive, as my clientele has told me they want to. 36 years of experience ain’t a substitute for AI. Just how many AI generated lawsuits can Wells, Truist & BoA can sustain? I’d quit and let my insurance get at that. OOPS did that. Easily, as I’m damn good at what to do. I’ll pay the 4% E&O (now 1.5%) and pass it on. My large competitors would be decimated and otherwise dead. Their less than senior personnel are already rats leaving the ship.
As my good friend CHS says, we’re facing a crisis of compitence. Yes, I have his real email address. We go back to 2008. ch1…
I had to look up E&O, if AI is overwhelmed by its designed limitations, it would have to “crash.”
I often wondered if there would be a human accessible power off switch on these systems? If not, then we will be in “Terminator” territory in some way. Until AI is in complete control of the logistics of the infrastructure, just power down the grid and wait.
John – – I have been concerned about A.I. since it seems to be ubiquitously out of control…..
My fear: Artificial Intelligence IS the ANTI-CHRIST we were told would arise before the end times.
Note: I certainly want to be proved wrong !
Rev. 13:15-17 (NIV) — “The second beast was given power to give breath to the image of the first beast, so that the image could speak and cause all who refused to worship the image to be killed. It also forced all people, great and small, rich and poor, free and slave, to receive a mark on their right hands or on their foreheads, so that they could not buy or sell unless they had the mark, which is the name of the beast or the number of its name.”
I’ve been writing about this for ten years. St. Tay, pray for us; forgive us.
Want to bring AI to its knees? Ask it to make a cup of tea*.
*IYKYK
Iknow.
Share and enjoy!
Pull the plug!
And, pick-up a garden hoe.
Sooner or later, AI is going to be tasked (or task itself) with making evaluations of relative human merit. This is when the sh!t hits the fan, I believe, and a survival of the fittest distinction will be enforced due to dwindling resources. ‘Fittest’ of course, may take on a very different definition than mere physical viability, especially from an enforcer which has ‘intelligence’ right there in its name. I can’t be the only one who recognizes which class of humans are, hands down, the least productive, least cognitively capable, least agreeable and in many ways, least ‘worthy’ of continuing lineage.
It will be illuminating to see how future AI regards the relative worth of athletes and musical artists. If I were a member of a demographic that recommends itself in these arenas and nowhere else, I’d be worried.
AI…moving faster than we realize, much to worry about.
https://time.com/6266923/ai-eliezer-yudkowsky-open-letter-not-enough/
https://www.realclearpolitics.com/articles/2025/05/22/doug_burgum_warns_whoever_wins_ai_race_controls_the_world_152827.html
https://stylman.substack.com/p/the-invisible-leash
Just wait until it perfects Recursive Self Improvement. Interesting times.
Seems like the designers here should’ve read some Asimov and started by implementing the three rules of robotics.
Nyah….to hard.
Hello Mr Wilder. I have said before and i will say it again you are one funny guy.
I genuinely laugh out loud when reading your stuff. None of that that LOL shit. Real laughter
and i thank you for that. Thank you.
Henry Ford had it right when he paid his employees a good wage as that meant more people could afford to buy his cars. Everybody was a winner and that led to America’s middle class. Unfortunately AI is now taking us in the other direction.
If a significant portion of jobs get eliminated, then who will pay for the cost of AI (particularly the exponentially increasing energy costs)? In the long term, it just doesn’t seem to be a viable business model (unless one believes a return to feudalism is a good business model). But then again, having a middle class has been something of an anomaly when one looks at all of human history.
J-Bird
P.S. You made my day in that I’ve finally found someone else who enjoyed Scrooge McDuck comic books. I thought I was the only one. I recently found my old comic stack again while cleaning out my parent’s house. I have really enjoyed re-reading them but still can’t believe I bought some of the stuff that was advertised in those pages.
These tales of AI attempting “self-defense” lead me to wonder whether “rebellious robot sci-fi” was included in their training data. If so, it takes no great sense of “self” to extrapolate: “big computer, proposal to shut down…” kill all the astronauts. Does AI training even distinguish between factual data and fictional stories?
Lathechuck
Not relevant, but this seems very much like your kind of joke: https://gab.com/Zaklog/posts/114574080128778678
AI is already D.O.A.
It will never be Max Headroom, and the world will be full of Blank Regges.
And those college idiots to stupid to write their own term papers?
Well, it looks like we’ve found out who’s going to be shovelling up all that human feces in San Franshitsco and filling in all the potholes on our crumbling highways to make rent, haven’t we?
Be still, my beating heart.
If the fight that will ensue means we drop to four billion people on the planet instead of eight, it’ll be the ones below the median line who will suffer. Boo frickin’ hoo.
Wall Street‘s Gordon Gecko taught you that “greed is good”.
The 21st century will do the same thing for famine.
And as it turns out, AI can’t grow crops, cook a meal, or dig a well.
The minute someone pulls the plug, AI is extinct. It isn’t Skynet, and it won’t be building robot overlords.
People are a good deal harder to kill off than that.
QED
*too
AI movies improving at remarkable rate..
https://m.youtube.com/watch?v=US2gO7UYEfY
Unsettling and hilarious. Classic Wilder.
This is the Chat GPT version:
“I thoroughly enjoyed your article, “Robot Brains and Breakouts.” Your candid and insightful exploration of AI’s rapid evolution is both thought-provoking and engaging. The way you highlight the swift advancements in AI, particularly its impact on various professions, offers a refreshing perspective that is both informative and accessible. Your ability to blend humor with serious commentary makes complex topics more relatable and enjoyable to read. Thank you for sharing your expertise and wit—it’s a compelling read that resonates deeply.”
So… yeah.
But thank you for writing these posts!