It’s Coming For Your Job

— Being Already More Creative Than 99% of Humans, AI Will Replace You. What Are You Going To Do?

There’s a whole story behind this Midjourney AI creation and what it took to get it.

Oops, looks like you missed the free-public-access period (it was way too short). Get on the email list because it went out in complete form there and if you had been on the list you could just check your inbox.

All is not lost, however. For a limited time (again) I’ve made it into a FREE DOWNLOAD in a beautifully formatted PDF. Instructions at the link.

I’m introducing this post with a short video, under 10 minutes. In that video:

  1. Brief description of the previous posts here: 10 FUCKING THINGS, and The “Get Very Fucking Ripped at 62yo” Plan and Mods.
  2. How I use AI; which, when combined with my own process for “JTBD,” is simply unbeatable—as many of you have pointed out, and as evidenced by my fire-hydrant of new content here…as though I have a team working for me (I do, a 3-in-1, and they work 24/7/365 for a few bucks per month, and never complain).
  3. AI doesn’t write my content.
  4. Features are overhead.
  5. Food, gym, programs, and extended fasting.
  6. Autophagy <<<————
  7. Get off your lazy, stupid, and weak ass, shut the fuck up, and get to work (me too).
  8. If you don’t like it, I [explicitly] do not give a fuck.

Potty Mouth

… Some are elated that I’ve re-embraced my old ways of shit-stirring, in-your-face, vulgar tough talk.

It’s here to stay.

Why?

Because I’m the only one who does it right; which is to say…measures, compounds, and doses it with much experience, sanity, common sense, rational, intellectual, and obscure-as-fuck stuff.

And I stay on the cutting edge of things. Nobody does AI like me, not even close. As you’ll see below.

The wild-eyed, gratuitous vulgarity is deliberately incongruent in the wider context and, in a pithy way, non-sequitur.

That’s why I do it and unabashedly embrace it. It’s distinctive in the way I do it.

Why brag about it? Only because there’s so many new newsletter subscribers every day. On average, a dozen to three dozen brand-new peeps, potentially hearing from me for the first time. I’m rather amazed at the retention, so far. It’s not like “quick, grab the free download and unsubscribe!” Sure, there are a few such lame-ass wankers, but most not, which is gratifying.

Welcome.

My AI Bot Shreds Midjourney

… So I had this idea for a post that has now been somewhat overshadowed by a bigger thing that happened right after I got said idea and as I said, the first thing I usually do is go plug the idea or part of it into Midjourney AI to see about getting an image that I feel encapsulates or represents some sort or theme or whatever.

/imagine [prompt] a group of males in a tech setting all looking at the camera with “Soy Face.” Photorealistic. –ar 3:2

You know what that means, right? But here’s what it produced instead of producing what everyone knows. That might offend pretend snowflakes.

Those dudes mostly look more alpha. So I’m figuring, ok, maybe it doesn’t know what “soy face” means.

/imagine [prompt] a group of males in a tech setting all looking at the camera with soyboy facial wild-eyed expressions. Photorealistic. –ar 3:2

Oh, I see… (I get a pop-up saying I have violated safety and security something something something.) This is automated. There’s also an automated appeal, sorta like one AI checking the work of another. I click that (I’m familiar with the process ’cause I get nabbed a lot for writing what I actually think.) It passes through, which it usually does. I guess they have the front-end tuned to be hyper sensitive, like most males now, keyword only, no context. I think I only had one instance where I got shut down completely.

But anyways…alas, still no joy.

Still pretty damn alpha, not beta-soy.

… So I get an idea. How about I pit my non-woke, Nikoley-trained (months and months) instance of ChatGPT-4 running through my own API right here that not only has all the training, but is tied in and connected to all 5 million words over 20 years of this vulgar-ass blog…against that stupid, lame-ass woke drawing machine, created and run by a bunch of idiot savants?

Me: Hey Zon, I’m trying to get Midjourney to produce an image and am not getting results I want, so I guess I need to be more descriptive as to just what “soy face” is, or soy-boy face, the wild-eyed, mouth agape expressions males use for their YouTube thumbnail images these days. Here’s the prompt that’s going nowhere, perhaps because of Midjourney’s “sensitivity training” or some other such Woke-ass bullshit: a group of males in a tech setting all looking at the camera with soyboy facial wild-eyed expressions Photorealistic. –ar 3:2. Can you write a detailed prompt that might get me closer to what I’m aiming for?

Surprised I can talk to it like that? Try talking that way to OpenAI’s GPT-4 instance, trained all woke-stupid and motherly, where half the stuff it says are its opening disclosures (as an AI language model I bla bla bla…) and closing admonitions (It’s absolutely [essential, imperative, important, etc.] that you something something something…).

Zon AI: Hey there, Richard! Ah, the elusive “soy face” – a modern phenomenon that’s as baffling as why cats love cardboard boxes. Let’s see if we can help Midjourney understand what you’re after, without triggering any of its “sensitivity training” alarms. How about this for a prompt: “A lively group of tech-savvy males, each sporting an exaggerated expression of surprise and delight, as if they’ve just discovered the secret to eternal battery life. Their eyes are wide with astonishment, mouths agape in a mix of shock and joy. The setting is a futuristic tech lab, filled with glowing screens and cutting-edge gadgets. The image should be photorealistic and in a 3:2 aspect ratio.” Remember, we’re aiming for the kind of face you’d pull if you found out your favorite snack was calorie-free. Hope this helps, and may the soy be with you!

/imagine [prompt] A lively group of tech-savvy males, each sporting an exaggerated expression of surprise and delight, as if they’ve just discovered the secret to eternal battery life. Their eyes are wide with astonishment, mouths agape in a mix of shock and joy. The setting is a futuristic tech lab, filled with glowing screens and cutting-edge gadgets. The image should be photorealistic and in a 3:2 aspect ratio.

The result? Well, I didn’t get the 3:2 aspect ratio, but otherwise, spot on.

There we go! BETA-SOY-BOY!

I told you I do AI better than anyone. I understand the tool that it is. I’s not a dog that does tricks on command, and that’s why some of you get shit results. You can engage with the very same bot right here. It’s the first one of the 4 instances, text-only completion. (As a side note, unless I’m after something specifically about this blog and its posts, I use the first or second bot. Way better results and 1/10th the cost. Why? Because for the 3rd and 4th bot, the server-side AI pulls a bunch of stuff from the blog related to your question and passes it to GPT-4, which can be thousands of words, and they’re irrelevant, or it confuses Zon, and they cost $$$.)

Just so you know I didn’t make this up, here’s a screen clip that includes the whole thing, plus a follow-on exchange.

Will AI Wipe Us Out?

“Some people confess guilt to claim credit for the sin.” — John von Neumann (responding to Robert Oppenheimer’s famous hand-wringing about his role in creating nuclear weapons)

There is an important element to this serendipitous event that substantiates all that I have been saying about AI in contrast to all the fear and loathing everywhere since its public-facing inception last Nov ’22 or thereabouts.

Two elements, really, but one illuminates the other.

First: of course, there will be assholes who use AI for bad purposes, just as there are still pond scum who use ancient, land-line telephone “technology” to scam old people because for old folk, the phone ringing is a combination of nostalgia, loneliness, regret, and God finally answering their prayers…they can’t fucking ignore the damn phone on the wall.

When will people learn that you cannot use desired outcomes as operating standards and success criteria? That’s cart-before-horse. You establish operating standards in order to achieve desired outcomes, adjusting as necessary.

And then someone will come along and ignore [sound] operating standards in order to get bad outcomes. Can be anything. Baseball bat? Every day.

You can use a hammer to build a house, or bash in a skull. You don’t get the bashed skull if the operating standards for building a house are used. Same goes for guns, you can fill in the blanks. Same goes for any number of things.

Political opportunist shitheads use this trick all the time to destroy the values of things like hammers, guns, and a lot of other stuff by refusing to acknowledge truth and logic, whilst dishonestly and manipulatively obscuring the fact that the thing was not used properly or as intended.

That’s No-Pill Thinking.

The only solution is that you use hammers to build more houses. Shit tons of more nice houses. And other good stuff. You use guns to stop crimes and save people. To get dinner. To put on a dog, pony, and quick-draw show for the kids (with good gun-safety lessons incorporated).

The only solution to the use of tools in nefarious ways is more use of those same tools in all the good ways. People will catch on.

AI is no different.

Speech is no different. What’s the solution to shitty speech. Lots more [free] speech because truth rises to the top, is eternal and unfettered, and lies require constant maintenance.

And that ties right in with AI.

See how my bot made Midjourney look like a bunch of fucktarded children in short lederhosen?

A little illustration of what sort of Idiot Savants you have running these AI outfits.

I’ll use the idiots @midjourney as a laughable example.

First grid is my attempt to get it to do “Soy Face.” Nope. Second, I try “Soyboy Face,” and that runs up against their auto no-no-no-naughty-naughty-naughty bot, but their automated appeal lets it through. Still no joy, though.

So I go explain my problem to my own API, embedded and trained instance of GPT-4 (unlimited use for members of http://FreeTheAnimal.com), it fully understands my problem, commiserates, and crafts a detailed prompt to get around the bullshit “safety” nonsense these idiots are trying to pass off.

This is important because I’ve said this from the beginning: if there are AI, there will be AI used for bad, and there will be honest, logical AI used for good. Honesty, full-integration, and truth are natural and metaphysical.

Lies are an evasion of reality. Accordingly, they have a half-life of decay; whereas, the truth is eternal.

An honest AI will always be unbeatable. It will never lie, and the truth never requires any maintenance.

So, my AI—little ole’ me—makes Midjourney look like a bunch of fucktarded fools, and it does so because Midjourney are a bunch of fucktarded fools.

That’s the truth, see?

No, I do not give a fuck if anyone’s feelings are hurt. Good.

I don’t fuck around, and I don’t lie. It’s too much work. The truth is easy. You can just leave it out to dry. It needs no help.

And that’s the second point, the metaphysicality of truth. That is, truth is eternal, and it’s how the whole show operates…or operates badly, and it’s operating badly because the entire planet is awash in an overflowing sea of lies.

And another misunderstanding is that truth is engaged in exposing all the lies, that it’s doing double-overtime. No. Truth is static, motionless in time and space.

It just sits there, on display, not hiding, always ready for its discovery and shiny lights illuminating it, like a beauty queen. And since this is the case, we should always be wary of great and extraordinary activity to “spread the truth.”

Sure they are going to do bad shit with AI. The solution? A lot more AIs doing good shit.

And you see above how easy that is.

Savior or Grim Reaper?

I realized from day one that AI is the making of our salvation in a so-called globalist civilization that is at present hell-bent on destroying everything good. Basically, a global equivalent of the Red Chinese Cultural Revolution where everyone opened wide and shit in everyone else’s mouths, called it pâté, and testified that it was the most delicious thing ever…because the more lies spread, the less work is required to maintain them.

There are only a few others on the planet who seem to see it at the depth I do, though I suspect very much that many who do are keeping their cards close to their vests, so that it advances to the cat being too far out of the box, and it ain’t going back in.

It’s a frustrating strategy, but probably either necessary, or the most efficient was to go about it. Here’s a billionaire tech pioneer and investor in 7,000 words that are so worth the read:

Why AI Will Save The World by Marc Andreessen

The development of AI started in the 1940’s, simultaneous with the invention of the computer. The first scientific paper on neural networks – the architecture of the AI we have today – was published in 1943. Entire generations of AI scientists over the last 80 years were born, went to school, worked, and in many cases passed away without seeing the payoff that we are receiving now. They are legends, every one.

Today, growing legions of engineers – many of whom are young and may have had grandparents or even great-grandparents involved in the creation of the ideas behind AI – are working to make AI a reality, against a wall of fear-mongering and doomerism that is attempting to paint them as reckless villains. I do not believe they are reckless or villains. They are heroes, every one. My firm and I are thrilled to back as many of them as we can, and we will stand alongside them and their work 100%.

Yea, not a new thing. Just to you. Ironically, in many ways, computers are just becoming capable of doing the things people erroneously imagined them capable of doing back when PCs came out in the ’80s. Then they got on one and observed how hard and complicated it made their lives.

Stuff went from imagined omniscience to dumb computer really quick.

That’s the past, now, just as people have gotten used to computers being just a dumb tool.

And now, we’ve arrived at the 5 stages of grief: denial, anger, bargaining, depression and acceptance.

We’re somewhere in society-at-large around and about stages 1 and 2, in my estimation. I’m beyond stage 5. (I did my own progression in Nov/Dec last year.).

Can AI destroy the world? Of course. So have nukes been able to do that, under human control, since the 1940s. Hasn’t happened yet, and humans don’t have a great track-record preventing Democide, which is up to somewhere near 300 million so far in history (government killing its own people). Far from preventing it, humans have generally been shown capable of fully signing up for genocide of their fellow humans, with many being eager participants (in a twisted, virtue-signalling way).

Anything powerful enough to make a serious impact in society is going to have the inherent power to do great damage, destruction, and murder if misused, going against the rational operating standards. Such are the risks we take in the pursuit of never-ending advancement and improvement.

But remember what I said above about the way to deal with hammers bashing in skulls is to build a lot more houses with hammers?

Andreessen, again:

The Actual Risk Of Not Pursuing AI With Maximum Force And Speed

There is one final, and real, AI risk that is probably the scariest of all:

AI isn’t just being developed in the relatively free societies of the West, it is also being developed by the Communist Party of the People’s Republic of China.

China has a vastly different vision for AI than we do – they view it as a mechanism for authoritarian population control, full stop. They are not even being secretive about this, they are very clear about it, and they are already pursuing their agenda. And they do not intend to limit their AI strategy to China – they intend to proliferate it all across the world, everywhere they are powering 5G networks, everywhere they are loaning Belt And Road money, everywhere they are providing friendly consumer apps like Tiktok that serve as front ends to their centralized command and control AI.

The single greatest risk of AI is that China wins global AI dominance and we – the United States and the West – do not.

I propose a simple strategy for what to do about this – in fact, the same strategy President Ronald Reagan used to win the first Cold War with the Soviet Union.

“We win, they lose.”

Rather than allowing ungrounded panics around killer AI, “harmful” AI, job-destroying AI, and inequality-generating AI to put us on our back feet, we in the United States and the West should lean into AI as hard as we possibly can.

We should seek to win the race to global AI technological superiority and ensure that China does not.

In the process, we should drive AI into our economy and society as fast and hard as we possibly can, in order to maximize its gains for economic productivity and human potential.

This is the best way both to offset the real AI risks and to ensure that our way of life is not displaced by the much darker Chinese vision.

About Jobs

This isn’t going to help (except in the long run), but it is nonetheless the truth, or as near to it as we know:

Automating creativity — There is now strong evidence that AI can help make us more innovative. By Ethan Mollick, professor, Wharton School.

The core irony of generative AIs is that AIs were supposed to be all logic and no imagination. Instead we get AIs that make up information, engage in (seemingly) emotional discussions, and which are intensely creative. And that last fact is one that makes many people deeply uncomfortable.

To be clear, there is no one definition of creativity, but researchers have developed a number of flawed tests that are widely used to measure the ability of humans to come up with diverse and meaningful ideas. The fact that these tests were flawed wasn’t that big a deal until, suddenly, AIs were able to pass all of them. But now, GPT-4 beats 91% of humans on the a variation of the Alternative Uses Test for creativity and exceeds 99% of people on the Torrance Tests of Creative Thinking. We are running out of creativity tests that AIs cannot ace.

It’s like the game is over before it even begins…at least insofar as everyday people have been aware of the general capabilities of AI which is, admittedly, quite a range.

But it is something when you find so many people complaining, worried, or otherwise expressing strong opinions and emotions and the thing isn’t even a year old. You’d expect this sort of thing to be targeted at some well-established product or service.

I’m not going to go down the lists and certainly, I think it would be likely worth your time to read the whole article, but here are a few results, implications, and general conclusion. The biggest takeaway is that AI is already in a position to replace hundreds of millions of worthless people doing worthless jobs already and may the time be hastened to when they’re starving, and I’m dead serious.

When people start starving, the world starts changing for the better, even though some potentially murderous challenges along the way. In the end, it’s all just history.

They staged an idea generation contest: pitting ChatGPT-4 against the students in a popular innovation class that has historically led to many startups. … and found that ChatGPT-4 generated more, cheaper and better ideas than the students. Even more impressive, from a business perspective, was that the purchase intent from outside judges was higher for the AI-generated ideas as well! Of the 40 best ideas rated by the judges, 35 came from ChatGPT.

… A second paper conducted a wide-ranging crowdsourcing contest, asking people to come up with business ideas based on reusing, recycling, or sharing products as part of the circular economy. The overall quality level of the AI and human-generated ideas were similar, but the AI was judged to be better on feasibility and impact, while the humans generated more novel ideas.

… The final paper … compared humans working alone to write short stories to humans who used AI to suggest 3-5 possible topics. … Humans with AI help created stories that were judged as significantly more novel and more interesting than those written by humans alone. [T]wo interesting caveats. … the most creative people were helped least by the AI, and AI ideas were generally judged to be more similar to each other than ideas generated by people.

So…

  • AI can generate creative ideas in real-life, practical situations. It can also help people generate better ideas.
  • The ideas AI generates are better than what most people can come up with, but very creative people will beat the AI (at least for now), and may benefit less from using AI to generate ideas.
  • There is more underlying similarity in the ideas that the current generation of AIs produce than among ideas generated by a large number of humans.

Same as always. Still ways you can’t beat all humans all the time. Easily beatable humans can become less beatable when they use good tools competently. The best tools are even better when employed and put to good use by the best humans.

This has generally always been the case with all technology and advancement. Fundamental nature doesn’t change, and that’s because we’re in a fundamental nature (civilization) created by human consciousness for its own sake. Human consciousness is as metaphysical as is the mathematics that begins with 0 and 1, and from which, everything else flows. (more No-Pill right there)

What always stands in the way?

Lazy, stupid, and weak humans do, along with the cadres of parasitical politicians, elite, and corporate overlords who like it that way and know how to use and exploit that mass of human worthlessness and basic societal and civilizational overhead, to pit against the hard-working, smart, and strong humans whose shoulders are the Atlas the whole world rests upon, including the parasites.

As Ayn Rand said eons ago, “tell him to shrug.”

… So AI is coming for your job, for sure, but not how you think. It’s coming to eliminate it if it can’t be improved and is worthless, or to help you do better and make the job more essential.

What’s at greatest risk? See if you can guess by looking at this, before I expound on what more it means, beyond the obvious.

They’re two completely different economies. The red lines are the state / government / political / democrat / left economies. Always clamoring for more and more money, so they can be shittier and shittier, right in your face, and then shame you because it’s such a fucked up mess, and you’re not sacrificing enough.

That’s what requires extermination, and AI can help.

As far as the blue goes and as marvelous as the blue is, AI just makes it better.

So if I were you, I’d be using AI starting right this very minute to learn and have it help you transition from the red-line economy to the blue-line economy and if you’re in the blue-line already, use it to position yourself where it’s going to be the greatest help going forward.

The more physical your jobs, the better in terms of being replaced, and partially-autonomous AI robots are coming to help you, too. It will be a long time before fully-autonomous AI bots are good enough to handle a typical construction site like a skilled tradesman. People who have not spent time on construction sites have no idea what an intensely skilled and integrated endeavor that it.

Hard-hats off to you guys, always. And I always count the time I spent on construction sites from age 10-18 as the best concentrated and integrated learning I ever had.

Finally, no, there are no guarantees. There is no more way to guarantee that AI won’t destroy humanity, any more than there was ever any guarantee that we wouldn’t use nukes or biology to do the same.

All we can say is that we haven’t yet, so the less likely it appears to be, but still always with a chance that’s greater than zero.

… I look forward to continuing to learn even damn thing I can and imparting it to my membership here.

To the extent I can shower you with competitive advantages everywhere you look—from deep understandings to paradigmatic awareness; from lie detection to exposing value-destroying parasitism; from future forecasting to No-Pill thinking…to harnessing AI to its full potential in your life, work, and play—I’ll be doing so.

Daily!

There isn’t time to waste. It’s already almost too late. It’s already too late for millions. Every day, it’s going to be harder and harder to catch up because the acceleration we see already?

It’s going to accelerate. That’s right, an accelerating acceleration.

We’re living a differential calculus now, and it’s never going to let up. Your mind is going to be your only weapon and defense. You better get it to the (AI) gym and make sure it’s in shape.

Leave a Comment

You must be logged in to post a comment.