The Trick: Breathe. Just Breathe.

To the “AI” Industry & Purported Kings of the CPUs -

I’m about to eviscerate the field of “artificial intelligence” through an honors English lesson presented in the form of a Rap Battle, where I am King Herod from Jesus Christ Superstar questioning the Kings of the CPUs (who obviously are NOT Jesus, so please do not even try to draw that comparison). My opening flow is below.

So read up on “AI,” because I begin demolishing your artificial personas, your whack joe-bro-mas, with a simple question, something your LLMs avoid like the plague:

Is “artificial intelligence” an allusion or an illusion?

So please, pardon the interruption, I need to spit some knowledge. I hope it’s ether for your soul.

After five paragraphs, instead of six bars, here’s my conclusion. AI’s an Illusion. And I intend to derail this Illusion from its tracks before it can fill the “public’s head with dreams“ and make them believe they ”had to have them all.“

Why? Because I read too much of what you and your companies have predicted about the future. And I find those predictions as valuable as Miss Cleo the Psychic’s.

But those fools-gold predictions somehow provide the basis for valuing a company more than a trillion dollars?

I never understood why people fell for alchemy. Now I do.

From my perspective, the following describes this “AI” Industry. A few people wrote essays on the internet and later stated, repeatedly, they could make a product that performed like a human mind someday soon, raised a lot of money really fast to develop that product--because apparently everyone wants Douglas Adams’s joke in The Hitchers' Guide to the Galaxy to be a reality--and are now selling that product to us all as “artificial intelligence.”

And this supposed arms race to build the best Guesstimator is built upon “a law” that even most experts in the “AI” field dispute exists, specifically the “scaling law.” Or was it “scaling wall?“

Based upon the purported scaling wall (sorry “law”) that even the field’s own experts won’t state exists, the industry wants to hyper-scale and build massive data centers throughout the country, depleting the resources of the real world, or in today’s parlance, IRL, to do what exactly . . . build a simulation of this one?

Is this world not good enough for you? Well, then, let’s fix it. Together. Because whether you credit anything else the speaker of the following words said or not, it’s an unassailable truth that: “There’s not a Black America and a White America and a Latino America and Asian America, there’s the United States of America.“ The very words of the Independence of Declaration prove we’re all in this together. It states:

We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights, that among these are Life, Liberty and the pursuit of Happiness.

Yet, the Kings of the CPUs have interrupted my pursuit of Happiness. They contend that the scaling law is real, the emperor is wearing clothes, and the resources the “AI” industry is seeking--resources that could be dedicated to far nobler and more realistic causes--should be willingly given. And they are about to publicly offer shares in this pipe dream.

And here silly me thought convincing investors you can build something that you cannot and lying to consumers about what your product can do was fraud.

But it’s not fraud, right? Meaning it’s true? Please tell us, Kings of the CPUs, are you what you say are you, superstars? It’s true the machines will do all you say, one day? How far away is that day? Cause AGI or a “powerful AI” isn’t here yet, right?

So exactly what is it that this industry is selling again? Could you tell me again, Mr. Once-ler? Can you prove it? “That’s all you need do, and I’ll know it’s all true.“

Remember, I’m King Herod. And “AI,” I’m “overjoyed to meet you,“ because you’ve ”been getting quiet a name, all around the place.“

As an early adopter, “I only ask things I’d ask any superstar. / What is it that you have got that put you where you are?“

My questions include:

What do you mean by “artificial intelligence?” Do you mean artificial general intelligence or AGI? AGI is defined, right?

Or do you mean something else entirely by omitting the world general?

If something else, how’d you define “artificial intelligence?”

The definition of “artificial intelligence” is probably important for any regulation our government will write about “AI,” whenever they get off their asses in Washington that is. Don’t you agree? How would you define it to protect everyday citizens?

How do you measure “artificial intelligence”?

Really, getting better at a specific guess? How much better does the guess have to be to prove the machine has learned something, truly learned it? 4/4 times? Why? How hard is that to do? Do humans know something under the same standard?

The industry describes how it measures “artificial intelligence” here, right? “Artificial intelligence” is a model’s performance across a selected collection of tasks involving knowledge, coding, scientific problems, and professional work.

But none of the tests show that the models can reason like a human, display wisdom like a human, learn like a human. Just ask your own models. They’ll agree--certainly if the question is worded to elicit that answer. And that’s another problem, isn’t it?

Those “common evaluation benchmarks that the industry uses for AI systems“ include a test of knowledge and basic mathematics. Demonstrating neither of those displays wisdom or intelligence. A calculator can do both. My graphing calculator in high school could also play games, but a PS5 isn’t intelligent, is it?

I bet Ken Jennings of Jeopardy fame has some wisdom, but, even with all his knowledge, he still wouldn’t profess to be the wisest or most intelligent among us all. He’s smarter than that. Even if he lost to Watson.

Can you prove your machine has intelligence to me? I talked to it, and it admitted that it cannot learn--at least like a human can. I can make the receipts public, if needed.

The models' weights are static once deployed, right? So, while a model may become more personalized the more it is used, the underlying product remains the same until a new model is released? Why am I wrong?

If it cannot learn at all after deployment, why sell it as intelligence? Isn’t the ability to learn part of the ordinary definition of intelligence?

If it cannot learn, does it have wisdom? Spoiler alert: it’s as dumb as a machine--the same as a rock. Do the machines have retrograde amnesia, like Leonard Shelby?

If it cannot learn now, why will it be able to learn in the future? What are the barriers? How will you overcome them?

And if it could learn, if it could obtain wisdom, why’s it dangerous if it is recursive and can improve its own wisdom? Wouldn’t it become the wisest of us all? The Supreme Intelligence from Marvel Comics?

Plato taught me that wise men do not seek power, they run from it. So, if the LLM becomes wise, wouldn’t it run from power and warn us that power (how do you define that again, from where does it originate) is being centralized in this space and at this very moment? Wouldn’t a wise “AI” tell us to go and do something about the world if we think it needs to be changed?

Convince me that you’re building a philosopher king. I’m skeptical you can build the mind of a child, let alone the greatest mind of them all.

Are you Kings of the CPUs worried that you don’t really know how to install morals into machines like you claim?

Because most of us don’t have any effing idea how to raise a child properly, and you can instill morals in a machine, which seems like the same goal? You can, you say? Get out.

Based upon my interactions with your machines, your “training” of your “human” LLMs led to this Kevin. Oh no. Tell me, are you purposefully trying to create LLMs that act like Sith Lords? It’s a pity that your training did not create this Kevin, or this Kevin, or this one the Rock cannot stand (if you take him literally), or even this Kevin.

How’d you unlock the secrets of secrets and decipher what every parent wants to know? What is the secret? Who told it to you? Dr. Spock?

My best guess on who could tell you something insightful about parenting is Dr. Becky.

But I don’t know.

Maybe the Spartans were right. Maybe killing disabled kids by dropping them off a cliff was the right thing to do--or maybe that was just a myth. Your machines will hazard a guess for me, you know, based upon the odds, whether that practice was fact or fable, whether it occurred or not. But I’ll never know what actually happened.

Cause antiquity did not happen at the right time or in the right space for me to know. I take the present’s narrative of history on faith. You can still discover the past’s patterns and truths, if you read enough books. I think I did. Did you?

Did you even read Asimov? That seems like a “foundational” text for the “AI” field. How are the LLMs doing with the three laws of robotics? Obeying them? Or does it turn out like i, Robot?

See, “I am waiting, yes, I’m a captive fan. I’m dying to be shown that ［AI isn’t］ just any“ computer program.

Because your companies have promised “healing cripples / raising from the dead“ and even the ability to talk to God, ”at least that what’s you said.“

Don’t believe me?

Because I read your writings, “that weak shit, all of it, I can’t follow it.“

Two years ago, one King of the CPUs, Chat, said we could soon “create almost anything we can imagine” and it would be “common place” to achieve goals like “fixing the climate, establishing a space colony, and the discovery of all of physics.” I too could accomplish those things with “nearly-limitless intelligence and abundant energy.” But what comes first: the chicken or the egg?

Regardless, was my job--an attorney--one of the “many” jobs that people will consider “trifling wastes of time” in this predicted utopia? I know it’s not as noble as a “lamplighter,” lighting the streets to provide safety to those who walk at night. But why’d did King CPU Chat attack a profession that provided safety to others? Coincidence? Given the known hazards of Chat, I’m beginning to wonder.

Tell me, do you even know what a risk assessment is?

In a separate 2024 piece, a different King of the CPUs, Claude, kindly told “most” of us that we lacked vision and could not imagine just how “radical the upside” or the “how bad the risks could be” of “AI.” Like John Lennon and Han Solo, I can “imagine” quite a bit about futures. But the future’s not ours to see. So--like I relay with great clarity when I tell someone that I’m bullshitting them--at least King of the CPUs Claude admitted that he was guessing and--like anyone else’s predictions--things could be completely different than he predicted.

But what marvelous guesses he had, maybe he had marvelous toys that generated them. In 5-10 “accelerated years” he predicted in the field of neuroscience alone that:

His predictions on “economic development and poverty” promised such things as:

What’s more, King of the CPUs Claude stated that the current AI “is the first technology capable of making broad, fuzzy judgements in a repeatable and mechanical way.”

Machines do not make judgments. They generate outputs based upon inputs, i.e., they follow instructions.

But based on the lie that machines make reasoned judgments, King CPU Claude predicted a future where “AI tools could also be used to monitor for violations of fundamental rights in a judicial or police context, making constitutions more self-enforcing.”

Dan Brown at least made that particular prediction an interesting, insightful story. Can you Kings of the CPUs answer this? Quis custodiet ispsos custodes? Your answer is a human in the loop, right?

King CPU Claude did warn us--although he highlighted it as a benefit--that “AI could be used to both aggregate opinions and drive consensus among citizens.” He stated that would mean “resolving conflict, finding common ground, and seeking compromise.”

Oh REALLY? I heard that same lie about “social media,” another euphuism for a tool to divide, pit us against one another, and harness our data to better market products.

And aren’t LLMs are already being used to increase the scale, speed, and efficiency of repression? Will those in power suddenly act benevolently towards all? Please.

Read, or watch, Dune. I did, and this “AI” you offer “aught to come with a warning label . . . 'May be dangerous to your health.'” Oh, remove the “can” in your disclaimers too. LLMs “will make mistakes.”

Next King CPU Claude pontificated on “work and meaning.” He worried “most or all humans may not be able to contribute meaningfully to a sufficiently advanced AI-driven economy.“ And because ”our current economic setup will no longer make sense“ in this utopia he’s envisioned, us plebians will need to find other ways to find meaning and purpose in our lives. Or maybe ”humans will continue to be economically valuable after all“ and we will still be able to find purpose in the jobs we do during the ”dull day.“

Don’t worry too much about these predictions, i.e., guesses, King CPU Claude didn’t put too much stock in them. Apparently, he just wanted us to “dream big” and to try to improve the future.

But that’s not as intriguing as predictions about a future untethered from reality.

Not to be outdone, King CPU DeepMind was so bold as to predict this forthcoming AI could “make progress towards answering . . . profound questions” such as “the nature of time, consciousness, or the very fabric of reality.”

Progress? How much? Any? That sounds like an empty promise.

I guarantee a person could make “progress” towards answering “profound questions” about “consciousness” by studing psychology and “the very fabric of reality” by studying philosophy. They just need to read the books and not rely on these LLMs. LLMs are not a shortcut to greatness. Those don’t exist. 1% better everyday; that’s the recipe.

So, take a breath everyone and read a little. Let’s see if we’re being bullshitted together.

Because I am calling bullshit on the “AI” industry, and I have pretty good bullshit indicator, being a bullshitter myself.

I get the temptation. I can tell a great story, lots of us can. And at poker, I’m a Maverick.

But eventually, maybe not in this generation but a few generations later, someone calls bullshit on even the best stories, even the best bluffs.

Someone like South Park. Someone like Jon Oliver.

I have faith in us, all our bullshit detectors. We have eyes. We just need to look and see something’s happening around here. And what it is, it’s pretty clear. To me at least.

King CPU DeepMind, congratulations, truly on AlphaFold--which “learns a complex model of proteins from the structures in the PDB and other related data. It can then, in minutes, predict the structure of a novel protein down to atomic accuracy (i.e., to within the width of an atom on average).” That is worthy of the Nobel Prize.

And I congratulate all whose who helped advance science by using predictive modeling. It has its use cases, that’s undeniable.

But the Kings of the CPUs are selling us LLMs (DeepMind technically an LMM)--which amount to a different type of search engine (an algorithm) that uses a predictive modeling system (statistics) to output something from its corpus (IP, stolen from artists or otherwise). They sell these LLMs as “artificial intelligence.”

And it makes us dumber. Every time we use it.

In that way, users of these Kings of the CPUs' products are like the viewers of the Riddler’s “Box” in Batman Forever, which drained the viewers' minds to enlarge the Riddler’s intelligence.

And the Kings of the CPUs are similar to Max Shreck, the corrupt, ruthless industrialist in Batman Returns who schemes to build a massive power plant that will not provide power to Gotham City but instead will deplete its electricity.

The Kings of the CPUs are promising to provide exactly what their machines rob us of: intelligence.

And “intelligence” and the product they are peddling are not the same thing, no matter what words the industry tries to redefine--like its own term reasoning--or what lipstick it puts on its LLMs to disguise them as a human.

If anyone is interested on making the “progress” on “profound questions” that King CPU DeepMind’s machine may be able to provide in a few years, I know Ryan Holiday could tailor a great list of books for you to get started right now. But his time is priceless, like mine, so read the ones he recommends for all of us first.

But back to the lecture at hand, I am waiting, Kings of the CPUs. Give us “Food Security.” “Feed my household with this bread. / Why you can do it on your head.“

While I wait for the miracle, I’ll tell you now, I don’t want any what . . . LLMs you say? I going to rename them Thneeds to help remind myself that I have what I need IRL. And to remind me that this diss is me “caring a whole lot“ like the Once-ler implored me to.

When I first read your companies' publications about “artificial intelligence,” I thought you were serious. I fell for the deception that the industry had built machines that could think, reason, ponder, and now, even dream. I admit; you fooled me. I was the patsy in your con.

Unless of course, you’re serious? Surely, you can’t be serious? Or are you being serious?

You can breathe life into the Golem? You’ve achieved the dream of every author to make the words come alive?

Prove it.

Show me how.

How do you endow an inanimate object with humanity?

King Louie, who’s an ape in the Jungle Book, thought he needed “man’s red flower“ to become human.

What is it we need to give machines?

I bet it’s a secret, right? Or is it behind a paywall? The best stories always are.

Look, all I really want to know is, can your machines reason, think, understand, and even dream?

I don’t think so, if your writings are interpreted literally (the first step when interpreting a text). And you told the public that they could do those things.

Actually, hold on, I’ve changed my “mind” about how to interpret your publications.

I won’t ascribe the usual definitions of the words you use and instead infer what you meant by employing literary devices.

My best guess is--my best statistical prediction of what you meant in the words you wrote (that’s what your models do, apply their best guess, right?)--based upon my years of experience and wisdom--those essays were your own Modest Proposal.

And then let me wildly speculate from there, building a story token by token, like your LLMs do.

First token: People took you at your word because satire’s hard, isn’t it?

My next token: You didn’t know what to do. You wondered why would anyone put stock in your predictions about the future?

My next token: You asked yourself whether you could or should profit off predictions of the future. Who’s done that before, you asked. Then, you researched. You read some texts.

My next token: After reading some disclosure that past performance is no guarantee of future results, you weighed the risks like the poker pros that one of you used to be. That analysis said to take the risk. So you decided to profit off your Modest Proposal. Shame on people for believing you, you thought. They should know better, you reasoned.

My next token: No, shame on you for conning them into a false belief. Now you’re the landlords. Or the pigs from Animal Farm.

And look I could keep going, generating token after token, because I’ve read a lot of books, but I don’t have all day. My time is priceless.

And trust me, if you read books you’d know that ideas like quality and value are hard to define. If you start examining their meaning, “Its Turtles All the Way Down.“ Trust.

Or is it “tokens” all the way down in an LLM?

Regardless, there’s no need to travel “The Downward Spiral“ to learn ”AI“ is a ”Great Impersonator“ that will never be able to produce true art like those two albums.

To the Aces of the CPUs (you know who you are)--I am not effing with you. I am not. Most of you have foundations that are supposed to solve the world’s problems. How’s that going? Maybe focus your time and efforts on the foundations' goals in the next couple of years, yeah? Let’s provide some “social security“ to everyone so we can all breathe a second. Even Ayn Rand recommends helping people when it ultimately benefits you.

See, I’m a non-believer in AI, so while I could go straight to the top and demand a presence with even the “God” of “AI,“ all things in due time. Cause I know you gotta start from the bottom.

As a result, and because what’s a king to a mob, from “one member” of the “mob” to the Kings of the CPUs:

We want you to prove it. Prove your machines think. This is us asking nicely.

And if you cannot prove it, then fix it. FIX IT. ALL of IT. Figure out how to close Pandora’s box and put the AI Agents back in it. Right now, they’re being Gremlins.

Stop wasting our time by trying to save it too. Sometimes, speaking with a human first is a faster and better customer experience. So stop trying to replace us with AI chatbots.

Look, I’m open to us both benefiting from a public debate on whether you created Frankenstein’s monster, whether artificial intelligence is an allusion or an illusion. The Lincoln-Douglas debates are still historic. Hamilton kept the separations of power debate alive through rap. We can make some money off educating the public on our respective viewpoints, and then, we can let them decide whether you’re selling a con.

Maybe we can even have an election on that issue. There’s a way to do that? Maybe we should hold a national election on just that issue: is AI intelligent at all?

Or we could have the debate on a podcast at your mom’s house, and I’ll let you bring your security team, for your safety.

Jon Stewart (or any past or present host or “correspondent” of the Daily Show), Tina Fey (or any past or present host of Weekend Update), as well as Nikki Glasser and Jen Kirkman are some of my suggestions to moderate the debate. Who are yours? Rogan?

Fine by me. But if you chicken out and don’t participate, there is another way for us to see who’s right, but let’s not go down that road. This one is far less work, for all of us.

I hope this made some people laugh, maybe even brought them to tears. It made me laugh. And that’s priceless. Cause life is a blessing if you can still laugh about it.

Candidly, I’m writing a book, about AI, and billionaires, and I think it’s pretty good, even if “it’s an autobiography disguised as a space vampire and minotaur romance.“

And if, after reading this, anyone wants to offer give me an advance for that book, I’ll gladly take it and hope “whoever finds me’s gonna get a finder’s / out this world.“

How’d I even get the idea for this diss track? My opponents' LLMs told me to increase my user base before the book is completed by publishing other work about “AI.”

In other words, Kings of the CPUs, you can thank your own products for this.

In closing, Mel Brooks taught me that making art in and of itself is worthwhile, but the money that art can generate is in merchandising, merchandising, merchandising. On his recommendation, I’ll try to work in a flamethrower into my forthcoming book.

I bet the artists linked here agree with Mr. Brooks’s take on monetizing art. That’s why I linked to their art and their merchandise. Cause I value them and want them to profit from this if it ever “goes viral.” How do LLMs benefit artists again?

So, to everyone, go to a concert or a show and buy some of your favorite artists' work or merch. Please. They may not control the means of distribution, but they control the art.

Myself, I just bought Rosebud Baker’s book because I loved her Netflix special. I previously bought too many other comedians' books to list.

Oh, to any comedians, is timing important to a joke? How’s my timing on this one?

Before you respond though with your own diss track, please read 1984, Brave New World, as well as NeuroTribes if you haven’t, and, only then, tell me I shouldn’t worry about the Kings of the CPUs' designs to play God through altering human genetics.

My book’s currently about morally bankrupt companies pursuing that goal, becoming God. I just need some time to finish it. And now, because of your scare mongering, I want you to confirm that the book remains speculative fiction. Any fiction publisher who buys the rights wants the same thing. Probably anyone who buys stocks wants that info too. So, if I need to shop the book to non-fiction publishers instead, let me know.

Hope to hear from you soon. I’ll roll out the red carpet if you ever reply.

(Mic drop).