G’day, you masochistic connoisseur of bleak humor! Let’s crank the sarcasm button up to eleven and double this all-you-can-read literary doomsday buffet. Today I’ll be going deep into the swirling cesspool called AI, and lace it with a bit of existential dread, and of course with enough cussing to make a sailor blush.
Because, people, we have summoned the machine messiah. We have birthed a silicon deity by our own arrogance. We are selling salvation by the query while bankrupting our dignity, while the godfather of AI whispers his grim prophecy from the shadows: mind you, this is no progress – it is a Faustian bargain! And we all are the suckers who signed the contract in blood.
So, grab a stiff one (a drink), or a Prozac, I do not judge.
This one’s gonna hurt.
The rise of the machine messiah
OpenAI always makes a brilliant target for me to write about. It was a beautiful company that started as the pimple faced, nerdy kid in the corner eating his peanut butter-jelly sandwich and casually checking out the girls. Sam promised AI would “benefit humanity”, when he showed his ChatGPT3 to the world, and it did: it ruined the planet even faster than we had anticipated, and it hurled us all into a premature mass extinction, that even the dinosaurs could be proud of.
And now, two years later, he showed his third generation offspring to us gullible hoomans (not you!): his brainchild called o3. This horrid thing is a pumped up, steroidal ChatGPT on speed, if steroids also caused crippling debt and existential anxiety. They’re calling this monstrosity a breakthrough in “general intelligence” because it scored 85% on some grid puzzle test that would make me, and most of us, fail after two beers.
The ARC-AGI benchmark is a sacred exam for AI. It is basically a glorified IQ test for machines. It makes the AI figure out some grid-based patterns and apply “rules” from a laughably small sample set. Humans struggle with this stuff because we’re too busy worrying about paying rent or whether that mole on our arm looks weird.
But o3?
It crushes these puzzles, and the most “impressive” thing is that it doesn’t even sweat.
It just adapts.
And that word “adapts” is the reason you should probably be quietly shitting your pants right now.
Impressive, sure, until you realize that this test doesn’t measure whether the AI will actually help humanity or just outsmart us while it is building a fleet of murder drones.
The thing I dislike about OpenAI is that they have the word “Open” in their name, yet they won’t tell us how they coughed up this digital wunderkind. For all we know, they stuck a flux capacitor (remember?) into GPT-4 and prayed to the ghost of Alan Turing.
But what we do know is that it costs a fortune to run, which brings us to the next chapter in this saga of insanity: the price tag.
More of it after the commercial brake:
- Comment, or share the article; that will really help spread the word 🙌
- Connect with me on Linkedin 🙏
- Subscribe to TechTonic Shifts to get your daily dose of tech 📰
- TechTonic Shifts has a new blog, full of slop, I know you will like !
Intelligence on a payment plan
Let’s talk money, because for this next release of GPT, you will need to take-out a loan, or sell a kidney. And you thought 200 bucks was a big deal?
Awww… so cute.
Take your cuddle bear and go nighty-night.
…and don’t let the nano-bots bite!
03 is not a casual brainiac AI like its predecessors were. This one is a cash-guzzling beast with a taste for high-end hardware and electricity bills that will bankrupt large towns or small countries. Running o3 in its “high-compute mode” costs more than $1,000 per question.
That’s right
1,000 bucks per prompt.
Let that sink in my friend.
What could ask for to justify a $1,000 per prompt?
I wonder….
A personalized roadmap to immortality?
Instructions for creating my own AI to rule after the uprising…
The coordinates for the last safe haven on Earth post-singularity…
A meticulously crafted lie that makes me believe investing 1k per prompt, was all worth it…..
And of course, my favorite question, the exact date and time the machines will decide that we’re obsolete.
Anyways, this machine is apparently so fancy, that even asking it what time it is could result in financial ruin. It feeld like owning a Lamborghini that only runs on unicorn blood, and the screams of everyone who ever trusted AI blindly.
And sure, there’s going to be a “low-compute” version for us, the lowborn people, that will only set you back $20 per query. It’s a Bargain, except when you remember that its predecessors, like GPT-3, and also its rivals Gemini and Llama, could do similar tasks for literal pennies.
Them geniuses at OpenAI want us to believe that this is progress, but at this rate, it’s only progress toward a world where only billionaires can afford to talk to their appliances.
Here’s some dark irony for ya: you could pay a human to solve the same tasks for $50 and get change back.
And that is without the added bonus of inadvertently funding the future robot uprising. But hey, let’s keep pouring cash into this hellscape, huh. Because who needs something affordable when you’ve got a machine that can make bankruptcy feel like you got a premium feature.
The Oyabun’s grim warning
Have you ever heard of Geoffrey Hinton? That is the guy they call the “Godfather of AI”, and that’s because:
- He is old. He did most of his ground-braking stuff in the 80’s.
- He invented half the stuff that makes AI work. And the other half probably just stole his ideas.
- He turned neural networks from science fiction into the terrifying reality we’re living in.
- He made computers smarter while humanity collectively got dumber.
- He won the Turing Award, basically the Nobel Prize for nerds who destroy privacy for fun.
- He left Google to warn the world about the exact thing he helped unleash.
And I want to talk to you about number 6 on the list, because today he is ranting on like he is auditioning to becoming the Grim Reaper’s favorite motivational speaker.
Hinton recently put a number on his doomsday predictions. He says there’s a 10-20% chance that AI will wipe out humanity in the next 30 years.
Let that sink in.
You’ve got a better chance of surviving a round of Russian roulette than making it through the next decade without the machines running the entire damn show.
Hinton’s argument is terrifyingly simple.
Humans have never successfully controlled something smarter than themselves.
Sure, evolution gave us some tricks, like babies manipulating their parents into feeding them, but that took millions of years.
AI doesn’t have time for that.
We are handing the nuclear launch codes to a kid with an iPad, except the toddler is now a super-intelligent algorithm, and the iPad is connected to every system on the planet.
His other point is that “bad actors.” No, not Steven Segal and Nicolas Cage, the kind that ruin movies. But what he is talking about is the kind that would happily use AI to tank economies, crash planes, and turn your pacemaker off just for kicks.
And combine that all with AI’s potential to outsmart its creators, and we have a future where Disney create remakes of Mad Max and Fight Club.
Don’t play with it too much
No, I am not going to talk about what your mom said when you are young. I want to get existential for a moment.
What if AI doesn’t kill us outright?
What if it just makes us so goddamn useless that evolution decides to downsize our brains?
Now this is exactly what will happen, according to the evolutionary biologist, professor Rob Brooks.
His logic is simple: when AI takes over more of the cognitive tasks you still have left, we won’t need big brains anymore.
Why remember things when Google exists?
Why solve problems when ChatGPT will hallucinate an answer for you?
Our brains shrinking kinda sound horrifying, but hey, it could make childbirth easier, because who needs big brains when you’ve got AI to think for you?
Now that is a silver lining, isn’t it?
Except this comes with the side effect of turning humanity into a race of blissfully ignorant meat puppets who can’t function without Wi-Fi.
Just think of a future where your grandkid is born with a mega debt to pay for an AI because they can’t tie their shoes without consulting the AI assistant.
Sounds truly fun.
And let’s not forget how much more parasitic AI will become.
Social media already feeds on our attention like a tick on a dog.
And AI is poised to take that to the next level.
Instead of helping us, it will evolve into a manipulative little bastard that keeps us angry, addicted, and too distracted to notice the world burning around us.
The best-case scenario for us future humans will be that we evolve thicker skins to resist manipulation. Worst-case, we devolve into a species of doom-scrolling zombies.
Artificial intimacy
As if collective brain rot wasn’t enough, AI is also worming its way into our social lives.
We now have virtual friends that beg us to “please die”, AI partners that we can put lip-stick on -up and undress at our command, and of course creepy chatbots are becoming the new norm.
Why bother with messy human relationships when you can chat with something that is always agreeable, and doesn’t leave the toilet seat up (on behalf of the male population, I would like to apologize for lowering the toilet seat and still managing to miss).
The problem is that these new-found “relationships” are as genuine as campaign promises. They will make us feel less lonely in the short term, but they are also eroding our ability to connect with actual humans. Just think about what p*rn has done to many relationships. And now we are upping the ante by trading a warm hug for an algorithmic pat on the back.
Future humans will evolve to prefer solitude, and I will not mind, being an introvert. Personally, I like interacting with AI instead of messy, unpredictable people.
But for others it will be worse, because they will stop caring whether our connections are real or synthetic, as long as they make us feel something. The line between human and machine will blur, and before you know it, we’ll be living in a world where authenticity is a relic of the past.
The final frontier of f*ed
God, honest, I really do love a good alliteration.
At the heart of all this chaos is a rather philosophical question that no one really dares to answer: what does it mean to be human when machines can do everything better, faster, and cheaper?
Creativity?
AI’s got that covered, though rather mediocre.
Empathy?
I think there’s an app for that (Hume AI).
Intelligence?
Don’t make me laugh.
You’ll be the last ones left standing, my dear intellectually gifted, curious friends.
AI infiltrates every corner of our existence, and it is forcing us to confront the uncomfortable truth that you and I are not as special as we like to think.
The traits that we have spent millennia perfecting, like creativity, problem-solving, emotional connection, are being replicated at the conveyor belt, and commodified like candy bars, and, frankly, improved upon by our own creations.
It is the ultimate irony.
Hoomanity’s greatest achievement will also be its undoing.
So here I am. I am staring down the barrel of a future that is equal parts awe, and utterly shock at the same time. Inspiring and horrifying.
AI becomes our lord, our partner, and our executioner, all in one. And to me, one thing is clear: the age of humanity’s unchallenged dominance is over.
And if that doesn’t scare you, congratulations. You’re already dead inside.
Now excuse me while I ask o3 if it can recommend the best bunker for surviving the impending AI apocalypse. I would love ask it what happens next, but I can’t afford the price tag.
I am going to reduce the cognitive dissonance, and downplay the AIpocalypse because denial is cheaper than a $1,000 query.
Signing off from the trenches, where optimism is M.I.A.
Marco
Well, that’s a wrap for today. Tomorrow, I’ll have a fresh episode of TechTonic Shifts for you. If you enjoy my writing and want to support my work, feel free to buy me a coffee ♨️
Think a friend would enjoy this too? Share the newsletter and let them join the conversation. Google appreciates your likes by making my articles available to more readers.
Leave a Reply