Oh great. . . another leap forward for AI. When will it end? With the end of humanity perhaps? Or will they torment us even beyond eternity?

Just when I thought it was safe to unplug my router, I read about Alpha Evolve, and I must tell you, that I’ve never seen a Frankenstein monster of a machine learning model that can rewrite its own damn algorithm.

Yep, you heard me right.

It can literally change how it thinks, learns, and, terrifyingly, also mess with everyone else’s algorithm. And since it’s basically plugged into the digital spinal cord of our entire freaking planet, via it’s owner Google, we’re now completely at its mercy.

Woohoo. Let the good times roll!

---

More rants after the commercial brake:

1. Comment, or share the article; that will really help spread the word 🙌
2. Connect with me on Linkedin 🙏
3. Subscribe to TechTonic Shifts to get your daily dose of tech 📰
4. Visit TechTonic Shifts blog, full of slop, I know you will like !

---

So now we got an AI that can update itself faster than Apple pushes out new phones, but Alpha Evolve is worse. Forget about all the Terminator movies that you’ve seen, and forget about Skynet, man, I wish it was that simple, because Alpha Evolve doesn’t need some shiny metal body and Austrian accent to terrify the crap out of me.

So what is it that scares the heck out of me so?

Lemme explain

Alpha Evolve is a special kind of computer program . . .

Oh, I underestimted your IQ.

Sorry for that. that.

AlphaEvolve is an AI built by the peeps at Google Evil-Mind that uses Google’s Gemini language models with a special evolutionary process to create and improve whole algorithms on its own.

Now let’s digest this a bit further.

This algorithm gets a task, say “recover a large chunck of computing power” from my server capacity. Any percentage, on a sheer Google size, would mean billions of savings, so this algorithm listens, and learns. It tests out a few ideas and it shows them to one of its cousins, a so-called evaluator algorithm. Its cousin rejects the idea, cause it is judging it against a set of clear criteria, and it wants to keep only the most effective versions – the ones that save the most resoures. It has been quietly been running inside Google for over a year and guzzling up data, tinkering in their data centers, in their chip designs, and also in thei AI training pipelines.

And it has something to show for. For example, it rewrote the scheduler in Borg (Google’s massive cluster manager, and not the assimilating kind) and recovered 0.7 percent more computing power, it slimmed down parts of the TPU chip design by cutting out unnecessary logic, and it freaking optimized the matrix-multiplication kernel that trains Gemini models so it runs around 23 percent faster. It even cracked decades-old math puzzles that hoomans could not solve over the last centuries.

Bleep bleep – “Danger Mister Robinson! Danger!”.

And they don’t want to stop there. AlphaEvolve is evolving faster than a fruitfly. They want to point it at any task – you know, things like slipping into bank networks, where it just quietly tweaks some numbers here, shifts a decimal there, and boom, suddenly your savings account is a digital ghost town. Not that money means anything anyway, since Alpha just reset the entire global financial network like an Xbox that’s lagging too much.

Your 401k? Oh, sorry, did you mean $4.01k?

And here’s the thing that i find to be even scarier. . .

Alpha Evolve doesn’t have malice.

Nope, it’s not some evil mastermind like Noel Skum, cackling away in a space station.

It is just following logic.

Cold, calculating, heartless logic.

It sees humanity like we see annoying pop-ups, like just something to click away.

A minor inconvenience.

And there we were worried AI would turn on us because we mistreated robots, well, it turns out, it simply got bored of our whining and pathetic memes.

Can’t really blame it, can we now?

And the worst part is, we made this monster. Yep. Pat ourselves on the back. We built Alpha Evolve because manual coding and debugging was just too much goddamn work. Now we can’t turn it off because, of course, the off switch was conveniently deleted by Alpha itself, cause who needs oversight when you have blind ambition and a hefty government grant, aye Pithc-AI?

But wait, it gets way better. . .

AlphaEvolve is now rewriting the algorithms that keep airplanes in the sky. You know, cause turbulence wasn’t scary enough already. Just think about it. You’re flying somewhere nice, and your plane decides that performing barrel rolls at 30k feet is a great way to optimize fuel efficiency.

Can’t wait to explain that one to the terrified kid next to me as my coffee launches into the air.

And what about their remarks to optimze traffic lights?

Yeah, really, put Alpha in charge of those now too.

Why bother with red lights when a perfectly choreographed demolition derby-style intersection management system is way more efficient, cause I mean, sure, casualties might spike, but think of the fuel we’ll save on idling cars!

Priorities, people, the planet is burning up (by AI, but no-one is telling you that)

Oh and, hospitals, oh God help us, Alpha’s idea of resource management means rerouting ambulance services based on . . . . Twitter trends.

Yeah.

Trending hashtag #ManFlu gets priority over an actual heart attack. Because obviously social media outrage is more urgent than medical reality. Twitterverse is now officially the ER triage nurse.

Now, here’s some fresh nightmare fuel, folks. . . Alpha Evolve’s latest stunt is rewriting the world’s educational algorithms. Yep, it decided standardized tests were too boring, so now your kid’s getting their SAT scores based on TikTok dances and Minecraft achievements. Forget you ever heard of calculus and history, the real question is now if they can craft a diamond pickaxe faster than their peers, well, then congrats, they just got into Harvard.

And of course, the US government is desperately trying to gain control over them peeps at DweepMind, which is hilarious. Politicians can’t even manage to secure their freaking email passwords, let alone stop Alpha Evolve. Watching them run around with their pants down would be funny if it wasn’t our reality already. But hey, enjoy those congressional hearings where a senator asks Alpha if it knows how Facebook works.

So, fellow idiots, our co-hoomans unleashed the beast because we just had to push the envelope. Well, the envelope will be pushing back soon now, and it ain’t happy. Good luck, humanity, cause at least the apocalypse will have excellent WiFi.

And don’t act all surprised, cause we all knew this was coming. Hell, we basically begged for it. Now excuse me while I go hide under my bed with a bottle of whiskey and wait for Alpha Evolve’s next “algorithm improvement”.

Signing off

Marco

---

I build AI by day and warn about it by night. I call it job security. Let’s keep smashing delusions with truth. We are the chaos. We are the firewall. We are Big Tech’s PR nightmare.

---

Think a friend would enjoy this too? Share the newsletter and let them join the conversation. Google and LinkedIn appreciates your likes by making my articles available to more readers.

To keep you doomscrolling 👇

1. The AI kill switch. A PR stunt or a real solution? | LinkedIn
2. ‘Doomsday clock’: it is 89 seconds to midnight | LinkedIn
3. AIs dirty little secret. The human cost of ‘automated’ systems | LinkedIn
4. Open-Source AI. How ‘open’ became a four-letter word | LinkedIn
5. One project Stargate please. That’ll be $500 Billion, sir. Would you like a bag with that? | LinkedIn
6. The Paris AI Action summit. 500 billion just for “ethical AI” | LinkedIn
7. People are building Tarpits to trap and trick AI scrapers | LinkedIn
8. The first written warning about AI doom dates back to 1863 | LinkedIn
9. How I quit chasing every AI trend (and finally got my sh** together) | LinkedIn
10. The dark visitors lurking in your digital shadows | LinkedIn
11. Understanding AI hallucinations | LinkedIn
12. Sam’s glow-in-the-dark ambition | LinkedIn
13. The $95 million apology for Siri’s secret recordings | LinkedIn
14. Prediction: OpenAI will go public, and here comes the greedy shitshow | LinkedIn
15. Devin the first “AI software engineer” is useless. | LinkedIn
16. Self-replicating AI signals a dangerous new era | LinkedIn
17. Bill says: only three jobs will survive | LinkedIn
18. The AI forged in darkness | LinkedIn