Okay, so lemme get this straight from the start. The tech industry is banking our collective future on “artificial intelligence”. Every new product comes infused with it and if you – like me – are trying to read up on it, you come across phrases that are so vague and buzzwordy that it is practically begging to be mocked by anyone with more than six functioning neurons, and billions upon billions of dollars are being thrown around like confetti, all because some nerd in a hoodie typed a few fancy prompts into ChatGPT last November 2022.

Venture capitalists, them greedy little farts, they chucked an obscene $131.5 billion at AI companies in 2024 alone, and that is 52 percent more cash tossed into the abyss than the year before. Hell, in the last three months of that year alone, more than half the world’s venture capital money ended up in the noses of AI bros promising the moon and delivering polished dog shit.

Now, lemme mansplain this real slow for you . . . “AI agents” are the flavor-of-the-month tech toys that perform complex tasks for their lazy-ass human masters. Big Tech executives have been pouring tanker-trucks of bullshit about their developments all over the media, and they’re swearing high and low that these Oompa Loompas will “replace knowledge work” and deliver some kind of mystical “fundamental shift in business”.

Seriously, I haven’t heard such overblown horseshit since the invention of the Segway.

But hold on tight to your overpriced ergonomic chair, cause I tell ya, you gonna be laughing your ass off.

Let me enlighten you.

Researchers at Carnegie Mellon, somewhere in May of this year, they published a painfully honest study (”benchmarking LLM Agents on real world tasks”) which showed that AI Agents based on Google’s Gemini 2.5 Pro, supposedly one of the top dog of AI agents, screwed up 70 percent of the tasks it was given like browsing the web, writing code, running programs and communicating with other workers, and even when counting the half-assed tasks, you know, simple stuff like emailing, googling crap, or simply copy-pasting shitty code, the best Gemini could manage was a pathetic 61.7 percent failure rate.

Here are the first sentences from the abstract:

“ We interact with computers on an everyday basis, be it in everyday life or work, and many aspects of work can be done entirely with access to a computer and the Internet. At the same time, thanks to improvements in large language models, there has also been a rapid development in AI agents that interact with and affect change in their surrounding environments. But how performant are AI agents at accelerating or even autonomously performing work-related tasks? ”

(You can find their entire research on https://the-agent-company.com).

Now if Gemini was the best, the rest were of course straight beyond horse shit.

GPT-4o from OpenAI – a staggering 91.4 percent of its tasks went down the toilet. Meta’s Llama-3.1-405b had 92.6 percent fail rate, and Amazon’s Nova-Pro-v1 ,get this, it botched 98.3 percent of the time.

Ninety-friggin’-eight point three percent!

For fcuk’s sake, I’ve seen people who were totally pissed on schnaps with better completion rates.

Now let’s continue, because this here comedy show is about to get way, waaaay better because here comes Gartner. . .

---

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 !

---

AI slap-ons, strap-ons are taking over the market

Yeah, you know Gartner, our beloved corporate tarot-card reader, and they predicted that 40 percent of these AI agent projects will implode by 2027 due to pesky little things like, um, runaway costs, flimsy business value, and hilariously reckless security screw-ups.

Jeeez, talking about a kick in the ass. . .

Then there’s this guy with the unfortunate name of Anushree Verma, and he’s a senior analyst there, and he even threw more shade in when he said, “Most agentic AI projects are early stage experiments or proofs of concept driven mostly by hype and often horribly misapplied”.

Now, that , my beloved smart ass friends, is corporate speak for “you got conned, you suckers”.

There is even an epidemic they call “agent washing”.

Hahahaha, doesn’t that sounds suspiciously like a money-laundering scheme? You know, the one where companies slap “AI agent” labels onto regular old apps to ride this new wave of hype?

Oh, thaaaat kind of washing.

But isn’t everyone doing that nowadays, like, slapping AI onto their crappy-ol’-ass-of-a-product and call it innovation. Now take our ‘friends’ at Apple for instance. The are facing a class action lawsuit for calling a their autocorrect thingy “Intelligence” on their iPhone 16. And in Philly, someone tried to peddle a fake “AI financial analyst”, he got caught, and had to eat a $225,000 fine for his trouble.

Gartner’s conclusion of analyzing this tsunami of AI ‘Innovations’ out there is that out of thousands of alleged “AI agents,” a pathetic 130 are legit.

The concept of hype is hardly new, I mean, who could forget the blockchain/NFT/Web3 mess, that beautiful pyramid scheme, that turned out to be nothing more than a tech-infused Ponzi party where crypto snorters defrauded grandmas out of their retirement savings. But still, Web3 startups peaked at a mere $8 billion per quarter. But these days, the AI laughs at those numbers, because one single AI outfit recently bagged $10 billion in one go.

Tjakkaaaa!

People, we have officially gone from stupid to ludicrous.

Mainstream media, Wall Street, hell, even STUPOD himself (Da POTUS) – they have all swallowed the AI story without question, after they created a global circle-jerk that crypto scammers only could fantasize about during their coke-fueled nights in Miami. But unlike Web3, if the AI tanks, we are all screwed economically speaking.

There’s one economist at NYU named Caleb Maresca, and he clearly is a real killjoy at parties, because he warns the mere anticipation of transformative AI could gut wages and jack up interest rates to hellish levels.

---

What hasn’t been said about AI?

It is supposed to be a disruptor.

A total game changer.

The next great leap forward.

And we’re all promised a future sitting around doing jack shit while the AI Loompa slaves away.

But, be real for a sec, are we all really that gullible?

My favorite economist-slash-buzzkill, Maresca, says hold your farkin’ horses – just the faintest whiff of transformative AI (the stuff that kills jobs) is enough to send our economy straight into the toilet, nuking wages, and sending interest rates through the roof. “My findings reveal expectations of transformative AI can substantially affect current economic conditions”, is what he wrote, “even before any technological breakthrough occurs”.

Now that’s economist-lingo for “we’re screwed, and it is only gonna go downhill”.

And he continued to drone on-and-on, like a member of parliament, saying “Expectations of transformative AI substantially affect current economic conditions”, making the painfully obvious observation that bosses salivate at the mere whisper of slashing labor costs.

These friggin’ CEOs are practically drooling over the chance to automate your sorry ass out of existence.

Sam CTRL-Alt-Delete man, who will go down in history as the man who single handedly farked up the 21st century somewhere at 20%, he said that AI would need a “new social contract”

OMG. The moment a douche like him decides to use a term like “social contract”, you know us plebs are gonna get fucked.

Now, another douchebag, our dearly beloved Sebastian Siemjiemschniematkovski of Klarna – y‘all know him as the guy who single-handedly slayed 22% of his staff in favor of OpenAIs ChatGPT, and recently doubled back on his promise to cut even more staff, because – as it turns out – people want to deal with ‘real people’ when they’re in trouble and not a soulless Natural Language Processor.

And Musk, sorry, y’all know him by now as Noel Skum of course, he is actually the one that started this whole lay-off fest in the first place, when he took over Xwitter and slashed the workforce by 80% – 6.000 souls – and it could still ‘operate’ (for whatever that’s worth). He succeeded to claim that jobs might soon be “optional for humans”, because billionaires love nothing more than an optional underclass.

So Maresca, our Mr. Economic Doom himself, analyzed it even further – he said “When AI steals your job, the wages don’t vanish magically into economic nirvana, they flow directly into the yacht fund of whoever owns the AI”. In other words, the rich get richer, and you, dear reader, are left holding your rapidly depreciating “human capital”.

Wanna know what future awaits?

Think Russia, but worse.

Just exchange O- with Bro-ligarchy and you’re there: Trump’s new America. The bro-ligarchy | LinkedIn

The top 1 percent swimming in money, and the rest of us will be fighting over the remaining crumbs. Half the population scrapes by with less than 10k in savings, except unlike Russia, nobody even wants your damn labor anymore.

So, what’s Maresca’s big suggestion?

“Pressure politicians to spread the wealth from AI,” he timidly advises, as if politicians have ever voluntarily parted with campaign funds, “At an individual level, prepare for a future where human capital is worth jack shit”, he adds helpfully, all the while clarifying that he is “not a licensed financial advisor”.

Gee thanks, Caleb, that is truly super reassuring.

In the end, AI’s utopia depends entirely on society’s capacity to not fuck it up, which means we’re screwed. Totally. Big time.

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