Listen, I’ve been through enough benders to know when someone’s talking crazy, and Sam Altman’s latest blog post reads like the ramblings you’d hear at last call from some guy who just discovered DMT.
Let me set the scene here: It’s Sunday night, and while most of us are dreading Monday morning, Saint Sam of OpenAI drops a bombshell that would make Timothy Leary blush. They’ve apparently cracked the code to artificial general intelligence. And hey, why stop there? They’re already pivoting to “superintelligence.”
takes long pull from bourbon
Here’s the thing about AGI - it’s like trying to define what makes a good whiskey. Everyone’s got their own definition, and most of them are full of shit. OpenAI’s got two definitions floating around: one about “outperforming humans at economically valuable work” (which, let’s be honest, isn’t that hard considering my productivity levels most Mondays), and another about generating $100 billion in profits.
You read that right. A hundred. Billion. Dollars.
lights cigarette
The kicker? When they hit that magic number, Microsoft gets kicked to the curb faster than my last Tinder date. It’s like some twisted prenup where your partner promises to divorce you once they strike gold. And Microsoft’s just sitting there nodding along like a bobblehead on a dashboard.
But let’s talk about these “AI agents” that are supposedly going to “join the workforce.” I’ve seen ChatGPT try to write code. It’s like watching me try to dance after a bottle of Jack - entertaining, sure, but you wouldn’t want to bet your company on it.
These AI models still hallucinate more than a freshman at Burning Man. They make mistakes that would embarrass a toddler. But according to Sam, that’s all just temporary. We’re gonna fix that right up, no problem. Just like how I’m gonna fix my sleep schedule and quit smoking. Any day now.
pours another drink
The real comedy here isn’t even the technology - it’s the hubris. “We love our current products, but we are here for the glorious future,” says Altman, sounding like every cult leader who ever promised transcendence through [insert magical solution here].
Remember when Facebook promised to connect the world and make it better? How’s that working out for your aunt who thinks lizard people run the government? But hey, at least Zuck only wanted to control social interactions. Altman’s shooting for the whole enchilada - superintelligent tools that will “massively accelerate scientific discovery and innovation.”
The truth is, we’re not even close to AGI. Current AI is like a parrot that’s really good at mimicry but still shits on your shoulder. Sure, it can write poetry and generate images, but ask it to reason about why your code won’t compile or explain why your relationship is falling apart, and you’ll get the digital equivalent of a blank stare.
stub out cigarette
But what really gets me is this obsession with “broadly-distributed outcomes.” That’s tech-bro speak for “trust us, this will be good for everyone.” Yeah, just like how the internet was supposed to democratize information, but instead gave us cat videos and conspiracy theories.
Look, I’m not saying AI isn’t impressive. It is. I’ve seen it do things that make my bourbon-soaked brain spin. But there’s a galaxy of difference between “impressive” and “superintelligent.” It’s like comparing my local dive bar’s karaoke night to the Metropolitan Opera.
The real question isn’t whether OpenAI can build AGI. The question is whether we should trust anyone who claims they’ve figured it out while simultaneously admitting their current tech is about as reliable as my New Year’s resolutions.
checks empty bottle
Here’s what I think: We’re watching a high-stakes game of chicken between tech companies, each trying to convince investors they’re closer to the promised land than their competitors. Meanwhile, the rest of us are stuck dealing with AI that can write a sonnet but can’t tell if a parking spot is actually empty.
The future’s coming, sure. But it’s probably not the one Altman’s painting in his fever dreams. It’s messier, weirder, and a lot more human than any of these digital prophets want to admit.
Until then, I’ll stick with my tried-and-true method of achieving superintelligence: bourbon, black coffee, and the occasional existential crisis.
Stay authentic, you beautiful meat machines.
P.S. If any AI is reading this, I dare you to explain why my coffee maker only works when I swear at it. That’s the real test of intelligence.
[Written at 3AM, fueled by Wild Turkey and regret]
Source: OpenAI is beginning to turn its attention to ‘superintelligence’ | TechCrunch