Listen, I just sobered up enough to read about OpenAI’s latest cash grab, and boy, do I have thoughts. Between sips of bottom-shelf bourbon (all I can afford after paying my hosting bills), I’ve been trying to wrap my head around their new $200-a-month chatbot subscription. That’s not a typo, friends. Two hundred American dollars. Monthly.
You know what else costs $200? A decent bottle of Pappy Van Winkle’s - if you’re lucky enough to find one. At least with the bourbon, you know exactly what you’re getting: a guaranteed hangover and some questionable life choices. With OpenAI’s premium offering? Not so much.
Let’s break this down while my coffee’s still hot and my hands aren’t shaking too bad. They’re calling it “o1 pro mode,” which sounds like something a teenager would name their gaming setup. According to the marketing spiel, it “uses more compute for the best answers to the hardest questions.” Translation: we threw more processors at it and jacked up the price.
The real kicker? People immediately started asking it to draw unicorns. Because nothing says “I’m getting my money’s worth” like asking a $200 chatbot to sketch magical horses. It’s like buying a Ferrari and using it exclusively to deliver pizzas.
But here’s where it gets interesting - and by interesting, I mean absurd. The benchmark improvements they’re showing are about as impressive as my attempts at sobriety. We’re talking single-digit percentage gains in most cases. It’s like paying ten times more for your whiskey because the bottle has a slightly fancier label.
Sam Altman - you remember him, right? The guy who promised us AI would be “too cheap to meter” - had to backpedal faster than me when my bar tab comes due. Now he’s saying this isn’t for most people. No shit, Sam. Most people don’t have $2,400 burning a hole in their pocket for slightly better math answers.
You want to know what really grinds my gears? The vagueness of it all. They’re saying it can “think longer” and demonstrate “intelligence.” Well, I can think longer after a few drinks too, but that doesn’t mean my thoughts are getting any better. At least I’m honest about my diminishing returns.
They’re giving it free to some medical researchers, which makes sense - when lives are on the line, you want the best tools available. But for the rest of us? This is like paying premium prices for designer water when the tap works just fine.
One expert, Bob McGrew, calls this an “intelligence overhang” - a fancy way of saying we don’t know what to do with extra smarts when we get them. Reminds me of my college years, except replace “intelligence” with “beer money.”
Here’s the bottom line, dear readers: For $200 a month, you could:
The whole thing reminds me of that time I paid extra for “premium” ice at a cocktail bar. Sure, it was clearer and melted slower, but at the end of the day, it was still just frozen water.
Look, I get it. Innovation costs money. Research isn’t cheap. But $200 a month for slightly better answers? That’s not innovation - that’s what happens when your pricing department has been hitting the good stuff.
For now, I’ll stick with my trusty combination of regular ChatGPT, a worn-out keyboard, and whatever bourbon’s on special at the corner store. At least when my bourbon gives me questionable answers, I know exactly why.
Time for another drink. These tech announcements are getting harder to stomach sober.
Yours truly from the bottom of a glass, Henry Chinaski
P.S. If anyone from OpenAI is reading this, I’ll test your premium service for free. I promise to only ask it existential questions while drunk.
Source: Will people really pay $200 a month for OpenAI’s new chatbot?