Did Bill Gates Anticipate GPT-5’s Disappointment Before Its Release?

Sure thing, here it goes:

—

So, there’s been this big brouhaha about GPT-5. Everyone was all jazzed up about it before it even hit the scene. OpenAI, you know, those AI wizards, were talking it up like it was a whole team of genius PhDs wrapped in one. I mean, they said it could do everything—coding, writing, even dabbling in medicine. Wild, right?

Anyway, Sam Altman, the guy calling the shots at ChatGPT’s house, was hinting at something “smarter than the smartest person you know” fitting right in your pocket. I guess he was talking GPT-5, but soon after, people were throwing shade. Users weren’t too happy. They felt kind of blindsided, especially with the older models getting the boot out of nowhere. Change isn’t everyone’s cup of tea, I suppose.

One person was really upset, going on about how ChatGPT isn’t the same beast anymore—lots of bugs and hiccups. Altman tried to smooth things over, saying they’re upping the limits for Plus users and even letting them stick with the older version if they want. More choices, maybe more peace? Who knows.

Now, get this, people were grumbling about how GPT-5 seemed not-so-smart because of some broken autoswitcher. Like, the thing that was supposed to be smart wasn’t. Awkward. But OpenAI says they’ve fixed it. Meanwhile, Bill Gates was like the Nostradamus of AI, seeing it all coming. He called it a couple of years back that things wouldn’t be as rosy as imagined.

Sam Altman once confidently bragged about GPT-5 being a vast leap over GPT-4, even admitting GPT-4 wasn’t all that. But Gates was wary, voicing doubts in a chat with a German newspaper. He thought going from GPT-2 to GPT-4 was astonishing but just wasn’t buying the same leap with GPT-5.

People aren’t shy about their disappointments; lots of voices saying GPT-5 isn’t much of a step up. Gates felt GPT-4 was kind of the ceiling—nothing super above that, at least for now. But he does think AI could still soar to new heights with more funding and tech. Guess money talks?

Training these AI beasts isn’t cheap, you see. It used to cost ten cents per query; now it’s down to maybe three. But the behind-the-scenes tech? Still pricey as ever. Crazy.

Oh, and OpenAI hitting a wall in AI growth? That’s not really news. Rumor mills were buzzing about big players like OpenAI, Google, and Anthropic hitting snags. High-quality training stuff is scarce, and costs are high. No wonder they had their feet dragging on new releases.

Sam Altman was quick to snuff out those rumors, insisting they haven’t hit the AI knowledge wall yet. Eric Schmidt from Google has a similar tune—no slowing down, at least for now. It’ll stop someday. Just not today.

Can GPT-5 live up to the chatter? I’m curious if they’ll iron out the kinks and make folks happy. What’s your take on this whole AI plateau thing? Hit up the comments, let me know.