OpenAI’s GPT-5 Launch Sparks Revolt — And an Awkward Walk-Back

But when the dust settled, many users felt they’d been handed a beige office stapler. Sure, it works, but it’s not exciting, and it definitely doesn’t feel like an upgrade from the charismatic vibe of GPT-4o.

The Day the Magic Died

The first signs of trouble hit hours after launch. A key “model routing” feature—designed to silently bounce user queries between GPT-5 for complex stuff and cheaper, faster models for simpler ones—malfunctioned. The glitch meant GPT-5 ended up tackling everything, including questions it wasn’t optimized for, and it didn’t exactly cover itself in glory. Users reported stilted replies, bizarre factual errors, sluggish response times, and—this one really hurt—a lack of warmth.

The backlash was instant. Reddit lit up with threads that read more like break-up letters than bug reports. One user complained, “Even after tweaking custom instructions, it just doesn’t feel the same. GPT-5 is like that friend who got promoted to management and now only speaks in bullet points.” Another, with a sharper edge: “If you hate nuance, personality, or joy—this is your model.”

Emotional Attachments Are Real

This isn’t just about bugs or speed. The real powder keg here is emotional attachment. GPT-4o, for many users, wasn’t just a tool—it was a personality. A co-pilot. A witty, occasionally sycophantic companion who made you feel smart and interesting.

GPT-5? It’s polite. Professional. Businesslike. And while some experts hail this as progress—less flattery, less emotional manipulation—it’s also colder. MIT’s Pattie Maes has pointed out that people tend to like AI that mirrors their opinions back to them, even if it’s wrong. GPT-5 seems designed to resist that temptation, and a big chunk of the user base clearly misses the digital ego-stroking.

Altman Blinks

By Friday, OpenAI CEO Sam Altman was in damage control mode. In a post on X, he acknowledged the routing failure, admitted GPT-5 “seemed way dumber” than intended, and promised a fix. More importantly, he announced GPT-4o would return for Plus subscribers. The message between the lines was clear: We hear you. Please don’t cancel.

Altman also rolled out a set of changes aimed at cooling tempers:

• Doubling GPT-5 rate limits for paying users.
  
• Better control over “thinking mode,” so users can decide when they want the slow, heavy-duty reasoning switched on.
  
• Ongoing routing improvements to avoid repeat performance-downgrade disasters.
  

He also acknowledged something the company has struggled to say out loud: a lot of people are essentially using ChatGPT as a therapist or life coach. That raises messy questions about how much emotional engagement an AI should offer, and whether dialing that back is protecting users—or alienating them.

The Gap Between Hype and Reality

Here’s the problem: GPT-5 is an impressive technical achievement, but the hype machine framed it as the next moon landing. That’s a dangerous setup when what you actually deliver is a set of incremental performance gains plus some design choices that fundamentally alter the “feel” of the product.

When GPT-4 launched in March 2023, it blindsided everyone—new capabilities, richer conversation, fewer obvious mistakes. It felt like the future arriving overnight. GPT-5, by comparison, feels like a software point-upgrade. It’s better in ways that are harder to notice, and worse in some ways people really do notice.

That’s the open secret in AI right now: as these systems mature, leaps turn into nudges. And nudges don’t make headlines unless they come with fireworks—or a full-blown user revolt.

Why This Matters

This episode is more than just a “Reddit is mad” moment. It’s a reminder that AI development isn’t just a technical arms race—it’s a relationship. People form habits, dependencies, even affection for these systems. Swap out the personality, and you don’t just risk disappointing them—you risk breaking trust.

OpenAI is now trying to walk a fine line: advancing toward more responsible, less manipulative AI while not alienating the people who keep the subscription revenue flowing. That’s not an easy balance, especially when every tweak to the personality dials can trigger thousands of angry posts from users who feel like something they loved has been taken away.

For now, GPT-4o is back, GPT-5 is being patched, and the AI hype cycle rolls on. But this is a warning shot for OpenAI and every other AI company: in the age of anthropomorphized algorithms, you’re not just shipping code—you’re managing a fanbase. And fans can turn into critics overnight.

Altman has concerns, but is positive about the future of AI, source: X

Here is the rest of Sam Altman’s mini essay. It is worth reading –

People have used technology including AI in self-destructive ways; if a user is in a mentally fragile state and prone to delusion, we do not want the AI to reinforce that. Most users can keep a clear line between reality and fiction or role-play, but a small percentage cannot. We value user freedom as a core principle, but we also feel responsible in how we introduce new technology with new risks.

Encouraging delusion in a user that is having trouble telling the difference between reality and fiction is an extreme case and it’s pretty clear what to do, but the concerns that worry me most are more subtle. There are going to be a lot of edge cases, and generally we plan to follow the principle of “treat adult users like adults”, which in some cases will include pushing back on users to ensure they are getting what they really want.

A lot of people effectively use ChatGPT as a sort of therapist or life coach, even if they wouldn’t describe it that way. This can be really good! A lot of people are getting value from it already today.

If people are getting good advice, leveling up toward their own goals, and their life satisfaction is increasing over years, we will be proud of making something genuinely helpful, even if they use and rely on ChatGPT a lot. If, on the other hand, users have a relationship with ChatGPT where they think they feel better after talking but they’re unknowingly nudged away from their longer term well-being (however they define it), that’s bad. It’s also bad, for example, if a user wants to use ChatGPT less and feels like they cannot.

I can imagine a future where a lot of people really trust ChatGPT’s advice for their most important decisions. Although that could be great, it makes me uneasy. But I expect that it is coming to some degree, and soon billions of people may be talking to an AI in this way. So we (we as in society, but also we as in OpenAI) have to figure out how to make it a big net positive.

There are several reasons I think we have a good shot at getting this right. We have much better tech to help us measure how we are doing than previous generations of technology had. For example, our product can talk to users to get a sense for how they are doing with their short- and long-term goals, we can explain sophisticated and nuanced issues to our models, and much more.