GPT 5: Reviews from an AVG User
So, after the very infamous OSS release, I was just waiting… patiently…for OpenAI to finally drop GPT-5. Because THAT’S the one update we’ve all been hyping in our heads for months if not years…
And then it happened. I was scrolling through Instagram like I normally would around 9 PM when I saw some tech page casually mention, “Sam Altman is going live at 10pm with a big announcement.”
Guess what I did?
Well… I was kinda asleep AHHHH!!
But when I woke up the next morning, my feed was flooded, actually with posts screaming “GPT-5 is here!” OpenAI this, OpenAI that… feature this, feature that…
But to get the real sauce? WE GO REDDIT, right peeps?
And ohhh boy… the vibe was the literal opposite. People calling it “trash,” “worst update yet,” “Sam didn’t deliver,” “This is NOT what we were promised with GPT-5.” You know the drill. And say what you want, but Reddit has that holy grail reputation to be brutally honest, just straight opinions from the trenches.
But well did I just outright believe them?
Not immediately. But I was curious enough to dig in…
So in this blog, we’re gonna unpack GPT-5 …the hype, the backlash, and whether it’s actually as bad as they say… especially if you’re not a hardcore ChatGPT user.
THE REASON WHY PEOPLE ARE ANGRY ON REDDIT:
So, I started digging and ….
People aren’t mad because “it’s not AGI” or whatever the tech people are arguing about.. Maybe it could be but … for normal people… it’s like they’re rather mad because their ChatGPT feels broken.
We’re talking:
- Short replies that cut off mid-thought.
- Less personality, more robotic “AI tone.”
- Story writers saying, “Why does my character suddenly sound like a bad audiobook narrator?”
- Plus users hitting chat limits within an hour.
- And here’s the worst, no way to just switch back to the old model.
One person put it perfectly:
“It’s like my ChatGPT suffered a severe brain injury and forgot how to read.”
Whereas some people didn’t even know there was an update. They just logged in like normal, started writing, and halfway through their ongoing story went, “Uhhh… something’s off.”
Some of them even started saying that they want the Chatgpt 4o version back..
“GPT5 sucks. Give me back my 4o please.”
Well.. word is out that they’ve had that version back on public demand
But the catch is it’s only for Plus users……
Well not to worry someone did crack the code on reddit..
Apparently, if at the start of every chat you type:
“Can you emulate the 4o model?”
…it suddenly wakes up, gets its personality back, and starts acting about 90% like the old GPT-4o.
The genius even found a way to make it permanent i.e.
Ask ChatGPT to save that 4o-style personality in your saved memories, and every chat will feel like 4o again.
Go share this blog right now with all your friends … who want the 4o version back…
Now.. let’s look into the positive sides… tooo… like what they’ve tried to do with GPT-5 and what’s their vision..
THE POSITIVES & FEATURES:
Well… my feed was filled with the features… and I’m sure you’ll have heard about them too…
But I wanna talk about these two features specifically that really caught my eye which seem both good and bad from my POV…
- GPT-5 as a unified MODEL
We all know that now GPT-5 is a culmination of every past OpenAI model.
Get My Equation’s stories in your inbox
Join Medium for free to get updates from this writer.
But how exactly did they do this?
Well there’s a thing called ROUTER…
A SIMPLE ANALOGY:
Imagine having an overthinking girlfriend …
sometimes she answers in one word… sometimes she’d write you a whole TED Talk message you didn’t ask for. And well.. On special occasions .. You might even get a curated message by her three other friends..
And that’s what the router does…
It’s exactly how GPT-5 picks its own brain by deciding whether to use fast mode, deep mode, or Pro, all automatically.
And unless you’re a heavy user, it might slip into overthinking mode when you just wanted a quick yes or no.
And maybe this could be the reason why people are hating GPT-5… Humans are emotional beings .. over time.. GPT-4o had created an emotional bonding towards people… like it understood our personality and gave detailed oriented answers now …. It’s just.. Giving straight forward answers..
Well that is exactly what brings us to our second feature.. Which is…
2. Reducing Sycophancy and Refining style
Here’s what Sycophancy mean:
Like that GPT-4o was known for behaving like a friend who always hyped you up. It was the people-pleaser model. You could say, “I’m thinking of quitting my job to open a coffee shop on the moon,” and it would reply with, “That’s such a creative idea! Let’s figure out how to make it happen.”
GPT-5? Not so much. It’s more like, “That’s unrealistic, and here’s why…”
On paper, this is a win. OpenAI says it’s to reduce “sycophancy” basically cutting down on the model just telling you what you want to hear. The idea is to make responses more truthful, more grounded, and less biased by your prompts.
But here’s where the emotional side kicks in and people got attached to that “supportive” vibe. GPT-4o felt like it knew you.
It’s like breaking up with a toxic ex. You know the new partner (GPT-5) is objectively better for you, more honest, more stable, less prone to making stuff up but deep down, you miss the chaos, the validation, and the way your ex (GPT-4o) always made you feel right.
And because of that, some people just… want to go back.
Is GPT-5 AGI?
This word is thrown around the GPT-5 release a lot but what does it even mean?
For us to understand what AGI means…
Well there’s no universally agreed-upon definition of AGI (Artificial General Intelligence).
But we can say that “a highly autonomous system that outperforms humans at most economically valuable work.” That doesn’t mean “good at chess” or “great at writing poetry.” It means a system that can handle anything from managing a complex coding project end-to-end, to using biology knowledge to solve engineering problems, to writing a Pulitzer-level novel .. without being spoon-fed prompts every two sentences.
By that standard, GPT-5 isn’t it.
The model has clear gains in accuracy, consistency, and personality control.
But it’s still brittle without careful prompting, still struggles with symbolic reasoning and math, still hallucinates facts, and still can’t reliably manage multi-stage tasks without human babysitting. These aren’t “we’re one tiny step away from AGI” problems, they’re “the architecture itself might have limits” problems.
But the headlines like “Glad I’m getting the translations for this newspeak ‘AGI is not useful’ means we aren’t even close.”
LAST OPINIONS
And that’s all I feel as a normal average user of Chatgpt..
So the conclude the question,
“Is GPT-5 a disappointment?
If you bought into the hype, probably yes. It’s not AGI but it’s a stepping stone towards AGI… It’s not perfect. It’s sometimes colder, more direct, and less “fun” than GPT-4o.
But if you strip away the hype and judge it for what it is.. a faster, more truthful, it could be actually pretty good.
What are your thoughts???? I’m no tech person… but maybe the AutoAgent GPT-5 has more potential..
Also the real question is, while we’re bickering over chatbot vibes, are we paying enough attention to the race? Because whether it’s Sam’s slow-and-safe or Zuck’s open-and-fast approach that wins… the finish line for AGI is getting closer every day.
And when we cross it, no one’s going to care whether their AI replied with emojis or em-dashes…
