In the wake of OpenAI’s long-awaited release of GPT-5, it appears that the company has learned one big lesson: the realization that it’s wildly out of touch with its customers.
One of the biggest revelations in the wake of the launch was the extreme degree to which users became attached to the preceding GPT-4o AI model. Many of OpenAI’s “power users” freaked out when OpenAI suddenly deprecated the model, meaning that they could no longer choose to use the uniquely sycophantic GPT-4o.
Users who loved GPT-4o were so upset, in fact, that OpenAI reversed course and reinstated the model for paying subscribers — but it remains apparent that the company was totally blindsided by just how passionately that subset of its user base felt not only about the model, but about ChatGPT’s various “personalities” as well.
In a recent podcast interview with The Verge, OpenAI’s longtime head of ChatGPT, Nick Turley, essentially said as much when referencing the backlash to GPT-5, which primarily took place on Reddit.
“I think the Reddit comments are really interesting to read because they show just how polarized users can be, where you can get some people with really strong opinions who love 4o and people who have really strong opinions on GPT-5 being better,” he told The Verge reporter Alex Heath. “The level of passion that people have for their choice is quite remarkable. And it recalibrated me a bit.”
Unfortunately, that recalibration may be too little too late.
“The first hire I made after ChatGPT was a data scientist, because I was so confused,” Turley told the website. “I would be talking to every user and they would tell me a different story as to why they were loving ChatGPT, and it was just deeply confusing to me and I had to get to the bottom of it.”
That admission itself is something of a head-scratcher. The era Turley was discussing would have been late 2022 or early 2023, and per his LinkedIn, he had been a vice president and head of ChatGPT since August 2022, when he jumped ship from Instacart — a clean vertical trajectory that would suggest he had the prowess to lead such a massive effort.
ChatGPT has gained upwards of 700 million active weekly users under his leadership and has, as he rightfully brags on his LinkedIn, become the fifth-most visited website in the entire world.
Over time, as Turley noted in the podcast interview, he did eventually figure out the “Cliff’s Notes” version of ChatGPT’s main use cases — writing, coding, “chit-chat,” and “searchy stuff like informational queries.”
In a broad sense, those three listed uses are accurate — but flattening the strange rationale behind the ways people use AI is kind of like saying that people like to use social media for fun and posting. There’s so much more there, and it’s clear that the man in charge of ChatGPT has not yet plumbed those depths.
Case in point: while Turley acknowledged that “a few things have changed,” ChatGPT is “not completely different from a year ago,” and those primary three use cases are largely “here to stay.”
“Obviously, we have done a bunch of work on the product,” the ChatGPT czar told The Verge. “That work you can break down into sort of pure model improvements like the behavior, the personality, its capabilities, its likelihood of refusing a request.”
Though he gets a bit closer when referencing the chatbot’s personality, it remains obvious that Turley has the broad strokes down, but doesn’t yet know what’s happening underneath the surface.
Even OpenAI CEO Sam Altman seems to have a better grasp on the situation than his ChatGPT VP.
In a post on X shortly after the initial 4o freakout, Altman suggested that OpenAI has been “closely tracking” the “attachment some people have to specific AI models” for a year or so behind the scenes, especially as more and more people use the chatbot as a “sort of therapist or life coach.”
Though Turley also admitted in the interview that OpenAI has been “tracking this type of thing for a while” and is “concerned about a world where people become overly reliant on AI,” it’s pretty obvious that the man in charge of ChatGPT has only recently begun asking why people feel that way about the chatbot — and that seems like a pretty glaring knowledge gap to us.
More on OpenAI: OpenAI Has a Major Problem on Its Hands With GPT-5