4 things we learned from OpenAI’s GPT-5 Reddit AMA

The day after launching the potent new GPT-5 AI model, OpenAI CEO Sam Altman and a number of other researchers and engineers visited Reddit for the venerable Ask Me Anything thread.

Although a wide variety of technical and product-related topics were covered in the conversation, a few subjects stood out as being very significant to posters because of how frequently and passionately they were brought up. Some of the most noteworthy things we discovered at the OpenAI AMA are listed here.

Pining for GPT-4o

The most common issue in the AMA was the lamentations of users who cherished GPT-4o and felt personally assaulted by its discontinuation. As one user wrote, “BRING BACK 4o GPT-5 is wearing the skin of my dead friend,” that isn’t hyperbole. “What an evocative image,” Altman responded. Okay, we’re working on something right now. We heard you on 4o.

Additionally, this was not a single request. Another post made the case that GPT-4o and GPT-4.1 should be kept alongside GPT-5 since they have unique personalities and inventive rhythms. Altman acknowledged that they were currently “looking into this.”

One person asked, “Why are we getting rid of the variants and 4o when we all have unique communication styles?” The most of the demands were a little more subtle. Bring them back, please!

Altman conceded the point in a succinct but straightforward response. Okay, we hear you all on 4o,” he added. “Thanks for taking the time to provide us with the criticism (and the enthusiasm!).” We want to restore it for more users, and we’ll monitor usage to decide how long to keep it available.”

It’s intriguing that a large number of frequent users appear to like the older model’s design over the arguably superior newer models.

Filtering history

Another hot topic was ChatGPT’s safety filter, which was criticized for being overzealous by many users both before and after GPT-5. As an example, one user explained how they had been marked for talking about historical subjects. Their comment regarding Gauguin was flagged and removed because the artist was a “sex pest,” and the user’s own clarification question was flagged as well.

Altman’s response was a mix of agreement and sobering. “Yes, we will keep making improvements to this,” he stated. “It’s really difficult; sometimes it’s hard to tell where the lines are.” Although he acknowledged that the line separating safe and dangerous content is far from perfect, he emphasized that OpenAI intends to offer “very wide latitude” and that “people should of course not get banned for learning.”

New tier

An other query focused on a weakness in OpenAI’s subscription structure: “Are you guys going to introduce an additional plan for non-professional solo power users? Some people find the $20 plan to be insufficient, and the $200 tier is excessive.

“Yes, we will do something here,” was Altman’s concise response. Just a confirmation that the notion is on the table, no further information. This succinctness gives room for everything from “next week” to simply stating, “the discussion starts now.” For heavy users, however, who are limited by the Plus tier but are unable to defend enterprise pricing, the price difference is a significant concern. The way devoted individual users interact with the platform may change if OpenAI does develop an intermediary tier.

The future

Altman provided some further details regarding the present and future status of ChatGPT and GPT-5 at the conclusion of the AMA. Writing that “we expected some bumpiness as we roll out so many things at once,” he began by acknowledging that there were some problems with the release. However, it wasn’t quite as smooth as we had imagined.

GPT-5 didn’t look as amazing as it should have till now because of that bumpiness.

According to Altman, “GPT-5 will seem smarter starting today,” “Yesterday, we had a sev [severity, meaning system issue] and the autoswitcher was out of commission for a chunk of the day, and the result was GPT-5 seemed way dumber.”

Additionally, he guaranteed the eventual reintroduction of GPT-4o, at least for those same subscribers, and increased access for ChatGPT Plus users, with double the rate limits. The American Medical Association did provide a more lucid picture of what OpenAI is prepared to alter in response to public outcry.

At least the reinstatement of GPT-4o for Plus users recognizes that other metrics are more important than raw capabilities. In addition to the personality types planned for GPT-5, future editions of GPT-5 and beyond might be produced with more intentional flavors if consumers are this outspoken in preserving an older model.