.jpg)
Reactions to OpenAI's much-awaited August 7 debut of GPT-5 have been conflicting. After months of anticipation for GPT-4's 2023 launch, customers were let down by what seemed to be minor enhancements. Many voiced dissatisfaction about the loss of previous AI models, while others claimed that the responses were "too formal". Updates are now being released by the Microsoft-backed business to allay these worries.
The GPT-5 launch
The CEO of the company has publicly acknowledged the difficult beginning. Sam Altman reportedly acknowledged, "I think we totally screwed up some things on the rollout," over a dinner with reporters. "Our API traffic doubled in 48 hours and is growing," he told The Verge, highlighting positive indicators. We have run out of GPUs. Every day, ChatGPT's user base reaches a new peak.
Altman’s damage control
Making ChatGPT Warmer
A change in GPT-5's personality is among the first adjustments made by OpenAI. The business responded to comments on X by saying that "ChannelGPT should feel more approachable now, though changes are subtle". Little, sincere remarks like "Good question" or "Great start", rather than flattery, will stand out to you. Users throughout the world are already receiving the update.
The head of ChatGPT, Nick Turley, admitted to the criticism, stating that while GPT-5's straightforward approach was popular with some, it turned off a lot of people. "GPT-5 was straightforward and concise. That's nice. My personality is that of a robot — I'm German, you know. However, a lot of people don't, and they appreciate that ChatGPT would genuinely check in with them," he said.
New Modes of Use
GPT-5 now has Auto, Fast, and Thinking modes to provide flexibility. Fast responds more quickly, Thinking responds more thoroughly, and Auto alternates between the two. While dependability is still being worked out, Sam Altman acknowledged on X that "GPT-5 will seem smarter starting today." GPT-5 appeared much less intelligent yesterday as a result of the autoswitcher breaking and being out of action for a portion of the day.
Legacy Models Return
The elimination of previous models was a major source of controversy among users. Many people asked for the return of GPT-4o during a Reddit AMA. In response, Altman stated that OpenAI had taken note of the comments and had reinstated GPT-4o for Plus subscribers under "Legacy models." Through settings, it is also possible to enable other older versions, such as o3 and GPT-4.1. TechCrunch claims that Altman has committed to more transparent "transition periods" in upcoming model phase-outs.
Expanding Limits and Access
Additionally, OpenAI has raised its usage caps. With GPT-5 in Thinking mode, Plus and Team subscribers can now send up to 3,000 messages per week. Customers of Enterprise and Edu can now access the model as well. In addition to these modifications, the business is enhancing its user interface to make it simpler to determine which model is answering a question.
ChatGPT is still expanding quickly in spite of criticism, and it currently has close to 700 million weekly users. OpenAI is placing a wager that its more adaptable and warmer GPT-5 will win back user confidence. The deployment serves as a reminder to Altman and his team of the difficulties involved in updating a product that is utilised by hundreds of millions of people globally.
.jpg)
.jpg)
.jpg)
