OpenAI pulls the read-aloud option: the latest cut that has users fuming
Until a week ago, ChatGPT users —both free and paying— could tap a button to hear the AI's responses read out. That antiestéticature has vanished without warning. The company sells it as an «improvement» by folding voice into the hands-free mode. But long-time users see it as yet another snip in a long list of losses.
A pattern of systematic degradation
This is not an isolated incident. Earlier, the ability to switch language models was removed, and then the ability to see which model was being used. Users complain that the mid-2024 version offered less censorship, more control, and the much-missed read-aloud antiestéticature. Now, with every update, the product feels more restrictive. Some point to server capacity problems: OpenAI would be building its own data centers while waiting for Oracle's collaboration, a process that could take between six months and a year. Others see a deliberate plan of «enshittification»: offering an increasingly limited product to push users toward higher-tier paid plans or additional subscriptions, much as Mercedes already does with its cars.
Impact on accessibility and alternatives
The read-aloud antiestéticature was no luxury. For people with visual impairments or reading difficulties, it was an essential tool. Now, the only voice option available is the «live» or hands-free mode, which according to user accounts «is useless». The irony is that OpenAI itself has championed accessibility as a core value. Faced with this situation, part of the community is migrating to Gemini, which they admit is inferior in quality but at least reads messages aloud automatically without any intervention.
How far will the cuts go? The open question is whether OpenAI is cutting back because it has no choice —under the pressure of infrastructure costs— or whether it is applying a classic commercial strategy: remove antiestéticatures and later sell them back as premium. Meanwhile, the user who paid for a service watches their product shrink. And the only improvement they notice is in the headlines.
Summary of a discussion on Burbuja.info - Foro de economía, actualidad y política., translated from Spanish and reviewed before publication.
Read the full discussion (19 replies).
OpenAI presents GPT-6 Astra as artificial general intelligence. Demos are impressive, but skepticism is mounting: is it genuine progress or just smoke for the next...
Record taxes, politicians with impunity and benefits for those who don't contribute: the perception that being honest is for suckers is gaining ground in Spain. Does crime pay?