OpenAI, the artificial intelligence (AI) research firm responsible for ChatGPT, has postponed the launch of its much-anticipated Voice Mode feature, which was initially showcased at its product update event last month.
The company has stated that it requires "one more month" to further refine the feature, a decision that has sparked significant backlash from users and the AI community.
This feature enabled users to communicate with the chatbot through voice and engage in a conversation that sounded remarkably natural.
The demonstration drew parallels to the science fiction film "Her," which features a virtual companion voiced by Scarlett Johansson.
Shortly after the unveiling, Johansson issued a legal threat against OpenAI over the voice's similarity to her own, prompting the company to remove the voice from its library.
In explaining the delay, OpenAI acknowledged that it was unable to progress with a limited "alpha" release of the feature in June as originally intended, indicating that additional development is necessary.
Voice Feature Teased in May, Supposed to Launch in June
OpenAI initially hinted at the capabilities of its Voice Mode assistant for ChatGPT in May, promising an interactive and expressive feature that could engage in lively, real-time conversations and even interpret facial expressions.
The feature was slated for an alpha test with a select group of ChatGPT Plus subscribers by June.
This announcement sparked controversy on two fronts: first, due to the feature's unsettling realism and believability, and second, because of allegations that it had used Scarlett Johansson's voice to mimic her character from the movie "Her."
Now, OpenAI has announced that Voice Mode requires an additional month of development "to meet our launch standards," with plans for a broader release to all Plus subscribers in the fall.
This decision has fuelled further controversy surrounding the AI assistant.
The company has indicated that the delay is for additional safety tests, though specific details about these tests have not been provided.
Why the Delay?
OpenAI has provided some generic reasons for the postponement of Voice Mode.
The company has vaguely mentioned that it is "enhancing the user experience" and preparing its infrastructure to accommodate the millions of users expected to interact with Voice Mode.
However, another reason that OpenAI has subtly introduced may raise concerns: the company is "improving the model's capacity to identify and reject certain content."
This suggests that Voice Mode may be susceptible to misbehaviour, potentially indicating a deeper safety issue—or what the AI industry euphemistically refers to as "alignment"—a problem that has historically affected chatbots.
These issues can arise when chatbots generate inappropriate or harmful content, either spontaneously or in response to deliberate user prompts.
OpenAI had planned to introduce the new voice mode to a subset of its paid subscribers around this time.
However, due to these challenges, the feature will not be available until July, and even then, it may be significantly restricted, with full availability to all users delayed until autumn.
Voice Feature Delay Met with Backlash and Mockery
Despite the explanations provided, users were dissatisfied with the news, and OpenAI's announcement was met with a surge of criticism from the AI community.
Many critics were swift to highlight OpenAI's history of overpromising and underdelivering, contrasting its performance with that of its competitors.
Some have labelled the delay as a "rug pull," as they had subscribed to ChatGPT Plus based on the initial assurance that Voice Mode would be imminent.
Others have speculated that the impressive demos showcased in May were deceptively exaggerated, a tactic that would not be without precedent.
More broadly, many have simply ridiculed OpenAI's inability to deliver a product.
Meanwhile, its competitor Anthropic has just unveiled a significant update to its chatbot Claude, now known as Claude 3.5.
AI Youtuber Matt Wolfe has suggested that the delay is a strategic move by OpenAI.
Benjamin De Kraker, creator of FinalFrame AI, expressed his frustrations, noting that OpenAI often promotes features that are later discontinued due to lack of use or interest.
The delay has also prompted users to reconsider the worth of their ChatGPT Plus subscriptions.
Peruvian medical researcher Patrick Wieghardt indicated that he might terminate his subscription, while another user confirmed they had already cancelled theirs.
An AI newsletter publisher wrote:
"Right now I see no point in paying for what you're offering for free anyway."
Some observers predict that the full release of Voice Mode could be several months away.
Ryan Morrison, AI Editor at Tom's Guide, estimates that mid-November, after the US elections, is a more realistic timeframe.
This latest setback, unrelated to the previous PR debacle involving Scarlett Johansson's voice, has only subjected OpenAI to further scrutiny and deepened the disappointment among its supporters.
JinseFinance
Miyuki
Brian
Hui Xin