On Monday, OpenAI announced the release of its new AI model, GPT-4o, which features realistic voice conversation capabilities and the ability to interact through both text and images. This move aims to keep OpenAI ahead in the competitive field of emerging technology.
The new audio features allow users to engage in real-time spoken interactions with ChatGPT, including the ability to interrupt it mid-speech, mimicking natural conversation—a challenge that previous AI voice assistants have struggled to overcome, as demonstrated by OpenAI researchers during a livestream event.
“It feels like AI from the movies… Talking to a computer has never felt really natural for me; now it does,” OpenAI CEO Sam Altman wrote in a blog post.
Backed by Microsoft, OpenAI faces increasing competition and pressure to expand the user base of ChatGPT, the chatbot that garnered global attention for its human-like written content and high-quality software code.
During the livestream, researchers showcased ChatGPT’s new voice assistant capabilities. In one demo, ChatGPT used its vision and voice to guide a researcher through solving a math equation on paper. Another demo highlighted the GPT-4o model’s real-time language translation abilities.
OpenAI’s presentations had a science-fiction feel, with ChatGPT engaging in playful banter with its user. At one point, the OpenAI researcher complimented the chatbot, saying, “You’re amazing,” to which ChatGPT replied, “Oh stop it! You’re making me blush!”
Altman later posted “her” on X, referencing the 2013 Spike Jonze film where a man falls in love with his AI assistant, voiced by Scarlett Johansson.
At the event, OpenAI’s Chief Technology Officer Mira Murati announced that the new model would be offered for free, as it is more cost-effective than previous versions. Paid users will have higher capacity limits than free users. The GPT-4o model will be integrated into ChatGPT over the coming weeks.
Additionally, free ChatGPT users now have access to a “browse” feature, allowing ChatGPT to display up-to-date web information. Murati clarified to Reuters that OpenAI does not plan to monetize free users through ads.
After its late 2022 launch, ChatGPT quickly became the fastest application to reach 100 million monthly active users. However, its website traffic fluctuated throughout the past year and is only now returning to its May 2023 peak, according to analytics firm Similarweb.
These announcements precede Alphabet’s annual Google developers’ conference, where new AI-related features are expected to be showcased.
Reuters recently reported that OpenAI had planned to announce an AI-powered search product, but this announcement has been delayed, according to a source familiar with the matter.
On Monday, shares of Alphabet fell 0.4 percent, having dropped nearly three percent earlier in the day. Microsoft shares declined by 0.2 percent.
