ChatGPT parent OpenAI on May 13 showcased its latest AI model, GPT-4o, with a demo featuring voice interaction across text and images. This could keep the company "ahead of the race" in the global artificial intelligence landscape, Reuters reported.
The GPT-4o boasts advanced audio capabilities, allowing users to engage in real-time conversations without delays and even interrupt the AI during its speech—a significant milestone in replicating natural human interaction, the report said. OpenAI researchers showcased these features during a livestream event, likening the experience to dialogue straight from the movies.
OpenAI CEO Sam Altman expressed his enthusiasm in a blog post, highlighting the newfound naturalness in conversing with computers, a feat previously considered challenging. "It feels like AI from the movies ... Talking to a computer has never felt really natural for me; now it does," Altman wrote.
Backed by Microsoft, OpenAI faces mounting competition and the imperative to broaden the user base of its popular chatbot, ChatGPT, the report noted.
During the livestream, researchers demonstrated ChatGPT's enhanced voice assistant capabilities. In one demonstration, the AI guided a researcher through solving a mathematical problem, leveraging its vision and voice functionalities. Another showcased the model's prowess in real-time language translation.
The demonstrations bordered on "science fiction", the report added. It noted playful exchanges between ChatGPT and its human counterpart, where at one point, a researcher stated he was demonstrating "how useful and amazing you are", to which the chatbot replied with: "Oh stop it! You're making me blush!"
Mira Murati, OpenAI's Chief Technology Officer, announced that the GPT-4o model would be provided free of charge, citing its superior cost-effectiveness compared to previous iterations. Paid users will enjoy expanded capacity limits, offering enhanced capabilities.
The GPT-4o model is slated for integration into ChatGPT in the coming weeks.
There was a range of reactions to the GPT-4o demo, from those who expressed excitement and welcomed the technology.
This demo is insane.
— Mckay Wrigley (@mckaywrigley) May 13, 2024
A student shares their iPad screen with the new ChatGPT + GPT-4o, and the AI speaks with them and helps them learn in *realtime*.
Imagine giving this to every student in the world.
The future is so, so bright. pic.twitter.com/t14M4fDjwV
the usages in accessibility for this will be huge! Imagine a camera and an open channel of communication. I could ask things on the fly. I am excited to try this! https://t.co/TlpCIHivQg
— Lucas Radaelli (@lucasradaelli) May 13, 2024
And those who shared potential use-case applications for the tool.
I can’t help but think how transformative this is going to be for the disability community. Those who are blind or have low vision or have cognitive disabilities or even speech disabilities. https://t.co/1evNCvkVbY
— carden ♿️ (@cardenonwheels) May 13, 2024
Others were excited by the "futuristic" and "sci-fi" aspect of the tech.
The future is going to be onchain & AI powered! What a time to be alive!
— Neeraj Khandelwal (@nrjkhandelwal) May 13, 2024
Very cool GPT-4o demo! https://t.co/x81oMGrT0x
Future is there. #GPT4o https://t.co/wHv2yAB3sg
— Qingqing_Chen (@qingqingparis) May 14, 2024
But excitement also gave way to some caution. Privacy and copyright — a long-time criticism of AI training models made its presence known.
Super fascinating technology.
— Chad (@Chad_Nelson_) May 13, 2024
A bit concerned about the potential privacy and security implications.
These AI models used by bad actors could do some significant damage.
I'm hopeful that these advancements can be harnessed for good. https://t.co/FJdUsz32WI
Excitement and worry grows more in every new amazing AI launch. I hope @OpenAI also ensures this cannot be used for identity theft. AI must not affect people's safety by any means and therefore must be regulated. #regulateAI https://t.co/3SPQnDVmgZ
— Kerem Iseri (@keremiseri) May 13, 2024
There was also the usual wonder at the tech leap, but worry for the future of human jobs as AI advances.
If you haven’t see this…and if you are interested in the direction of AI and how AI continues to evolve, check this out.
— Boots2Classes (@Boots2Classes) May 14, 2024
ABSOLUTELY WILD.
I can’t help but wonder just what this does for our profession…man… https://t.co/1wqlD7Scky
Some even verged on warnings, calling into question the legality of AI training models.
So before anyone gets excited ask - where did they get the data to train on? Who did they compensate? If the answer is they got no permission & paid no fees you need to call these what they are - products built on theft & they should not be able to do this legally. #GPT4O #GPT5 https://t.co/xK6LqOUCMF
— Kristine (@schachin on Threads) 🇺🇦 (@schachin) May 13, 2024
And others picturing doomsday scenarios.
Great. AI having conversations with each other. Next step they'll be conspiring against humans to create WWIII if we don't get there first. #StateOfTheWorld today https://t.co/Wxzzegf056
— Rachel Gray (@RachelG69564736) May 14, 2024
Christmas 2025, we are all going to be re-enacting scenes from Her, and Black Mirror. 18 months, full fluency.
— Joe (@josephradhik) May 13, 2024
I still cannot bring myself to type or talk to an artificial being, but I can see so many humans using this for hours on end.
Pair it with smart glasses, and full… https://t.co/hLgqfsiLKi
Besides all this, there was as usual, the memes that made an appearance.
In a world not far away we don’t even need humans to have a fight on X, we’ll just send the AI’s https://t.co/3FVcwJE6gH
— Kristian (@KayAgeOfEl) May 13, 2024
(With inputs from Reuters)