OpenAI CEO Sam Altman teased that the new ChatGPT update “feels like magic” — and he wasn’t wrong.
The AI company basically planted a flag in the sand emblazoned with two words aimed at its Big Tech rivals: your move.
OpenAI CTO Mira Murati showed off the “Spring Update” to ChatGPT on Monday with a series of live demos. The newest version of the AI chatbot, powered by OpenAI’s new flagship AI model GPT-4o, can reason across audio, vision, and text in real time.
And it’s surprisingly human-like.
We’re getting crazy close to the movie “Her”
For starters, ChatGPT’s voice and conversational abilities took a huge leap forward thanks to GPT-4o, sounding capable of expressing emotion and varying its tone.
The new AI has what sounded like the voice of an American female in the demo — think something like Scarlett Johansson in Spike Jonze’s movie “Her” — although OpenAI researchers had it switch to a robot voice at one point. An OpenAI spokesperson said the audio output will be limited to a selection of preset voices at launch.
The voice didn’t just sound human-like. It also displayed an uncanny ability to mimic human dictation. The new ChatGPT giggles, adds humor, and moderates voice inflection depending on the prompts.
It seems to also be capable of picking up on some human cues. When a researcher was hyperventilating while practicing deep breathing, the chatbot said, “Mark, you’re not a vacuum cleaner.”
You can also interrupt the chatbot, which makes conversations feel more natural. You don’t have to wait for the AI to finish its response before jumping in with a clarifying question or switching the topic.
The response time was also lightning-fast. An OpenAI spokesperson said the chatbot can respond to audio inputs at a similar response rate to humans, taking 320 milliseconds on average.
Following the event, OpenAI CEO Sam Altman posted on X, formerly Twitter, with the movie title that was on many people’s minds after seeing the demos.
ChatGPT’s eyes also got an upgrade
The chatbot demonstrated elevated abilities to interpret a graph, help with coding, interpret emotions, and essentially tutor users on math equations by viewing video or images shown to a phone’s camera.
All the while, the voice assistant maintained a lighthearted and cheerful tone.
In a separate demo shared online, GPT-4o was even was able to analyze video of the space around a user, taking in that the person was wearing an OpenAI hoodie and surrounded by recording equipment, to guess that the person might be putting together some OpenAI-related announcement.
While the chatbot appeared to have a couple of hiccups, such as when it misinterpreted an image prompt or inaccurately started responding before the question was complete, these moments almost made the chatbot seem more human.
It all feels more human — and ahead of what we’ve seen from rivals
In one instance, ChatGPT started responding to a prompt before the researcher showed the equation on the camera, and the researcher stopped the chatbot in its tracks.
“Whoops, I got too excited,” the chatbot responded. “I’m ready when you are.”
It also seemed to reply with responses that seemed to mimic feelings of appreciation. When the researcher showed the chatbot a picture of writing that said “I heart ChatGPT,” it responded with, “aw” and said “that’s so sweet of you.”
In another instance, ChatGPT said the researcher was making it blush when he said he was talking about how “useful and amazing” ChatGPT was.
OpenAI made the announcements the day before Google’s big summer conference, Google IO, which is expected to reveal the company’s progress on its various AI products, like Gemini.
But the timing of OpenAI’s event — and its impressive demonstrations — will leave AI watchers curious to see if ChatGPT is ahead of Google’s Gemini, or if Google has something up its sleeve.
But for now, OpenAI’s spring update once again demonstrates just how impressive ChatGPT can be, especially when you compare it to the existing voice assistant space.
Amazon’s Alexa, Apple’s Siri, and Google are all on notice. Their voice assistants are known to give robotic and direct responses to questions — far from being truly conversational. The new ChatGPT powered by GPT-4o blows them out of the water with its human-like responses.
Apple, for its part, seems aware of the gulf between even older versions of ChatGPT and Siri, with a recent report saying the decision was made to overhaul the iPhone voice assistant after Apple executives spent weeks playing around with ChatGPT and the company realized how behind it was. There’s also rumors the two companies have been talking, and Apple could end up licensing OpenAI’s model for some yet-to-be-announced iPhone features.
Apple fans shouldn’t have to wait long for more information. The company is expected to unveil its AI updates at Apple’s annual Worldwide Developers Conference on June 10.
Meanwhile, Amazon had plans to release an “Alexa Plus” paid version of the voice assistant that’s powered by generative AI, Business Insider’s Eugene Kim first reported. The assistant is supposed to offer more conversational and personalized responses, but a release date isn’t clear.
But, just like it did with the first version of Chat GPT, OpenAI once again highlighted just how impressive its tech can be — and is leaving the rest of the tech industry to prove it can play catch up.