OpenAI on Monday announced the desktop version of its chatbot ChatGPT and the latest version of the AI language model that powers the chatbot.
Mira Murati, OpenAI's chief technology officer, said during the company's livestreamed event that the new model offers improved speed and interactivity compared to the company's previous model, GPT-4.
The latest version, known as GPT-4o, eliminates response delays from ChatGPT and allows users to interrupt the conversation bot with a new query, thereby eliciting a modified response, Murati said. Ta. Murati added that the improved ability works in response to text, audio, and visual prompts.
“We're looking at the future of interaction between ourselves and our machines, and we think GPT-4o is really changing that paradigm,” Murati said. “This is the first time we've made significant progress in terms of ease of use.”
OpenAI researchers gave a live demonstration in which the new product identified the pace of your breathing and offered advice on how to relax, and in another demonstration, the product provided instant translations to communicate with Italian speakers. We made it possible to have conversations between English speakers.
In another demonstration, the product provided real-time guidance by viewing researchers' attempts to solve a mathematical equation through a cell phone camera.
GPT-4o is expected to be released in the coming weeks, Murati said, adding that the company plans to make the product available in stages to prevent abuse.
“Our team has been working hard to mitigate abuse,” Murati said. “We will continue to work with our stakeholders.”
OpenAI has been aiming to release new products and upgrades since launching ChatGPT in November 2022, reaching 100 million app users within two months. This performance set a record for the fastest growing app user base.
In March 2023, Open-AI released the latest version of its AI language model, GPT-4. Days after the release of GPT-4, OpenAI's CEO Sam Altman told ABC News that the product helped him score in the 90th percentile on the Uniform Bar Exam. He also scored near perfect scores on the SAT math test and can write computer code proficiently in most programming languages.
GPT-4 has the potential to be “the best technology ever developed by humanity” that will dramatically improve our lives,” Altman told ABC News at the time.
Monday's announcement comes three months after OpenAI announced its video generation tool Sora.
Just as ChatGPT responds to written input and Dall-E provides images, Sora creates up to one-minute videos based on user prompts. OpenAI said in a statement in February that the video generator is being used by a group of product testers but is not available to the public.
The risks posed by AI-generated content have sparked widespread concern in recent months.
A fake, AI-generated sexually explicit image of pop star Taylor Swift went viral on social media in late January, garnering millions of views.Last week, a fake image of pop singer Katy Perry at the Met Gala was widely circulated online, and even Perry's mother was fooled.
In February, OpenAI posted a statement online outlining steps the company had taken to prevent Sora from being exploited.
“Before making Sora available for use in OpenAI's products, we take several important safety measures,” the company's website states. “We work with red teams, experts in areas such as misinformation, hateful content, and bias, who adversarially test our models.”
In response to increased scrutiny of AI-generated content, technology platforms are taking steps to regulate such posts ahead of the November election.
Meta announced in February that it would begin labeling images created by OpenAI, Midjourney, and other artificial intelligence products. Social media site TikTok also announced in an online statement last week that it would also start labeling such images.
A series of major product releases in recent years have coincided with management turmoil at OpenAI.
Altman abruptly resigned from his role as CEO of OpenAI in November 2023. After an employee revolt and a public apology from one of the board members, Mr. Altman was rehired to the position just four days later.
The decision included conditions for OpenAI to reconstitute its board of directors. Altman's return appears to include input from Microsoft CEO Satya Nadella, who invested $10 billion in OpenAI last year.
At the end of Monday's event, Murati hinted at future product announcements.
“We also care very much about the next frontier,” Murati said. “We'll be updating you soon on our progress toward our next big goal.”
ABC News' Victor Ordonez contributed to this report.
