GPT-5, the long-awaited OpenAI model that will power ChatGPT, will reportedly arrive in just a few months.
According to Business expertOpenAI is expected to release a modern Immense Language Model (LLM) this summer. Moreover, some enterprise customers who have access to the GPT-5 demo claim that it is significantly better than GPT-4. “It’s really good, even materially better,” said the general manager who spoke to the publication. According to reports, the modern model still requires the cooperation of red workers, which means adversarial testing for ethical and security issues. The successful merger of red teams will ultimately determine the release of GPT-5.
OpenAI’s ChatGPT is largely responsible for sparking the generative AI craze that has seen Massive Tech companies like Google, Microsoft, Meta, and Apple develop consumer-facing tools. Google’s Gemini is a competitor that operates its own free-standing chatbot, as well as work-related tools for other products such as Gmail and Google Docs. Microsoft, a major investor in OpenAI, is using GPT-4 in Copilot, its generative AI service that acts as a virtual assistant for Microsoft 365 apps and various Windows 11 features. Google is reportedly in talks with Apple as of this week about potentially adding Gemini for iPhone, in addition to Samsung Galaxy and Google Pixel devices that already have Gemini features.
Expectations are high for the upcoming GPT-5 model. In a recent podcast interview with Lex Fridman, OpenAI CEO Sam Altman added to the confusion by saying, “I think [GPT-4] kinda sucks,” before adding, “I expect the delta between 5 and 4 to be the same as between 4 and 3,” referring to GPT-4’s enormous improvement over GPT-3. “I think our job is to live a few years in the future and remember that the tools we have today will suck looking back at them, and that’s how we make sure the future is better,” Altman continued.
Altman didn’t say when GPT-5 will be released or whether it will be called “GPT-5” at all, but he did confirm that “we will be releasing an amazing modern model this year.” Altman wasn’t sure what made it so amazing, but he said the technical side was especially essential in determining what the final product would look like.
Among OpenAI’s myriad achievements, such as a video generator called Sora, controversy quickly emerged. OpenAI ultimately did not release any information about how Sora was trained, leaving the developers to question whether their data was used without credit or compensation. OpenAI is also facing multiple copyright infringement lawsuits from news outlets, including one of these Modern York Timesand another coming from The Intercept, Raw Story and AlterNet. Elon Musk, an early investor in OpenAI, also recently filed a lawsuit against the company due to its confusing non-profit yet somewhat for-profit status.
