GPT-5 Rumor: Sam Altman-Led OpenAI Set to Revolutionize Chatbots with 'Materially Better' Release

OpenAI allegedly intends to launch GPT-5 in mid-year, which is touted as significantly improved.

Anticipating the Arrival of GPT-5: OpenAI's Next Evolution in ChatGPT Technology

The Verge reported that in the coming months, OpenAI plans to release the next version of its model for ChatGPT, the generative AI tool that sparked the current wave of AI projects and investments.

According to two people familiar with the company, the generative AI company led by Sam Altman plans to release GPT-5 sometime in the middle of the year, most likely during the summer. According to another person familiar with the process, some enterprise customers have recently received demos of the latest model and related ChatGPT tool enhancements.

These people, whose identities Business Insider confirmed, requested anonymity so that they could speak freely. "It's really good, like materially better," said one CEO who recently saw a demo of GPT-5.

According to the CEO, OpenAI demonstrated the new model with unique use cases and data. He added that the company hinted at other yet-to-be-released model capabilities, such as the ability to call AI agents being developed by OpenAI to perform tasks autonomously.

The company has not yet set a release date for the new model, so current internal expectations for its release may change. One of those familiar with the situation stated that OpenAI is still training GPT-5.

After training, the tool will be internally tested for safety and "red teamed," which is a process in which employees and a select group of outsiders challenge the tool in various ways to identify issues before making it available to the public. One of the people familiar with the situation stated that there is no set timeframe for completing safety testing, so the process could delay any release date.

Navigating OpenAI's Enterprise Sales and the Evolution of ChatGPT Technology

Sales to enterprise customers, who pay OpenAI for an enhanced version of ChatGPT for their work, are the company's primary revenue stream as it expands its operations and Altman builds his growing AI empire.

The last major update to ChatGPT was released a year ago by OpenAI. GPT-4 was billed as being much faster and more accurate in its responses than its predecessor, GPT-3. Later, in 2023, OpenAI released GPT-4 Turbo to address an issue known as "laziness," in which the model would sometimes refuse to respond to prompts.

Large language models, such as OpenAI, are trained on massive amounts of data scraped from the internet to respond to user prompts in an authoritative tone that resembles human speech patterns. That tone and the quality of the information it provides can deteriorate depending on the training data used for updates or other changes OpenAI may make in its development and maintenance efforts.

Several Reddit forums have been dedicated to complaints about GPT-4 degradation and poor output from ChatGPT. According to one source, OpenAI employees hope that GPT-5 will be more reliable and impress both public and enterprise customers.

Many of the most important training data for AI models are technically owned by copyright holders. OpenAI and many other tech companies have opposed new federal rules governing how LLMs access and use such material.

Photo: Jonathan Kemper/Unsplash