What are ChatGPT’s Future Plans? Revealed By OpenAI’s Sam Altman

An attendee from a recently held private meeting, in a blog post, disclosed some confidential information shared by OpenAI’s Sam Altman regarding ChatGPT’s future plans during the meeting.

Sam Altman has been attending conferences and meetings ever since OpenAI launched its AI chatbot ChatGPT last year. The CEO of OpenAI, in a recent meeting held in London reportedly shared some confidential information during the closed-door meeting which was leaked by one of the attendees.

Raza Habib, the co-founder of London-based startup Humanloop, wrote down a blog post explaining all that Altman said during the meeting. This information that was supposed to remain confidential was leaked through the blog post which has now been taken down on the orders of the OpenAI founder.

The post had already created a buzz before it got deleted and some crucial information regarding the company’s future strategies has been exposed through the same.

“A common theme that came up throughout the discussion was that currently OpenAI is extremely GPU-limited and this is delaying a lot of their short-term plans.” said the blog post.

The blog post mentioned that the company’s short-term plans are presently being delayed due to the shortage of Graphic Processing Units (GPUs). The GPUs are the main chipsets that are used to train and run the AI applications which the company is currently finding difficult to acquire in adequate numbers.

Sam Altman also stated that this is eventually affecting the developers who largely rely on the company’s services since the shortage of GPUs has slowed down the existing services of OpenAI.

“The biggest customer complaint was about the reliability and speed of the API. Sam acknowledged their concern and explained that most of the issue was a result of GPU shortages” mentioned Habib in the blog.

As per the blog post, Altman very precisely listed down all the things that are not able to be executed by the company due to the unavailability of adequate GPU chips. The company’s plans to increase the context window tokens for GPT-4 users from 8,000 to almost 32,000, announced in March, have also been affected due to the same since only a handful of users got access to increased tokens.

Sam Altman also shared OpenAI’s future roadmap for API, which basically includes the company’s goal of providing efficient and cheaper GPT-4 tech. The blog post said, “In general, OpenAI’s aim is to drive “the cost of intelligence” down as far as possible and so they will work hard to continue to reduce the cost of the APIs over time.”

OpenAI also plans to introduce multimodality in the year 2024, which was supposed to be launched along with GPT-4, but only once the shortage of GPUs is mitigated, this can happen.