GPT-5 Release Date: Everything you need to know about OpenAI’s Next Chatbot

There’s a rumor that OpenAI’s ChatGPT is getting more powerful.

Two experts who spoke to Business Insider say GPT-5, the new AI model, is expected to launch this summer. Some businesses have already tried out a demo of the tool and are pleased with its capabilities.

Using this technology, OpenAI aims to create artificial general intelligence (AGI), or systems that are more intelligent than humans.

According to this lofty, sci-fi premise, an AI can think for itself without human supervision, which can produce more AI models of its ilk without human intervention. A breakthrough of this magnitude may either destroy or enhance the world.

As the race to build the best artificial intelligence heats up, here is everything you need to know about GPT-5.

What is GPT-5?

GPT-5 is the successor to GPT-4, OpenAI’s fourth-generation chatbot for which a monthly fee is required.

As the name implies, GPT stands for generative pre-trained transformer, an AI engine developed and improved by OpenAI to power ChatGPT. Like the processor in a computer, each new release of the chatbot is based on a brand new GPT with enhanced capabilities.

This AI chatbot can offer human-like responses and can even recognize and generate images and speech. Its successor, GPT-5, is reportedly expected to offer better personalization, make fewer mistakes and handle more types of content, such as video.

It is also worth noting that other companies, such as Google and Meta, have released their GPTs under their names, collectively known as large language models.

These artificial intelligence systems are based on the human brain and can generate text responses to conversations.

Is GPT-5 being trained?

In the latest report, OpenAI claims to have begun training GPT-5 in preparation for its release in mid-this year. Once its training is complete, the system will undergo various safety tests.

This process will include a “red team” test in which internal and external testers will test the bot and provide feedback on its strengths and weaknesses.

Based on a recent tweet from an OpenAI representative, it has been speculated that GPT-5’s learning process has begun.

In January, a leading researcher at OpenAI hinted that the company was training on a much larger GPU than usual. The revelation followed a separate tweet detailing how OpenAI had expanded its computing capacity.

The GPU, or graphics processing unit, is like a calculator that helps an AI model connect various types of data, such as associating an image with a text-based description.

According to Mr Altman, his company worked on GPT-5 on at least two occasions last autumn.

According to two people who attended his former venture capital firm Y Combinator’s alumni reunion last September, Mr Altman spoke of GPT-5 and its successor, GPT-6, as superior to their predecessors.

The CTN News reported in November that OpenAI was working on GPT-5, although no release date was announced.

Reports have also claimed that OpenAI’s boss has devised an audacious plan to procure the large quantities of GPUs required to train larger AI models.

According to the Wall Street Journal, Altman wants to raise as much as $7 trillion with the help of a global network of investors, governments, and energy providers to overcome supply shortages hindering technological progress.

What can GPT-5 do?

OpenAI’s chief spoke at the World Governments Summit in Dubai in February, saying that ChatGPT-5 will be more intelligent than its predecessors.

Mr Altman explained that this is more important than it may appear since these models are generally applicable. “So if it is a little bit smarter, it will be a little bit better at everything.”

According to Mr Altman, GPT-5 will require more data to train on, hinting at its brain power. He indicated that the plan is to use publicly available data sets from the internet and large-scale proprietary data sets provided by organizations. The last would include long-form writings or conversations, whichever format they take.

As Altman stated on Bill Gates’ Unconfuse Me podcast in mid-January, the multimodality concept will be important in the future: “Speech in, speech out. Images, eventually video. People clearly want that. We can push that further, but maybe the most significant progress will be in the area of reasoning ability..”

As a result, GPT-4 is currently capable of reasoning in very limited ways. Also, it is not reliable. If you ask GPT-4 most questions 10,000 times, one is likely to be pretty good, but it does not always know which one, and you want to get the best answer of 10,000 each time, so an improvement in reliability is necessary..”

GPT-5 versus GPT-4

To beat GPT-4, it must surpass GPT-4 Turbo, the next-generation model OpenAI released to subscribers in November.

The company’s most advanced artificial intelligence chatbot has knowledge of world events up to April 2023, as opposed to 2021 for GPT-4; it can analyze even longer prompts of up to 128,000 tokens or roughly the length of a 300-page book; it is better at following instructions; and it is capable of switching between tools based on user request, such as the Dall-E 3 image generator and Bing search engine.

GPT-4 has been tested by OpenAI and several researchers on real-life exams and has been shown to have a decent chance of passing the difficult chartered financial analyst (CFA) exam.

On the 2020 USA Biology Olympiad semifinal exam, it scored in the 99th to 100th percentile, aced the SAT reading and writing section, and excelled in the bar exam.

When will GPT-5 be available?

OpenAI has not yet announced GPT-5.

Also, it is unclear whether this project was affected by the turmoil at OpenAI in late November. On November 17, Mr Altman was ousted by the company’s board of directors.

Mr. Altman returned to the helm along with a new board following a period of tumult reminiscent of the dueling viewpoints on the future of artificial intelligence after five days of tumult.

Some ChatGPT users previously claimed the bot was running on a new AI model called GPT-4.5 Turbo, but that was false.

Looking at OpenAI’s GPT release timeline, we can see that the gap between updates is shrinking. In June 2018, GPT-1 was released, followed by GPT-2 in February 2019 and GPT-3 in June 2020.

The current free version of ChatGPT (GPT 3.5) was released in December 2022, followed by GPT-4 three months later in March 2023. Several recent updates have also been released, including a “turbo” version.

Will OpenAI announce GPT-5 at its developer day conference in November, or will June be the magic month again?