California: OpenAl on Thursday released GPT-4, a powerful AI system backed by Microsoft.
The latest version promptly responds to images and provides brief suggestions from photo of ingredients as well as writing captions and descriptions.
Millions of people used ChatGPT since it started last year in November. The modern version can process upto 25,000 about eight times as many as ChatGPT.
People have preferred its use for writing songs, marketing copy, poems, computer code and helping with homework.
However, the teachers suggested that the students should not use it.
This development led to the proliferation of human-like technology and increased competition between Microsoft Corp. and Alphabet Inc.’s Google, its backers.
OpenAI, which is the creator of the famous ChatGPT chatbot, explained in a blog post that its latest technology is “multi-modal,” which means it can generate content based on both text and image prompts.
While the text-input feature would be available to ChatGPT Plus subscribers and software developers with a wait-list, the image-input ability remains a preview of its research.
The release of GPT-4 by OpenAI, a highly-anticipated launch, demonstrates the increasing reliance of office workers on ever-improving AI for a variety of tasks. It also highlights the intense competition among technology companies to gain business from these advances. Recently, Google unveiled a “magic wand” for its collaboration software that can draft almost any document, just days before Microsoft is expected to showcase its AI-powered Word processor, which is likely to be supported by OpenAI. Additionally, a Microsoft executive has revealed that GPT-4 is assisting in powering its Bing search engine.
OpenAI has announced that its latest technology, GPT-4, represents a significant improvement over its previous version, GPT-3.5. According to OpenAI, in a simulated bar exam for US law school graduates, GPT-4 scored in the top 10 percent of test takers, while GPT-3.5 ranked in the bottom 10 percent. While the two versions may seem similar in everyday conversation, OpenAI says that the difference becomes apparent when the complexity of the task increases. OpenAI added that GPT-4 is more dependable, innovative, and capable of handling more nuanced instructions.
OpenAI President Greg Brockman demonstrated the technology online and showcased how it could create a real website based on a hand-drawn mock-up. The demonstration also revealed that GPT-4 could assist individuals in calculating their taxes. OpenAI CEO Sam Altman referred to GPT-4 as the company’s model that is most capable and aligned with human values and intent, although he noted that it is not perfect.
According to the company, GPT-4 is 82% less likely to respond to requests for disallowed content than its predecessor and scores 40% higher on certain tests of factuality. “Hallucinations,” or inaccurate responses, have been a challenge for many AI programs, but the company claims that GPT-4 has improved in this regard.
Rishi Jaluria, an analyst at RBC Capital Markets, said that Microsoft could gain advantages from the adoption of GPT-4. He explained that Microsoft is not only incorporating OpenAI’s latest technology into its products, but its Azure cloud is also powering the use of OpenAI. As many businesses are currently monitoring their IT spending in an unstable economy, this could be a strategic advantage for Microsoft. Jaluria suggested that whenever a company utilizes this technology, the workloads go through Microsoft Azure, and he considers this to be a critical time.
