The company that created the ChatGPT app, which can generate essays, poems, or computer code at will, released a long-awaited update to its artificial intelligence (AI) technology on Tuesday, claiming it would be more accurate and secure than its forerunner.
Since ChatGPT made its debut in late November, wowing users with its capabilities based on an older iteration of OpenAI’s technology known as a large language model, GPT-4 has been eagerly anticipated.
According to a company blog, “We’ve created GPT-4, the latest milestone in OpenAI’s effort to scale up deep learning,” and the AI system “exhibits human-level performance” on some academic and professional tasks.
According to the company, the model is “more creative and collaborative than ever before” and will “solve difficult problems with greater accuracy” than its previous iterations.
With its update, GPT-4’s text responses will be more accurate and, in the future, will originate from both image and text inputs, marking a significant advancement in the technology, though this feature has not yet been made public.
For instance, if a user sends a picture of the inside of a refrigerator, GPT-4 will not only recognize the contents with accuracy, but will also suggest recipes that can be made using those ingredients.
The next advance was being prepared, according to OpenAI, in collaboration with a partner business, Be My Eyes.
Much of the new model’s firepower is now available to the general public via ChatGPT Plus, OpenAI’s paid subscription plan and on an AI-powered version of Microsoft’s Bing search engine that is currently being tested.
Microsoft, which announced earlier this year that it would invest billions of dollars in the research firm, supports OpenAI.
The Windows manufacturer then quickly added the technology to its Edge browser, Bing search engine, and other products.
With Google announcing its own versions of the AI technology in response to Microsoft’s aggressive adoption of ChatGPT, Amazon, Baidu, and Meta have also entered the fray in an effort to keep up and beat Google.
decreased “hallucinations”
In contrast to its earlier chatbot, which had numerous interactions with ChatGPT or Bing’s chatbot in which users were exposed to lies, insults, or other alleged “hallucinations,” OpenAI claimed that the new version was much less likely to go off the rails.
“GPT-4 was strengthened and made safer over the course of six months. GPT-4 is 40% more likely to produce factual responses and 82% less likely to respond to requests for content that is not permitted “said OpenAI.
GPT-4 “is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it,” founder Sam Altman acknowledged, despite the anticipation.
“The algorithm’s power will grow, but it’s not a second revolution,” said Robert Vesoul, CEO of French AI startup Illuin Technology.
Vesoul questioned the safety precautions taken by OpenAI, which has already come under fire from billionaire Elon Musk for allegedly excessively censoring speech on its AI in an effort to avoid further embarrassing reactions.
“I’m not sure if I want an AI to prevent comments on uncharted territory. Should an AI decide whether or not I smoke?” said Vesoul to AFP.
Morgan Stanley, one of the companies collaborating on the GPT-4 rollout, will use AI to help its bankers and their clients.
In a statement from Morgan Stanley, Jeff McMillan said, “You essentially have the knowledge of the most knowledgeable person in Wealth Management-immediately.
The well-known online tutoring company Khan Academy and Stripe, a financial app that will use GPT-4 to combat fraud and for other purposes, are additional partners.
