OpenAI has introduced GPT-4, the most recent version of its immensely popular artificial intelligence chatbot ChatGPT.
The new model is capable of responding to images by suggesting recipes based on shots of ingredients, for instance, as well as producing captions and descriptions. It can also process up to 25,000 words, which is almost eight times more than ChatGPT.
OpenAI’s GPT-4
ChatGPT has been utilized by millions of people since its inception in November 2022. Common requests include the creation of songs, poems, marketing copy, computer code, and assistance with homework, despite the fact that schools discourage its use.
ChatGPT answers questions using natural human-like language, and it can also reproduce diverse writing styles like songwriters and novelists, using the internet as it was in 2021 as its knowledge bank.
There are concerns that it could one day replace people in a variety of occupations. OpenAI said it had spent six months on safety features for GPT-4 and had educated it on human comments. Nonetheless, it cautioned that it may continue to spread misinformation.
GPT-4 will first be available to ChatGPT Plus users, who pay $20 per month for premium access to the site.
New ChatGPT Version Not Fully Reliable
It currently powers Microsoft’s Bing search product. The IT behemoth has invested $10 billion into OpenAI. In a live demo, it generated an answer to a difficult tax query – albeit there was no way to validate its answer.
GPT-4 is a sort of generative artificial intelligence, similar to ChatGPT. Generative AI employs algorithms and predictive language to create new content depending on cues.
OpenAI claims that GPT-4 has more sophisticated thinking abilities than ChatGPT. For instance, the model can identify available meeting times for three schedules.
OpenAI also announced new partnerships with the language-learning software Duolingo and the application for the visually impaired, Be My Eyes, to develop AI Chatbots that can aid their users through the use of natural language.
OpenAI has cautioned, however, that GPT-4 is not entirely dependable and may hallucinate a phenomenon in which AI invents facts or makes reasoning errors.
Read more: Teenage girls in Iran are being chased by authorities after they posted a dance on TikTok
