OpenAI, a startup backed by Microsoft, has launched its latest artificial intelligence model, GPT-4, which is a successor to the technology behind the ChatGPT. GPT-4 is a powerful AI model that can generate content from both image and text prompts, and it outperforms its predecessor, GPT-3.5, in many aspects.
One of the most significant differences between GPT-4 and GPT-3.5 is that GPT-4 can use images as inputs to recognize objects in a picture and analyze them. Moreover, GPT-4 can generate responses of more than 25,000 words, while GPT-3.5 is limited to about 3,000-word responses.
GPT-4 is 82 percent less likely to respond to requests for disallowed content than its predecessor, and it scores 40 percent higher on certain tests of factuality. Developers can also choose the tone and verbosity of their AI’s style of conversation, such as assuming a Socratic style of conversation and responding to questions with questions.
GPT-4 has also shown its capabilities in the US bar exam and the Graduate Record Examination (GRE). It can also help individuals calculate their taxes and create a real website based on a hand-drawn mock-up. Be My Eyes, an app that caters to visually impaired people will provide a virtual volunteer tool powered by GPT-4 on its app.
However, GPT-4 has similar limitations as its prior versions and is “less capable than humans in many real-world scenarios.” Inaccurate responses known as “hallucinations” have been a challenge for many AI programs, including GPT-4. Additionally, GPT-4 generally lacks knowledge of events that occurred after September 2021, when the vast majority of its data was cut off, and it does not learn from experience.
Only the text-input feature of GPT-4 is available to ChatGPT Plus subscribers and software developers with a waitlist, while the image-input ability is not publicly available yet. GPT-4 powers Microsoft’s Bing AI chatbot and some features on language learning platform Duolingo’s subscription tier.
OpenAI Chief Executive Officer Sam Altman said that GPT-4 was “most capable and aligned” with human values and intent, although “it is still flawed.” GPT-4 can rival human propagandists in many domains, especially when teamed up with a human editor.
In conclusion, GPT-4 is a powerful AI model that can generate content from both image and text prompts. However, it has some limitations and challenges, including inaccurate responses and a lack of knowledge of events that occurred after September 2021. Nonetheless, GPT-4 has shown its capabilities in various fields and has significant potential in the future of AI technology.
For the latest tech news, follow Techbar Nepal on Facebook, and Instagram.
For Digital Products and Services: Maurisys Software.
