OpenAI’s new GPT-4 model is the latest milestone in its attempt in scaling up deep learning. ChatGPT Plus, OpenAI’s paid user platform, now supports GPT-4 and developers can join a waitlist to gain access to the API.
OpenAI claims that GPT-4 achieves “human level” performance across a wide range of professional and academic benchmarks and that it can also accept image and text inputs where its predecessor, GPT-3.5, could only process text. For instance, GPT-3 achieves a score within the top 10 percentile on a mock bar exam.
OpenAI claims that after six months of progressively aligning GPT-4 using lessons from an adversarial testing program and ChatGPT, the system achieved “best-ever outcomes” in terms of factuality, steerability and refusal to step outside of guardrails.
When Will Chat GPT 4 be Released?
It’s likely that GPT-4 will come this year. Nevertheless, we have yet to determine an exact release date for it in ChatGPT. According to the New York Times, it may come as early as the first quarter of this year.
It has been said by Microsoft that GPT-4 will be released in March and that it will be a multimodal model capable of producing text, graphics and video. Nevertheless, how Microsoft and OpenAI plan to make use of GPT-4 is unclear at this time.
It’s probable that this model is being released for experimental purposes at this time. As a result, it may take some time before it appears in services like ChatGPT and Bing Chat from Microsoft.
Rowan Cheung provided a Tweet in which he shared the latest information related to the release date of Chat GTP 4.
A rundown of the most significant events in AI last week:
1. Microsoft announces Visual ChatGPT.
It allows you to send, receive and edit images during chatting👇 pic.twitter.com/ww3qi1c2bi
— Rowan Cheung (@rowancheung) March 13, 2023
More developments in the field of technology have recently been announced and you can read more about them by visiting the links provided below:
- Sony Built a $600 Point-and-Shoot Camera for the Blind
- The Role of Cross-Browser Testing in Responsive Design
- Using Email Validation to Improve Your Deliverability
What Happened When GPT-3.5 Compared With GPT 4?
When comparing GPT-3.5 with GPT-4, “the distinction can be minor in casual speech, OpenAI said in a blog post. When the task’s difficulty increases over a certain point, the differences become apparent. GPT-4 is superior to GPT-3.5 in reliability, creativity and the ability to handle more subtle instructions.
According to a recent research blog post, OpenAI has introduced GPT-4, the next iteration of their AI language model that can read photographs and explain what’s in them. In GPT-4, you can also use pictures as prompts.
OpenAI writes:
It generates text outputs given inputs consisting of intermingled text and graphics over a range of domains, including papers containing text and photos, diagrams or screenshots.
Essentially, this means the AI chatbot will be able to determine the content of an image. Chief Technology Officer of Microsoft Germany Andreas Braun recently stated that “GPT-4 will provide radically different possibilities for example, films.”
We hope that you will like this model. No official word on when Chat GPT 4 will be available has been made yet. We promise to keep you informed as soon as we learn anything new. If you’re interested in reading our most recent posts relating to different forms of technology, be sure to check out our Twitter page.