OpenAI, creator of ChatGPT and some other popular artificial intelligence, has announced the release of its latest AI model, the GPT-4. The company claims that this new AI model is a significant milestone in its deep learning research and will revolutionize the field of AI. Here are the details…
GPT-4 is an image and text-understanding AI model that can perform various tasks, including answering questions, completing sentences, and generating text, among other functions. The new version is an upgrade over its predecessor, GPT-3.5, which only accepted text inputs. The new model can accept image and text inputs and perform at “human level” on several professional and academic benchmarks.
GPT-4 is currently available via OpenAI’s API with a waitlist, and in ChatGPT Plus, OpenAI’s premium plan for ChatGPT, its viral AI-powered chatbot. OpenAI spent six months iteratively aligning GPT-4 using lessons from an adversarial testing program as well as ChatGPT.
The company reports that this process resulted in “best-ever results” on factuality, steerability, and refusing to go outside of guardrails. OpenAI claims that GPT-4 is more reliable, creative, and capable of handling much more nuanced instructions than GPT-3.5.
While there are concerns about the potential dangers of AI, OpenAI has been proactive in addressing these issues. The company has implemented guardrails to prevent its AI models from going beyond certain limits, and has also released a framework for measuring and reducing bias in AI models.