OpenAI, with the release of a new version of its AI, has created quite a buzz, more so among AI enthusiasts. Recently, even as the OpenAI president and co-founder Greg Brockman gave a product tour, explaining its capabilities, one cannot help but wonder if it will hit the mark.
ChatGPT-4, the most advanced AI model yet, has caused heads to turn with its new features. What particularly has attracted attention is its ability to respond to images. In another revelation, Microsoft confirmed that its Bing chatbot has been running on GPT-4 for the past few weeks, with hopes of increasing productivity. And with this, Microsoft hopes to beat Google search.
OpenAI says GPT-4 will be able to process words as many as 25,000. What’s astounding is the claim that it can mimic a particular writing style. Further, it can also respond to images and generate responses accordingly. It’s important to note that OpenAI claims it is 82% less likely to give disallowed replies while being able to offer 40% more factual responses.
Even so, OpenAI has been rather direct in admitting the limitations of ChatGPT-4. “Social biases, hallucinations, and adversarial prompts” were the words used to describe its shortcomings. The fact that Bing Chat, powered by ChatGPT-4, has shown inappropriate responses in the past confirms this statement. Only time will answer the question that still lingers- ‘will ChatGPT-4 overcome the flaws of its predecessor?’
TO CONNECT WITH US:
WhatsApp: +91 9222 47 9222
Skype: Abservetech Support
Email ID: [email protected]