About three days ago, artificial intelligence company OpenAI released its most advanced language model. GPT-4introduced and everything changed. Although it may seem like an exaggeration, there is no limit to what can be done with the GPT-4 and it feels like you are carrying a knowledgeable Albert Einstein in your pocket. Let's see where and for what purpose GPT-4 is used in a short time.
We've talked about GPT-4 in detail, but let's briefly go over its capabilities. First of all, the GPT-4, unlike other models, has a multi-modal structure, that is, it can detect and parse both image and text. So GPT-4 can analyze the content of an image and relate that information to a written question. But for now DALL-E cannot render images like Midjourney or Stable Diffusion.In addition, the GPT-4 is better positioned for tasks that require creativity or advanced reasoning . In a demo, GPT-4 was asked to summarize part of a blog post using only words starting with "g" and the request was granted. Similar tests were done by independent people. GPT-4 is way ahead of GPT-3.5 in terms of ability to use the language. Even though GPT-4 is much more complete, the biggest problem is that it fabricates information and is definitely mistaken in misinformation. Therefore, it is important to check the GPT-4 outputs. Now let's look at examples of how GPT-4 will change the world.
Virtual assistant for the visually impaired
Since 2015, it allows blind and low vision people to ask sighted people to describe what their phone sees. Be My Eyesapplication was merged with GPT-4. This Virtual Volunteer can change the lives of people with disabilities. In its promotional video and blog post, Be My Eyes states that thanks to the GPT-4, it provides detailed information to its user about everything photographed. For example, if a user submits a picture of the inside of the refrigerator, the Virtual Volunteer will not only be able to accurately identify what's inside, but also be able to predict and analyze what can be prepared with these ingredients. The tool can also prepare recipes for these ingredients and explain how to make them step by step. Be My Eyes' Virtual Assistant can additionally read maps, identify the difference between two dresses or tell sticker prices. In short, it can do whatever is requested from the captured image.
