GPT-4 is the latest language model for the ChatGPT AI chatbot, and despite just being released, it’s already making waves. The new model is smarter in a number of exciting ways, most notably its ability to understand images, and it can also process over eight times more words than its predecessor. It’s also much harder to cheat now.
You will have to pay to use the new version although it is locked behind a ChatGPT Plus subscription for now.
How do you use GPT-4 and ChatGPT?
Shutterstock
The easiest way to access ChatGPT is through the official OpenAI ChatGPT website. There’s a lot of interest in it right now, and OpenAI’s servers regularly hit capacity, so you might have to wait for a slot to open up to use it, but just refresh a few times and you should be able to get access.
If you don’t want to wait, you can sign up for a ChatGPT Plus subscription. This gives you priority access and you should be able to use ChatGPT whenever you want if you are a paid member. However, there is currently a waiting list for new subscribers, so you may still have to wait a bit.
You will also need to login if you want to use GPT-4. The default, free version of ChatGPT currently runs GPT 3.5, a modified version of the GPT3 model used since 2020. GPT-4 is, for now, a subscriber-only feature, but as more development is seen, it may become more widely available.
What can GPT-4 do better than ChatGPT?
GPT-4 is the next-generation language model for AI chatbots, and while OpenAI isn’t specific about the changes it’s made to the core model, it’s keen to point out how much it’s improved over its predecessor. OpenAI claims it can process up to 25,000 words at a time — that’s eight times more than the original GPT-3 model — and can understand much more nuanced instructions, requests and questions than GPT-3.5, the model used in the existing ChatGPT AI.
OpenAI also assures us that GPT-4 will be much harder to spoof, won’t spit out falsehoods as often, and will be more likely to reject inappropriate requests or queries that might see it generate harmful responses.
But GPT-4 also has some exciting new capabilities that early adopters are already putting to good use.
GPT-4 can understand images
GPT-4 is a multimodal AI language model, meaning it can understand text and other media, such as images. This might sound familiar if you are had a chance with Stable Diffusion AI art generation, but is more capable than that as it can respond to images and queries. This has led to some exciting uses, such as GPT-4 creating a website based on a quick sketch.. or being able to suggest recipes to the user after analyzing an image of the ingredients they have on hand.
Let’s get into the details now.
GPT-4 is multimodal and now accepts images as input and generates descriptions, classifications and analyses. 🔥
Below is one such example of giving an input image of ingredients and asking GPT-4 to generate a list of recipes. pic.twitter.com/mJMq8zLgkk
— Sumanth 🚀 (@Sumanth_077) March 15, 2023
Everything is better in programming
ChatGPT has already proven itself to be a capable developer, but GPT-4 takes it to a new level. Early users were able to get it to build basic games for them in just a few minutes. Both Snake and Pong were recreated from scratch, despite the users having almost no programming experience.
He can pass the exams
ChatGPT performed well as a human, but was under stress and cracks and seams could often be seen. But with GPT-4, this is much less likely to happen. In fact, it can perform so well in human tests that the GPT-4 was able to pass the unique bar exam in the 90th percentile of test takers. He also passed the Biology Olympiad test with a 99th percentile. In comparison, ChatGPT only managed to do so in the 31st percentile.
GPT-4 can create its own lawsuits
The combination of enhanced thinking and text comprehension has a lot of potential for the DoNotPay team. It’s working on using GPT-4 to generate “one-click lawsuits,” where robocallers would be sued if they spam you. Such a system could also be used to scan medical bills and identify errors or to compare prices with other hospitals to reduce bills. Then he could even craft a legal defense using the No Surprises Act.
He can understand humor
GPT-4 has a much better understanding of what makes something funny. Not only can he tell better jokes when asked, but if you show him a meme or some other funny picture and ask him to explain what’s funny about it, he can understand what’s going on and explain it to you.
GPT-4 limitations
Like ChatGPT before it, GPT-4 is not perfect. It is certainly a a worthy contender for the Google Bard, but it still has a long way to go before it’s absolutely flawless and can do just about anything.
At the time of writing, GPT-4 is trained on data collected up to August 2022, so there is no knowledge beyond that date. This creates severe limitations on what AI can do and means that it becomes less accurate as time goes on due to lack of up-to-date information.
Like its previous language models, GPT-4 is also prone to “hallucinations”, where it claims incorrect information as fact. This reportedly happens much less frequently with this model, but it is not immune, raising concerns about its use in accuracy-sensitive environments. He is also quite limited in his ability to learn from experience, so he may continue to make the same mistakes, even when pointed out to him.
GPT-4 is currently limited to 100 messages every four hours, even for ChatGPT Plus subscribers, and if you’re not already a member, you’ll need to join a waiting list and state your reasons for wanting to use it.
Editor’s recommendations
Categories: GAMING
Source: newstars.edu.vn