ChatGPT 4 is on the way, and rumors say that it might make OpenAI’s ChatGPT’s already impressive language skills even better. To be clear, OpenAI’s next product is not likely to be called ChatGPT 4. But we got creative and put the names ChatGPT and GPT-4 together, which is the name of the improved AI model that will run it in the future. Let’s talk about GPT-4, how ChatGPT works now, and when OpenAI might put out its next big update.
Table of Contents
Open AI Launches Chat GPT 4
After all the rumors, it looks like the popular AI chatbot ChatGPT’s language module, GPT-3, will be changed next week. Andreas Braun, the CTO of Microsoft Germany, told that the GPT-4 language module would be released sometime next week. The new language model will include “multimodal models” like videos, Bruan told the publication, as quoted in the article.
Also Read: Bing AI Chatbot Vs Open AI ChatGPT! Are Chatbots Better Than ChatGPT?
Braun is said to have called Large Language Models (LLM) like GPT a “game changer” because they let machines understand “natural language” and respond in a natural way. When GPT-4 comes out, it could be used to improve ChatGPT and Microsoft services that use Bing AI.
Microsoft’s global team hasn’t confirmed the GPT-4 presentation yet, but on March 16, the company’s CEO Satya Nadella will talk about “The Future of Work: Reimagining Productivity with AI.” The report says that on March 9, Microsoft Germany CTO Braun spoke at an event called “AI in Focus” about the GPT-4 project. CEO of Microsoft Germany Marianne Janik also went to the event and talked about how AI is changing how businesses work.
Holger Kenn, who is the Chief Technologist of Business Development for AI and Emerging Technologies, and Clemens Sieber, who is a Senior AI Specialist, are also taking part. They both talked about how GPT-4 could be used.
The report says that “multimodal AI” like GPT-4 can do more than just turn text into pictures. It can also make music and videos with just a few simple steps. A senior AI specialist at Microsoft named Clemens Siebler is said to have shown some use cases, such as how AI can be used to make speech-to-text phone calls better.
Some AI platforms, like Otter.ai, already convert speech to text, and some video conferencing apps have live captions. GPT-4 can take these technologies to the next level. But people in the industry worry that new technologies like ChatGPT, which is powered by GPT 3, and Bing AI, which is powered by GPT 3.5, could make it harder to find jobs in the future.
The CEO of Microsoft Germany, Marianne Janik, talked about these worries, the report says. She said, “It’s not about getting rid of jobs; it’s about finding new ways to do things that have always been done the same way.” Janik also talked about how Germans work and said that the companies in that country have “a lot of history.”
What is ChatGPT 4?
OpenAI is making GPT-4, a new language model that can write text that sounds like real human speech. It will make the technology of ChatGPT, which is based on GPT-3.5, better. “Generative Pre-trained Transformer” (GPT) is what it stands for. It uses artificial neural networks and deep learning to write like a person.
The simulated neurons in ChatGPT have billions of parameters that can be changed to get the result you want. The huge amount of data helps the computer work more like the human brain, which has billions of neurons. OpenAI hasn’t confirmed any details about GPT-4, but they have confirmed that work is being done on it.
But there are many rumors going around. For example, Wired said that experts in the field thought that GPT-4 would have 100 trillion parameters as early as August 2021. Adding more parameters to an AI doesn’t make it work better, and it could change how quickly it responds.
Also Read: Grammarly Launches Its Own Generative AI Tool “GrammerlyGo”!
Other rumors say that it will be easier to make computer code and that you will be able to make both text and images from the same chat interface. People also want AI to be able to make videos. A multimodal model can work with text, pictures, and videos. Expert on machine learning Emil Wallner said on Twitter that GPT-4 might be able to do this.
GPT-3.5 can hold up to 2,048 words, but the two DV models, which may be GPT-4, can hold four times (8K resolution) and sixteen times (32K) that number. This means that conversations can go much deeper and help solve much bigger problems. At the end of the day, we can only guess what will happen. We only know for sure that GPT-4 is being worked on and that it could greatly improve ChatGPT results.
When Will Chatgpt 4 be Released?
GPT-4 is coming, and it’s likely to come this year. We don’t know exactly when it will be added to ChatGPT, though. The New York Times said it could happen as soon as the first three months of this year. Since it’s March, the launch will happen in a few weeks. Microsoft confirmed that GPT-4 will be out the week of March 13, and that it will be a multimodal model that can make text, images, and even video. Right now, it’s not clear how Microsoft and OpenAI will use GPT-4. It’s possible that the model is being tested, so we might not see it in ChatGPT or Microsoft’s Bing Chat for a while.
Will GPT-4 Be A Multimodal AI?
In a podcast interview from September 13, 2022 for the AI for the Next Era show, OpenAI CEO Sam Altman talked about how AI technology will change in the near future. It’s very interesting that he said a multimodal model would be coming soon. Multimodal means that something can be used in more than one way, such as with text, pictures, and sounds.
OpenAI sends text messages to people to talk to them. You can only talk through text whether you use Dall-E or ChatGPT. Speech can be used by an AI that can talk in more than one way. It can follow your instructions and either tell you something or do something for you. Altman never said that GPT-4 will have more than one way to get around.
Also Read: Open AI Launches ChatGPT and Whisper API! Check Out All the Pros and Cons.
But he did say that it was coming up soon. I think it’s interesting that he sees multimodal AI as a way to build new business models that can’t be done right now. He said that multimodal AI was like the mobile platform in that it allowed thousands of new businesses and jobs to start up.
Altman said, “I think this is going to be a huge trend, and this will be the interface for very large businesses. I also think that these very powerful models will be one of the real new technological platforms that we haven’t really had since mobile.” After that, there are always a lot of new businesses opening up, so that’ll be cool.” When asked what the next step would be in the development of AI, he said that certain things would happen.
Conclusion
OpenAI needs, among other things, money and a lot of computing power to keep going. The New York Times says that Microsoft is in talks to put another $10 billion into OpenAI. Microsoft has already put $3 billion into OpenAI. The New York Times said that GPT-4 should be out in the first three months of 2023. Matt McIlwain, a venture capitalist who knows about GPT-4, was quoted as saying that it might be able to do more than one thing.
Microsoft could also use the GPT-4 language module in its services, but at first it might only be available in the German market. Next, we can expect more information, especially at Satya Nadella’s talk on March 16 at 8:30 PM called “The Future of Work: Reinventing Productivity with AI.”
Open AI, which made ChatGPT and the GPT-3 language module, has been working with Microsoft to make their services better. Microsoft just put out the latest version of Windows 11, and now Bing AI will be used to search from the taskbar. Bing AI is built on Prometheus, which is a technology owned by both Microsoft and Open AI. You can use the AI chatbot in the Microsoft Edge browser and the Bing app for Android and iOS.
