Site icon

OpenAI announces GPT-4 AI with Video Generating Capabilities, much advanced than ChatGPT

OpenAI has recently announced its new AI ‘GPT4’ that has the capabilities of video generation, solving difficult problems with accuracy. Microsoft’s CTO had already confirmed that the firm’s new AI ‘GPT4’ is coming this week.

OpenAI’s ChatGPT had stirred up the AI and the tech companies, insisting the capabilities and beneficiaries of the new form of AI – generative AI. Since its launch, Google, Microsoft, Baidu and more got in the row in implementing artificial intelligence into their applications and services, for making them better in experience and convenience.

In the go, Microsoft’s collaboration with OpenAI now steps ahead with the announcement of GPT-4, an advanced version of GPT-3.5 that powers ChatGPT.

GPT-4 AI

“We will present GPT-4 next week, because we have multimodal models that will offer completely different possibilities – for example videos,” starts Dr. Andreas Braun, CTO of Microsoft Germany. Now, OpenAI launched GPT-4, its most advanced system till date with greater problem solving efficiencies, quick and fast than ChatGPT.

Calling the large language models as game changers, CTO describes that it’s been possible to teach machines to understand natural language, as people. With the help of AI, machines can now understand languages in a statistical way what was previously readable and understandable only for people. Say, “you can ask in German and get the answer in Italian.”

GPT-4 allows for multimodal capabilities, intimating that it could put video creations into ease. Microsoft names the multimodal AI as “Kosmos-1”.

Kosmos-1 on pre-trained results show that the AI could classify images, answer questions about image content, automated labeling of images, optical text recognition and language generation tasks. Though, the modal is only 22% correct currently, the researchers are pushing it to unleash the benefits.

If released, this feat of AI would enhance and enrich the image-related searches and generative AI models.

What’s new about GPT-4?

Asking Doubts to GPT-4 with Images (Credits: OpenAI)

Related Posts


Applications of GPT-4

GPT-4

Speaking about the multimodal AI, Holger Kenn, who is the Chief Technologist Business Development AI, Microsoft trained Kosmos-1 for deploying it into Microsoft products with one of the examples including mapping millions of requests into the APIs via the cloud.

OpenAI had already collaborating with organizations and companies around the globe such as Duolingo, Morgan Stanley, Khan Academy, by building innovative GPT-4 powered platforms.

In a practically tested example, Speech-to-text phone calls in a call-center could be recorded and the agents would no longer have to manually summarize and type in the content, as AI would do the task. According to Clemens Siebler, 500 working hours could be saved daily with a large Microsoft customer in the Netherlands who receives 30,000 calls a day.

AI-supported document processing, semi-automation through the processing of spoken language in the call and contact center are the foremost beneficiary by equipping this advanced GPT-4. Means, GPT-4 would enhance the company’s exposure in B2B marketing.

Here is how you can use GPT-4.

However, researchers admit that AI would not always answer correctly, so it was necessary for validation. Microsoft is currently creating confidence metrics to address this issue.

“We are building a feedback loop around it with our thumbs up and thumbs down,” says AI researcher who worked on the project.

Do you think this new generative-AI can proffer mankind to lead a better lifestyle? Drop your thoughts on the AI…

(For more such interesting informational, technology and innovation stuffs, keep reading The Inner Detail).

Exit mobile version