GPT-4: The next generation of generative AI will be able to create videos

Microsoft ‘s CTO announced that GPT-4 , the next generation of AI developed by OpenAI , will be released next week.


It is the year of Generative AI, and the impact of developments such as ChatGPT ratifies the expectant moment that this technology is experiencing. After being the object of desire from companies like meta, Amazon, Baidu, Google and other “monsters” of innovation, now the fourth version of the “Generative Pretrained Transformer” (GPT) is close and Microsoft shared more details.

It was the CTO of the Redmond company, Andreas Braun, who announced that the new generation GPT-4 will be shown on March 19 and that, in addition to faster processing for content generation, it will have multimodal models that will create video and another type of material, always using this generative capacity.

“We will present GPT-4 next week, where we have multimodal models that will offer completely different possibilities, for example, videos,” Braun said at an event developed by Microsoft Germany called “AI in Focus – Digital Kickoff”.

So far, technology based on GPT-3.5, such as the popular ChatGPT and others, are capable of producing responses to interactions via text. The ability to generate words in real time with a complex enunciative clarity and able to adapt that content to certain contexts – age, language and other variables – established a new paradigm in technology companies associated with Internet searches.

What ChatGPT thinks about GPT-4

We asked ChatGPT, the current text-generative technology powered by GPT-3.5, to find out what they think about the new version. Although he clarifies at all times that everything depends on OpenAI, he mentions that it could boost what it is currently capable of doing to new levels.

“If that happens, it will likely improve my ability to understand and generate text with greater accuracy, consistency, and context, which would allow me to provide more accurate and helpful responses to user questions and requests,” ChatGPT responded to RPP . “In addition, GPT-4 could enable the generation of more diverse and sophisticated content in different fields, which could open up new possibilities for natural language applications in areas such as marketing, healthcare, education, and scientific research.”

Differences between GPT-3.5 and GPT-4

With the arrival of this new version of generative AI, there are marked differences with the current GPT-3.5 and the capabilities associated with this technological leap. IN summary, there are 5 important aspects to consider:

Size: GPT-4 would likely be larger in terms of its data processing and storage capacity compared to GPT-3.5, which could improve its ability to understand and generate more sophisticated text.

Complexity: GPT-4 could have a more complex neural network architecture compared to GPT-3.5, which could allow for the generation of more coherent and detailed text.

Accuracy: GPT-4 might have higher accuracy in understanding natural language compared to GPT-3.5, which might improve your ability to answer questions more accurately and generate higher-quality text.

Speed: GPT-4 could be faster in data processing and text generation compared to GPT-3.5, which could allow faster responses and greater efficiency in processing large amounts of text.

Additional Capabilities: GPT-4 might have new capabilities or additional features compared to GPT-3.5, such as an increased ability to understand the context and intent behind text, or the ability to output text in different languages ​​or writing styles.