the health strategist
institute for continuous transformation
in health and tech
Joaquim Cardoso MSc
Founder and Chief Researcher & Editor
March 11, 2023
EXECUTIVE SUMMARY
OpenAI is set to release its latest language model, GPT-4, next week, according to Microsoft, one of the startup’s major investors.
- GPT-4 is expected to be multimodal, meaning it can interpret both text and images, and will be capable of generating text, images, and videos.
- This represents a significant step forward in the development of language models and is likely to have a major impact on the artificial intelligence sector.
- GPT-4 will also be able to manage the input and output of data in different languages, making it a useful tool for translation.
- Microsoft has emphasized that GPT-4 is not intended to replace jobs but to enhance and simplify work processes.
DEEP DIVE
GPT-4 will arrive next week and will be multimodal
GearRice
By Ronan Byrne
10/03/2023
We knew that GPT-4 would most likely be released in the first half of 2023.
Fortunately, OpenAI will not disappoint in this regard. Microsoft, one of the main investors in the startup specializing in artificial intelligence, has confirmed that the world will know the new language model next week. At the moment, yes, they did not reveal the specific day.
Although much of the technical news will be kept until the official announcement, those from Redmond have already advanced a key detail: GPT-4 will be multimodal.
What does it mean for a language model to be multimodal?
Basically, what can interpret both text and images.
The previous version, which brings ChatGPT to life, is only capable of producing texts.
With GPT-4, therefore, it would be possible to generate text, images and even videos.
Andreas Braun, director of technology at Microsoft Germany, spoke about it in statements collected by Heise:
“We will present GPT-4 next week. We will have multimodal models that will offer completely different possibilities. For example, videos.”
Andreas Braun.
Braun described GPT-4 as a game changer for the artificial intelligence sector, since the language model will take another step to get closer to the capacity of human interpretation.
Obviously, he still has a long way to go to match it.
Another benefit of GPT-4 as a multimodal language is that it You can manage the input and output of data in different languages.
Transferring this possibility to ChatGPT, you could make the request in Spanish, but the generated text is in English. Obviously, this novelty would also be of great help for translation tools.
From Microsoft they reiterate that the impact of ChatGPT in the daily life of so many people is similar to that caused by the iPhone more than a decade ago.
However, the popularity of the chatbot did not stop its aspirations to continue improving the language model that makes it possible. Hence the great attention they are dedicating to the presentation of GPT-4.
Likewise, the company directed by Saya Nadella has made it clear that GPT-4 is not here to steal people’s jobs.
For example, those who work in written media. The main goal is to simplify or enhance your work, not take it away.
“This is not about replacing jobs, but about doing repetitive tasks in a different way than before.”
Andreas Braun.
In fact, at Microsoft they recommend companies train their employees on the adoption of artificial intelligence in working life.
“Traditional job descriptions are changing. New and exciting professions are emerging as a result of the enrichment of new possibilities [impulsadas por la IA]”, commented Marianne Janik, director of Microsoft Germany.
Developing…
Originally published at https://www.gearrice.com on March 9, 2023.
Names mentioned
- Andreas Braun, director of technology at Microsoft Germany
- Saya Nadella, CEO
- Marianne Janik, director of Microsoft Germany.
