"OpenAI" breakthrough: trying to build an IoT model that extends human life by 10 years

Now OpenAI is trying to revolutionise the world of technology by developing a new model that the company claims can extend human life.

Aiming to 'restructure' proteins

Under the leadership of CEO Sam Altman, OpenAI has been collaborating with Retro Biosciences, a start-up in the field of longevity, for a good year now. 

Using biotechnology and protein rearrangements, they aim to extend human life by 10 years. 

To extend human life, they are studying so-called Yamanaka factors - a set of proteins that, when added to human skin cells, make them transform into youthful stem cells that can create any other body tissue.

However, such "reprogramming" of cells is not very effective. 

As a result, OpenAI has developed a new artificial intelligence model called "GPT-4b mircro".

"GPT-4b mircro outperforms scientists

"GPT-4b mircro is a language model that has been trained to suggest ways of redesigning protein factors to make their function more efficient.

According to OpenAI, researchers used the model's suggestions to modify two Yamanaka factors, making them more than 50 times more efficient - at least according to some initial measurements.

"These proteins look better than the ones that scientists have been able to create themselves," said OpenAI researcher John Hallman.

So far, just a reference model

"The GPT-4b mircro model does not work in the same way as Google's AlphaFold, which predicts what shape the proteins will take. 

Because Yamanaka factors are unusually volatile and unstructured proteins, OpenAI said they needed a different approach, for which its big-language models were suitable.

The model was trained on examples of many types of protein sequences, as well as information about which proteins tend to interact with each other. 

While this is a lot of data, it is only a small fraction of what OpenAI's example chatbots were trained on. GPT-4b is therefore an example of a "small language model" running on a targeted dataset.

Sam Altman is confident of success

Last week, Mr Altman said he is "confident" that his company knows how to build a general artificial intelligence (AGI) that goes beyond human capabilities.

Highly intelligent tools could significantly accelerate scientific discovery and innovation, far beyond what humans can do on their own, he added.