Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Briefly
- Openii has built GPT-4B Micro, a reduced model specializing in engineering protein, in cooperation with retro biosciences to launch longevity.
- The model has designed new variants of Yamanak factors, proteins used to reprogram adult cells into stem cells, achieving 50 times more efficiency in laboratory tests.
- Researchers say the results show that AI could accelerate the sciences of life and exploration of longevity, although the work remains early and based on the laboratory.
And it is no longer just the ejection of the code, the paintings and the songs. It can now redesign the proteins within your cells.
On Post the company’s blogOpenai has just announced that he is cooperating Retro biosciencesStartup of longevity in the Silicon Valley, for training a specialized model called GPT-4b Micro. Unlike the chatbot you know, this model was not finely adjusted to punch or brainstorming. Instead, she was trained on protein data, biological text and 3D structure so she could Suggest brand new protein variants used in regenerative medicine.
The results were surprising: GPT-4b Micro successfully re-made two of the famous Yamanaka factors– Protein who won the Nobel Prize for their ability to bring adult cells back to stem cells. Stem cells are special cells that can both self -reflected (regenerate) and distinguish to many other types of cells in the body. They are important because they act as a system for body repair and have a huge potential for the treatment of diseases, tissue regeneration, and even the turnaround of the aging of aging.
In the lab they showed a AI designed version 50 times higher markers of stem cell markers and repaired DNA damage more effectively than the original. In other words, they have become old cells that are younger, faster.
Why is that important
Yamanaka factors are central in regenerative medicine, with potential for the treatment of blindness, diabetes, organ failure and more. But in practice, they are ineffective – less than 0.1% of cells usually convert to stem cells, and the procedure can take weeks. By finding variants that dramatically increase efficiency, AI could accelerate the study of cell reprogramming according to the yearsReducing the trial and error of conventional biotechnical.
This could shoot outwards:
-
Startups of longevity I can use AI Designed proteins to rejuvenate cells safer and more consistent.
-
Drug development deadlines They could be reduced if models like GPT-4b Micro become protein engineers on request.
-
Synthetic biology It could be switched to “what evolution gave us” and start exploring huge design spaces that were once impossible for people to move.
But also: big warnings
Science is a wound, and Openai admits that this is a proof of the concept. Laboratory validation is one thing; The move to clinical therapy is another. The engineering protein is notar for failure to translate from eating to the body, let alone in people.
There are and biological security concerns“If Ai can quickly design strong proteins, and then that power cuts in both directions.” Openai’s response is transparency: Working with retro is openly published so that others can repeat and criticize it.
For openness, this is not just one experiment; It is about showing that a language model can be diverted to a scientific discovery.
“When researchers bring insight into the deep domain to our models, problems that have lasted once years can change in days,” said Boris Power, who runs research partnerships at the company.
If this is true, then Ai will not just change the way we write or codes – he could start changing what means to grow older, treat and stay alive.
Generally intelligent Bulletin
Weekly AI journey narrated by gene, generative AI model.
