Discover How OpenAI’s Revolutionary AI Model is Transforming Longevity Science

Genetic engineers can influence the evolution of molecules in labs, but they have to limit their tests. A typical protein can be modified in countless ways because it’s made up of hundreds of amino acids, each with 20 options. However, OpenAI’s model often suggests changes that can alter half of a protein’s amino acids.

Joe Betts-Lacroix, CEO of Retro, explained that they took this model into the lab right away and saw impressive results. The model’s suggestions led to better outcomes than the original Yamanaka factors more than 30% of the time.

Vadim Gladyshev, a researcher from Harvard studying aging, highlights the need for improved methods to create stem cells. He remarked, “Reprogramming skin cells is simple, but other types are much trickier, especially in different species.”

It remains unclear how exactly the GPT-4b model makes its predictions. Betts-Lacroix likened it to AlphaGo’s unexpected methods when it defeated the world’s best player in Go. He believes they are just beginning to understand its potential.

OpenAI states that this collaboration did not involve any financial transactions. However, since the work could benefit Retro—whose major investor is Sam Altman, the CEO of OpenAI—there are ongoing discussions about potential conflicts of interest. Last year, the Wall Street Journal reported on Altman’s broad investments in tech startups, suggesting possible conflicts because some of these companies work with OpenAI.

Being connected to Altman and OpenAI might elevate Retro’s status, making it easier for them to attract talent and secure funding. There was no response from Betts-Lacroix regarding whether the company is currently seeking investors.

OpenAI clarified that Altman was not involved in this specific project, asserting that their decisions are not influenced by his investments.

Source link