Sam Altman says of “mild singularity”; A new perspective of the future of artificial intelligence

In his latest memoir, “Mild Technique” released last Tuesday, Sam Altman, CEO of Openai, has offered his new view of the future of artificial intelligence and its impact on human experience in the next fifteen years.

This note once again presents an example of Altman’s special futurism; In the same tone that he has previously found in his writings, he promises to be the proximity of artificial intelligence (Agi), but at the same time diminishes and describes it as gradual. In this note, as always, Altman depicts a fundamental change in concepts such as work, energy, and social structure, and has talked about the path that Openai has taken to achieve.

Part of the note states that by year 3, “we will probably see artificial intelligence systems that are able to discover new insights.” Although this seems general, Openai officials have repeatedly pointed out in recent months that their future focus is the development of models that can provide innovative ideas and insights on the world.

During the introduction of the new “O3” and “O4-Mini” models last April, Greg Brockman, a founder and chairman of Openai’s board of directors, announced that these models were first used by scientists as a tool to produce new ideas. These developments are a sign that Openai will probably increase its efforts to develop a productive artificial intelligence that are capable of discovering and analyzing scientific insights.

Openai, of course, is not alone on this path. Google in May of a project called Alphaevolve Unveiled that he claims he could provide new approaches to solving complex mathematical problems. Also Startup FUTUREHOUSE With the support of former Google CEO Eric Schmidt, it has claimed that its operating tool has succeeded in a real scientific discovery. Company Anthropic This month, it launched a new initiative to support scientific research based on artificial intelligence.

If these projects are successful, they can transform the path of transformation in industries such as pharmacy, material science and basic science and automate the process of scientific research.

This is not the first time that Altman has been drawing a possible future path of Openai with blog notes. In January of this year, he introduced the year 2 as the “Year of Factors”; A phrase that came to reality with the introduction of the company’s three artificial intelligence factors – Operator, Deep Research and Codex.

However, achieving a system that can really create a “new insight” is a challenge far greater than just building an artificial intelligence factor. Some of the prominent figures in the field of science and machine learning are skeptical about the true ability of artificial intelligence to create authentic hypotheses. Hugging Face, Thomas Wolf, had argued in an article that current systems could not ask major and fundamental questions; Questions that are the cornerstone of scientific progress.

Kent Stanley, a former Openai researcher, also believes that current models lack the ability to create new hypotheses. He is now a new startup leadership called Lila Sciences It is responsible for attracting $ 5 million to establish a laboratory based on artificial intelligence to develop models with focusing on scientific hypothesis. From Stanley’s point of view, this is essentially a matter of creativity; That is, creating a sense of charm and authenticity in machine learning models.

While it is not yet clear that Openai will really develop a model that can produce real scientific insights, the Altman note can be seen as a roadmap that shows where the company is taking.