In his next futuristic essay, “The Gentle Singularity,” OpenAI CEO Sam Altman has reflected on how artificial intelligence will change human life in the next 15 years. The main focus is on the fact that by 2026, AI systems will be able to come up with “new, previously unknown knowledge and ideas.”
The new generation of AI – new scientific concepts
According to Altman, in the coming years, AI will not only analyze existing knowledge, but also be able to put forward new scientific hypotheses, innovative ideas and theories. This could affect areas such as science, pharmaceutical production and materials science.
Competition is intensifying
OpenAI is not the only player in this regard:
- Google has created new approaches to mathematical problems through its “AlphaEvolve” agent.
- FutureHouse (backed by Eric Schmidt) claims its AI tool has made real scientific discoveries.
- Anthropic has launched a program to support scientific research using AI.
A process viewed with suspicion
But there is also skepticism in the field. Thomas Wolf, chief scientific officer of Hugging Face, believes that “modern AI is not yet able to ask big questions.” Kenneth Stanley, a former lead scientist at OpenAI, has noted that today’s models have not yet reached the stage of “generating real hypotheses.”
Stanley is currently developing a startup called Lila Sciences — a $200 million investment to create a scientific laboratory using artificial intelligence. According to him, this is a complex problem that requires giving AI a concept of “what is creative and interesting.”
Is OpenAI’s direction clear?
Altman has previously revealed his plans for the future through blog posts: for example, he declared 2025 the “year of AI agents” and shortly thereafter introduced the Operator, Deep Research, and Codex agents.
So it’s possible that “The Gentle Singularity” is a hint at OpenAI’s next big step.
