First of all, it doesn't mean that AI has proved Riemann's hypothesis, nor does it mean that AI has replaced mathematicians. What's really interesting is that AI is beginning to demonstrate a more important ability: to engage with fundamental scientific problems that humans haven't solved yet, and to discover connections between existing knowledge that humans haven't discovered before.
If this capability continues to improve, AI could have an impact on science that goes far beyond "helping scientists write code, look up papers, and calculate."
Much of the scientific progress of the past few hundred years has depended on a small number of extremely good human scientists. The appearance of Newton, Maxwell, Einstein and others is very scarce, and a person's time, attention and experimental ability are also limited. Even if a genius works 20 hours a day, it is impossible to study ten thousand directions at the same time.
AI has no such limitations.
If an AI scientist can read a large amount of knowledge, make assumptions, make mathematical deductions, design experiments, analyze results, and correct his own assumptions, then it may gradually change from a "scientist's tool" to a "participant in scientific research."
More importantly, AI can replicate.
A good human scientist cannot replicate himself into a million in an instant, but an AI research system can theoretically run into thousands or more research examples, allowing them to explore different problems at the same time.
This means that scientific research may gradually change from "serial exploration" that relied heavily on a few talents in the past to large-scale parallel exploration.
What is really revolutionary is the next step.
If AI can not only study mathematics, but also connect robots, automated laboratories, supercomputers and real-world scientific instruments, then such a closed loop may be formed in the future:
AI makes assumptions.
AI design experiments.
The robot performs the experiment.
The instrument produces data.
AI analysis results.
AI revises theory.
AI proposes the next round of experiments.
This cycle can be repeated over time.
At this stage, AI is no longer just "answering human questions" but begins to participate in "generating new knowledge."
Further, if AI can help study the next generation of AI itself, then a more complex feedback loop could emerge:
AI improves AI.
Stronger AI discovers new materials.
New materials make better chips.
Better chips provide more computing power.
Stronger AI studies energy and nuclear fusion.
Cheaper energy provides more computing power.
More computing power further enhances AI.
This means that AI may accelerate not only one scientific field, but also mathematics, physics, chemistry, biology, materials, energy and computer science.
Of course, this is still a long way to go before it is truly realized.
AI is still prone to mistakes and cannot fully understand the real world. Mathematical reasoning must be strictly verified, scientific predictions must be confirmed experimentally, materials must be truly manufactured, drugs must be truly experimentally and clinically verified, and nuclear fusion must truly solve material, tritium cycle and engineering problems.
Therefore, what is really noteworthy about Claude's progress in the direction of the Riemann conjecture is not that "AI has solved the Riemann conjecture."
Rather, it allows us to see a future that may be emerging:
In the past, mankind relied on a small number of talented scientists to explore the laws of nature.
In the future, humans may have AI scientists who can replicate, parallelize, continuously operate, and continuously improve.
If this capability is eventually combined with robots, automated laboratories, supercomputing and advanced manufacturing, scientific research itself could become a production capability that can be scaled up.
By then, the real problem facing mankind may no longer be:
"Will AI replace scientists?"
Instead:
"When scientific discoveries can be replicated and accelerated on a large scale, what is the upper limit to the speed of technological progress of human civilization?"