From the perspective of @BioProtocol viewers, I find this article really opens up a very different perspective on science. For the past 400 years, science has operated in a linear loop – observation, hypothesis, experiment, publication. Now, @BioProtocol talks about “Agentic science” – where BioAgents can read and understand the literature, create hypotheses, design experiments, send them to automated labs, and then feed the results back to the network. The interesting thing is that each BioAgent only does one thing well, but when connected together, they form a collective cognitive system, both parallel and continuous. This completely changes the bottleneck: from lack of funding, lack of publication opportunities → to speed of data collection and modeling capabilities. I have a feeling that this is no longer a small improvement, but a fundamental upgrade to the scientific method since the Enlightenment. If realized, it would be a new “knowledge engine,” enabling discovery at network speeds, more transparently and decentralized than ever before.