The speaker argues that many people misunderstand the scientific process, assuming that peer review is equivalent to independent replication. He explains that true independent replication involves downloading the code and running it on one's own machine.