In one of his non-comic Dilbert books, he suggested that gravity could be explained by everything constantly doubling in size at once, which only works near Earth, but Adams doesn't care. He also said the theory of evolution would be disproved in our lifetimes.
His post about ID was criticized by biologists, most notably PZ Myers, a developmental biologist who runs the
Pharyngula blog, who responded
here, then
here,
here, and
here.
When people in Scott Adams' comments section pointed out that he should read books on evolution, such as those by Stephen Jay Gould or Richard Dawkins, he refused on the grounds that because Gould and Dawkins got
paid to write those books, they were biased and couldn't be trusted to talk about biology. (Given that Scott Adams has been paid to write books on things like management, I sense hypocrisy.) Adams apparently claims to find both sides bankrupt, while refusing to learn about evolution and parroting creationist arguments. He presents both evolution and ID as being on equal scientific footing, which is nowhere near the case, and when he was called on it, he claims he is being misinterpreted and/or that he is a humorist and shouldn't be taken seriously. He claims that evolutionists (particularly the quite popular
Talkorigins Archive) misrepresent and straw-man the ID arguments, while never actually giving any examples of this happening.
Scott Adams is a post-modernist, claiming that science is no good at finding the answers, or at least that every other method is as good as science.
Adams is lost in his own distrust of expertise and possible bias. If only there was a way to learn things that was very good at identifying and elminating bias and error in all their forms (error, fraud, unconscious cheating, self-deception, etc.) What we need is a process for gathering and analysing knowledge based upon standards of logic and ethics that is self correcting over time, a process where all the data, rationale, and methodology is required to be available to anyone who's interested in learning or evaluating what others have done and concluded. Maybe this process couldn't answer all of life's questions, but at least it should provide immense benefit to all mankind. I wonder what such a process would be like...
Oh, wait! There's science. Nevermind.
He should stick to the office jokes.