Biochemist Laurence Moran has had a tough go of it recently. After engaging in a somewhat tense exchange with the AI bot ChatGPT, he has accused it of fabricating and misquoting information related to “junk DNA”. Does this AI bot “lie”, as Moran believes? Let’s investigate.
At the center of the conversation appears to be Moran’s many entrenched views on junk DNA. From the start, Moran has argued evolutionary thinking has a major say in explanations of junk DNA, suggesting that “nothing in all of biology makes sense if you don’t know about evolution.” Philosopher of biology Paul Nelson countered that Moran’s stance on junk DNA is unique and fails to represent todays broader scientific understanding of the concept.
Nelson notes that Moran has dug his heels in on his assertion that there is a widespread misconception of biologists failing to give due credit to the regulatory function of non-coding DNA sequences. Moran himself even released a book with the title “90% of Your Genome Is Junk” which backs this idea up.
However, results from the ENCODE project contradict this opinion, leading Casey Luskin to report for Evolution News in 2013 that “at least 80% of the human genome is now known to be functional.” On the contrary, Moran’s view that most of DNA is useless is rooted in a principle called genetic load which claims that if close to all of the genome is functional each newborn human baby would be inundated with too many mutations from their parents.
A post on Uncommon Descent challenges the idea of genetic load however, stating that even if the genome has been designed to achieve a certain level of functionality, incorporating into the design some redundancy and flexibility could easily account for the mutations. In other words, it seems unnecessary to assume that 90% of the genome must be deemed useless.
This brings us to the crux of the discussion: What do we do with AI chatbots like ChatGPT? It is clear that its responses included fabricated quotes and misquotes about junk DNA, but it’s likely this was due to revealing a summary majority view from online sources. User beware: AI chatbots are not a substitute for genuine, original research put into understanding an issue.
A major factor in this conversation about AI bots is their sometimes limited parameters and the resistance to critically explore a certain topic. As Moran alludes to in his post, ChatGPT is quite up to date with world events until 2021 – any event mentioned beyond that date is unavailable for this AI bot to address. As a professor, this type of “shortcutting” seems like a dangerous path for students to take, as it prevents them from developing and honing critical thinking skills and thoughtful writing.
Moran’s back and forth with ChatGPT offers an important lesson in the world of Artificial Intelligence. Don’t be tempted to seek out “quick and easy” answers. Real knowledge comes from studying the facts and finding the truth within them. Only by investing time and effort into your studies can you critically answer the questions of our times – and surely lots of time, effort and research went into understanding junk DNA.
