In Colorado, HumanCode startup introduced BabyGlimpse, a $259 test that uses DNA from a couple to predict what their future child might look like and how they even might behave, including their preferred kinds of snacks. The direct-to-consumer genetic testing relies on the combination of each partner’s DNA. HumanCode uses “machine learning models trained on a mix of publicly available genomes and proprietary data to come up with what’s called polygenic risk scores for each trait,” according to Wired. But the genes don’t guarantee the predictions since other factors like environmental factors of where you live, what you eat, etc. all impact those things even more than DNA.
Some critics, however, say this information is not always helpful and could in fact affect how parents think about their baby for life. For example, if parents know their baby will be more likely to have lower math ability or weaker than average bones, parents may change the way they treat their child based on the genetic information they received, like not pushing them as hard in math or not putting them in sports for fear of bone injuries.