Hmmm - he seemed a bit Panglossian to me - everything can be improved so easily!
I really wonder how reproducible the research that he quotes is.. As I understand it, the reading of epigenetic information from DNA is far more primitive at this time as compared with the reading of DNA itself. For anyone wanting some details of epigenetics, this is the book:
https://www.amazon.co.uk/dp/B00JFVOLZK/ref=dp-kindle-redirect?_encoding=UTF8&btkr=1
Basically DNA gets decorated with methyl groups and acetyl groups that modify its expression - i.e. how much of its protein product is produced. These are generally stripped off in the germ cells, but not completely - so certain characteristics can get passed on that way. For examples parents that survived famine can give rise to children who are likewise more resistant to famine, but tend to grow fat in times of plenty.
I don't know why they both talk of a computer hard drive as an example of information that cannot be modified - because we modify the contents of our computer hard drives all the time!
David