^{here, you can see "in vivo" is not the end of the process}

The potential for creating diseases is there but is probably the most costly and inefficient route at the moment. The use at the moment of "in vivo" is not life integration but creation of circuits. Bear in mind is a different use case, as this is about hardware not about creating functional proteins for the bacteria.

Think of it as encrypted biological information for the bacteria. It means nothing the them, the peptide could look like biological junk but it can compute. I made a post on protein folding design. That's one of the routes and is getting a lot easier.

For instance, in the case of data storage. You don't need the bacteria to be alive. DNA is relatively resistant as shown by retrieved mamoths DNA if on cold, dry places.

The main use currently is "DNA" chips.

You can't store them properly in a bacteria so to speak as the DNA degrades and introduces errors, so as part of the micro arrays you hibridize and conserve in sillico. One of the current apporaches is, you use the bacteria to produce the parts and ensamble.

Manipulation of DNA in vivo is one of the main ethical concerns, but is quite far away it seems.

Glad you passed by :)

^{DNA nanotechnology self assembly}