In response to the observation that people with poor vision would have an easier time filling a digital ballot that can be augmented, and that, in general, computers would make mistakes less likely to happen, the lecturer observes that there might as well be people not used to technology that could even choose not to vote because of that fact. Next the lecturer addresses the point being made regarding the mistakes that counters make when counting the votes, and recognizes that it is something that happens, but also, that if a computer were to make a similar mistake, since they're made by humans, and human mistakes could show up in the way they are programmed, there could even be thousands of votes lost and no possibility of recounting them. And finally, he says that the systems used in banks and communication were improved over a long period of time, and that because elections happen so much less frequently, there wouldn't be enough time to develop confidence that computers can be fully trusted for counting votes.