It maybe just the same, that is, if the respondent likes talking to machines. Sometimes, people get bored that they just answer for the heck of it (and answer so that it will get over and done with), so the answers may not be a hundred percent true for them. Only, the advantage of it is that it is faster to come up with the data from the survey.
When a person does the interview, open-ended questions become more detailed as a person can probe and clarify for answers and there is a chance for one to change their answers as interviewers can erase and edit the data. But there is also a possibility for the interviewer to influence the answer of the respondent.