The answer to this question can either be extremely simple, or quite complicated, depending on where we wish to go with it. However, let us explore what entails to think "quantitatively" and "qualitatively" to establish the parameters within which we want to analyze the thinking processes of man versus machine.
When we use the term "qualitative" thinking, the most raw and pure definition that we can give to the concept of "quality" (straight from the dictionary, that is) is that it is a peculiarity of uniqueness that establishes character. To analyze or extrapolate qualities, many variables must be taken into consideration: how is quality manifested; what are the traits of quality; within which parameters is quality going to be analyzed; does the concept of quality change from one scenario to another?
Qualitative research involves empirical methods that include ethnographic research, phenomenology, grounded theory, foundational research, and ethical inquiry, among other interventions.
All this being said, all the variables and parameters that entail qualitative research must be entered into a computer in order for it to process and analyze qualitative information effectively and accurately. Can computers be that efficient, or is human agency still a necessary inclusion to complete these types of analysis? Research shows that, when it comes to qualitative research, computers alone can only tell us so much. The resulting data could be further applied and analyzed, but only through human intervention for further analysis. This is because only humans can think deductively to make conclusions of the information, and think inductively to formulate further inquiry.
Also, computers follow a binary linguistic input that follows an "if/then" format. Again, this input can only be possible through human agency and the only conclusive evidence that computer language can produce is entirely dependant on the information that was put in. There is no room for last minute variables, or sudden new information. The if/then format is quite strict within a computer interaction.
Now, quantitative research is different because it is based entirely on data that MUST be analyzed from an if/then method. In quantitative research there is little to no human agency because data needs to be analyzed at its rawest should be given the most unbiased treatment. As we enter pre and post data, the computer is able to clearly read patterns, determine trends, establish correlations, and make predictions of the same formula. Also, quantitative research can analyze the same data under different variables as long as the data remains the same. That is when computers completely take over humans and surpass the possibilities of "thinking" processes.
Therefore, computers are best used for quantitative research and human agency is most effective when using qualitative research.