The AI transition is underway, and despite the growing enthusiasm for these tools, a recent study by Harvard and Boston Consulting Group revealed that they are less effective when applied to tasks beyond their frontier of competence. In many ways, using AI is analogous to having a toolkit: some jobs need a wrench while others require pliers. My main value-add as a researcher and content writer are domain expertise and critical thinking. Despite this, I often find myself conducting tasks that fail to leverage these skills. This left me wondering if I am spending too much time using a wrench when all I needed for the job was a set of pliers?

At mallowstreet, we often speak with industry experts on current developments or ongoing research projects. This typically follows an iterative sequence of:

Upon reflection, I realised that this process is rife with inefficiencies. To fix this I decided to use SOFI, mallowstreet’s AI tool, to see if it could improve my output by handing over many of the more repetitive and time-consuming tasks. 

Typically, we rely on a third-party provider for transcription services. However, this imposes both a monetary cost and time burden. On top of this, the quality of the transcriptions is not always up to par. Given this hurdle I was more than pleased when SOFI solved for each of these problems – yet this barely scratches the surface of the range of benefits it delivered.

AI might not yet be capable of outperforming humans at critical thinking or analysis, but it is highly adept at extracting themes from text. Using it in this way meant that, rather than organising key words and quotes on Excel or a tool like NVivo, I could instead get an overview from SOFI at the click of a button. I was especially surprised to find that working with SOFI amplified my output because it gave me a second pair of eyes to surface more findings.

Next, I needed to identify quotes for the report. This is normally done by running searches through each individual transcript, which can be a slow and repetitive process. SOFI provided a much better way of doing this thanks to its Checklist function, which was flexible and surprisingly intuitive. For example, when I wanted to identify specific quotes on a topic, I simply asked SOFI to show me an exact example from the conversation. This allowed me to compare quotes across each interview simultaneously to find best one for my article or report. I could also ask SOFI more general questions to get a summary about how the speakers discussed a given topic.

As humans we naturally gravitate towards repetition and procedure. Having served in the military I am all too aware about the value this brings to one’s personal and professional life. However, my experiences using SOFI taught me that it is equally important to move out of our comfort zones and avoid falling into stagnation.

 What has been your experience using AI so far? Tell us in the comments below.

     Or alternatively, below are some additional readings about SOFI AI: