# general
I love reading your articles on TNS. I've seen several talks recently that cite the same research and I'm noticing some interesting patterns that are repeated here, which I understand are likely due to Ben producing the patterns in the original talk. We need to think more about value streams rather than optimizing typing speed. The best finding listed was the increase in job satisfaction, which is excellent. Expecting code and docs acceleration to impact the whole value stream will be a mistake many folks make. The other thing I've noticed with the Harvard/BCG research is that folks are only sharing half of the findings, i.e. the "inside the jagged frontier" half. This hides the important insights into what happens with the tasks outside the frontier. Crucially, the top performers for tasks on the inside were the worst performers for tasks outside. It's probably the most crucial finding in the research. It would great to spark a discussion on this as the question is really... how do we identify whether a task is inside or outside the jagged frontier?
It was a great talk
Great points well put. Thank-you. Had I had more than 25 minutes, I would have had more of an opportunity to go into greater depth. But, given the talk was for developers I wanted to move on to cover studies specifically designed to analyse the impact on those roles rather than the generic "knowledge workers" in the BCG study. Perhaps in future iterations I can cover this in more detail. I'm also on the hunt for new research - especially those with larger sample sizes and similar control groups. If you find any, do shout them out so we can all reap the benefits of those insights.
BTW, I would put forth the idea that if you think about the jagged frontier as being defined by what the McKinsey study suggested are the four strengths of AI (for developer tasks) - this may help identify the right tasks more accurately.