AI Overlords

There was an interesting confluence of news this week. Three separate stories taken together paint a broader picture that we may be nearing the inflection point where AI begins displacing more people than the new opportunities it creates. That or we just become disposable meat sacks grinding out metrics to satisfy the production quotas of our machine learning bosses.

First up is a Huawei developer scolded by a linux kernel contributor for grabbing KPI. Essentially, the huawei dev was submitting trivial patches to fix typos that are usually fixed by the kernel's housekeeping script, but the huawei dev wasn't submitting them as such. The takeaway here is more jobs are subject to KPI, but HR might be asleep at the wheel in determining the effectiveness of the KPI. If all management wants is some graph that says "our people are working" without stopping to inspect the quality of the work, over time the effort they're expending may drift from alignment with their mission. When peoples' jobs are focused on generating KPI, they'll find a way to game it.

Next we find more context around KPI at amazon. Amazon is known for relentlessly squeezing maximum productivity from their disposable meat sacks. So much so that there have been some recent articles expressing concern over their churn rate, which can exceed 100% for low-end positions. With turnover that high, and a company that big, some wonder whether (or when) amazon might hit the end of the labor pool, and what happens then? The optimist might say that'd be a great opportunity for amazon to improve labor conditions, but amazon has long been supplementing with AI and robotics, and they're probably more interested in "upgrading" their masses of worn-out meat sacks than improving conditions for them.

And finally, we have the introduction of GitHub Copilot. To be sure, this is early days for replacing higher functions with AI, but it drives home the point that we're headed in one direction across every industry. Big data feeding AI to figure out how to do our jobs for us, with KPI squeezing us for all we're worth until then. The last generation outsourced for efficiency. The next generation might be displaced.