Thesis Update Log Week 2

This week I sent out emails to recruit members for the thesis committee (my defense won’t be for months but my academic advisor wanted this done ASAP). Being so official and interacting with professors I’ve only interacted with personally a small number of times means that I spent a ludicrous amount of time and stress on this really simple task.

I also came up with the folling working title and abstract:

Understanding and Avoiding AI Failures: A Catastrophic Approach

In place of root cause analysis, we look over the history of AI catastrophes to find common themes between these failures to develop a guide as to what failures to expect based on the nature of a technology and means to mitigate the damage or a recommendation to unilaterally discontinue research in extreme cases. By focusing on thematic analysis instead of root cause, we can systematically identify when and where attention should be paid to safety of current generation AI systems.

The “Catastrophic Approach” part is a bit cheeky or dramatic, but also who doesn’t like fun titles? I’ll see what Dr. Yampolskiy has to say.

I should have gotten more done, but even for deadlined work I’m having trouble working this week. All I can do is take it easy this weekend and hope for better energy next week.