A collection of XKCD’s best works (in my opinion) on Artificial Intelligence and related topics.
Starting with my favorite for discussing sanitizing inputs and naming conventions. “Little Bobby Tables” is a classic.
The Three Laws of Robotics
There is on other way than Asimov’s way. Remains true for robotics and ton of other stuff. One of my favorite readings to assign to students is “The Last Question” short story by Asimov. There isn’t a better discussion of the concept of Entropy anywhere.
Yes – Terminators are bad. There are much more bad stuff that can happen before Terminators become a feasible prospect. Risks of AI should focus there rather than on sentient robot beings.
Explainability of Machine Learning remains an elusive goal for the most part. Without that we are just grasping in the dark and are likely to rely on ex-post pattern fitting to explain our results. Not ideal.
Computers vs Humans
Computers can do a lot of things. Computers also absolutely suck at a lot of things. AI will automate a lot of human labor but there remains a lot of aspects of human experience that don’t seem to lend themselves to computerization.
Terminators will save us from us as we are our worst enemy.
For Emacs lovers (including me). I have spent obscene amount of hours searching for a building the perfect dot-emacs file. Hours well spent.
Particularly relevant in the era of Cubic Fit models of Covid-19 data.
Only XKCD can explain an entire class worth of material in one panel.
My Data Visualization students take note.
P-Hacking explainer. I have done a few sub-group analyses in my day too.
This is a living document. I will add more material to it as I find them.