One of my big beefs with ML/AL is that these tools can be used to wrap bad ideas in what I will call “Machine legitimacy”. Which is another way of saying that there are many cases where these models are built up around a bunch of unrealistic assumptions, or trained on data that is not actually generalizable to the applied situation but will still spit out a value. That value becomes the truth because it came from some automated process. People cant critically interrogate it because the bad assumptions are hidden behind automation.

I often have the opposite experience when looking for technical documentation about programming libraries. For example I will be dealing with a particular bug and will google the library name plus some descriptive terms related to the bug, and I get back general information about the library. In those cases, it seems google often ignores the supplemental information and focuses only on the library name as If I were looking for general information.

What is worse is that the top results are always blog-spam companies that just seem to be copying the documentation pages of whatever language or library I was looking at.

Former CEO of the river poisoning company says there is no way to meet our river poison reduction goals, so we might as well build bigger river poisoning machines because they might help us figure out how to stop poisoning the river. /s

I feel like there was a time when the tech folks in silicon valley had a lot of credibility, and we are now living in a period where most of the world sees them as a joke but that fact has not yet entered into the culture of silicon valley.

Used get my haircut at one of those “no appointment needed” haircut chains. Then they got an app, and every time I went it was “Why aren’t you using the app? You need to use the app. Next time use the app. Download the app on your phone. It’s gonna be an hour wait because you didn’t use the app.”

Now I just go to a local place.

Unhelpful Linux User Archetypes:

The Configurator: All problems are configuration problems. The fact that a user has a problem means they configured their machine incorrectly. All help requests are an opportunity to lecture others about configuration files.

The lumberjack: Insists on logs no matter how simple or basic the question. “How do I get the working directory in the terminal?” -Sorry, I can’t help you unless you post your log. “What does the -r flag do?” -You need to post a log for me to answer that question. “Is there a way to make this service start at boot?” -We have no way of knowing unless you post your log. When a user posts their log, the lumberjack’s work is done. No need to reply to the thread any further.

The Anacdata Troubleshooter: Failed to develop a theory of mind during childhood. Thinks their machine is representative of all machines. If they don’t have an issue, the user is lying about the issue.

The Jargon Master: Uses as much jargon as possible in forum posts. If a user doesn’t know each and every term, that’s on them. If you did not commit to mastering every aspect of a piece of software before asking for help, were you even trying to solve the problem?

The Hobby Horse Jockey: All problems are caused by whatever thing the contributor does not like. Graphics driver issue? Snaps. Computer won’t post? Obviously, Snaps. Machine getting too hot? Snaps. Command ‘flatpack’ not found? Oh you better believe snaps did that.

The Pedantfile: Gets mad because everyone asks their questions the wrong way. Writes a message letting the user know they asked their question wrong. Message usually appears within a minute or two of someone providing a solution to the user.