(The title was pilferedinspired from a comment by a work colleague, who agreed to be henceforth referred to as [Your Name], as ChatGPT offered this placeholder for their signature.)

Artificial Intelligence or, more accurately, Machine Learning is an amazing tool for sifting through large amounts of data and discovering insightful patterns. A task where a human operator would generally get bored and become sloppy — or simply die of old age in the process — can be very effectively performed by a machine, and a result returned, sometimes in a matter of seconds.

Rather than exhibit true intelligence, however, those systems only learn as much as is present in the data they are given. This is also what they regurgitate. It is no wonder that outputs from those algorithms replicate the biases present in their input data.

Much research work has gone into identifying and reducing biases in training data, or actively de-biasing responses, but the final decision of what to do with the result of an ML process is entirely in the hands of a human being operating it.

tl;dr: Rather than focusing solely on painstakingly fixing each ML system separately, we should also leverage generative AI chatbots to help train humans to recognise, and critically think, when dealing with any ML system, de-biased or not.

Continue reading

It is bad practice to use make install. Period.

Why? Because it installs files everywhere on your system—if you’re lucky, only in /usr/local—with no guaranteed way to cleanly remove them afterwards.

Yet, sometimes, there is no other option, for example if some software is not packaged for your Unix of choice and you don’t have time to do it yourself. There are some easy and rather straightforward ways around it, which I usually recommend to beginners.

It happened again today. So I recommended the use of /opt/PKG-VER as an installation prefix and stow(8) to make the software seamlessly available to the rest of the system. Nothing fancy or novel, but I thought I’d share the summary email in the hope it would help others.

Continue reading

I like my mailbox organised. And I like things to be automated. Fortunately, email systems support aliases for their users, so more than one email address reaches the same person. This allows for automatic filtering depending on which address the message was sent to.

What’s even better is that these systems can match a pattern to make generic aliases (e.g., user-REPLACEME@example.net for user user@example.net). This way, you can create valid email addresses on the fly, without having to tinker with anything (e.g., user-gascompany@example.net for the gas company to contact the user).

Now, dash (-) is not the most common character used for that purpose. The plus character (+) is more commonly seen. Notably, but not alone, GMail supports it. If you have an account there, try sending an email to YOURUSERNAME+test@gmail.com.

And this is where my problem is.  Once again, I was happily filling in a form requesting my email address, put in an address with a + in it, and got it rejected because it “contain[ed] invalid characters.” It really annoys me that some people who call themselves professionals in IT-related fields do not seem to be able to understand a standard properly, if they have been looking for it, at least…

Continue reading