Dear Reader,

Is it just me, or is this so-called "AI bubble" so complicated and odd that it's hard to even come up with the words to speak about it meaningfully? Everyone has a lot to say (including me), but the words hover around the edges, never really capturing what is actually going on and what it might mean for our industry and our civilization. I don't want to make it sound ominous – one legitimate viewpoint is that it might not mean much of anything at all, but the hard part is coming up with the arguments to justify whatever position you take.

With that in mind, I was interested in a post at Fast Company by Faisal Hoque [1], who argues that what we are calling the AI bubble isn't really one bubble but is actually three bubbles:

• an asset bubble (or speculative bubble) caused by investors bidding up the price of AI-related companies to astronomical levels;
• an infrastructure bubble, with companies pouring billions of dollars into AI infrastructure that could very possibly exceed any conceivable demand; and
• a hype bubble, with futurists, marketing execs, and pop science writers dialing up the optimism for a breathtaking AI future that might not ever happen.

Hoque's analysis is primarily geared for investors, which is not so relevant to us in the techie space, but his insight about breaking down the problem into three separate problems is useful for anyone who is trying to make sense of what is going on right now in the industry.

According to Hoque, the three AI bubbles will inevitably burst, but that doesn't mean meaningful work isn't getting done. He brings up the example of the dot-com bubble, which scaled to a level where it practically swallowed the whole planet, then suddenly exploded, leaving a littered field of bankruptcies, but while the public was focused on lofty casualties including Webvan and Pets.com, companies such as Amazon, Google, and PayPal were working a little outside of the limelight to perfect technologies that would one day lead to transformation.

But enough with the new. Let me take a moment to salute something that isn't so new: us. This month is our 300th issue, which means we've been putting out Linux Magazine for 25 years. A quarter century is a long time for any magazine, but for a computer magazine, we are in rare company. I just want to give a shout-out to our great team, our great readers, and our great topic: Linux and the open source community. Linux keeps evolving, which means that we get to keep evolving with it, and we never run out of exciting new things to write about.

I've always thought the key to putting out a good magazine is not to impose too many unnecessary expectations. Lean into good, eloquent tech writing, and let the subject reveal itself. As long as Linux keeps changing, our authors and our readers will change with it, and we will chronicle it all as it unfolds.

And speaking of change, I would like to offer a note of thanks to longtime author Bruce Byfield, who is making his final appearance as a regular columnist in this month's issue. Bruce has contributed to Linux Magazine as a blogger, print author, special consultant, and all-around expert since 2009. Farewell, Bruce! We will surely miss your insights, your deep knowledge of the Linux community, and your abundant patience. Best of luck with your next steps!

Joe Casad, Editor in Chief