Bubbles

It’s hard to find any tech news lately that doesn’t talk about the “AI bubble” with commentators trying to guess when it might pop. But what does all this actually mean? I’m definitely not an expert in the financial world, but here’s what I’ve managed to make sense of through what I’ve been reading and watching.

There are different kinds of economic bubbles, but there are common threads. They occur in situations where investors buy into an asset at a rate that exceeds what the asset is worth. This is usually through speculation – where investors are banking on seeing a large return on their investment in a short period of time. Often there is a period of euphoria in which companies and investors see the asset as “the next big thing.” Eventually, as excitement about that particular asset begins to wane, or the kind of growth or profits from that asset don’t materialize, prices fall and ultimately crash as investors pull out their money and move on to something else.

Economic bubbles have occurred at various times in our history. In 1634, Tulip Mania caused tulip prices to soar before abruptly crashing in 1637. Railway mania occurred in the UK with speculators investing heavily in railroad building before suddenly crashing in the 1840s. The bubble many compare the current state of AI investment to is the Dot-com bubble, which grew in the late 1990s before bursting in early 2000.

The Dot-com bubble occurred as people started to use the World Wide Web and investment began to flow to a large number of tech startups. When many of those companies failed to be profitable, investors backed out and stocks fell, causing a tech-heavy stock market to plummet. This caused a huge number of tech companies to shutter, and for many others to lose value. Some of the larger tech companies we know today (Amazon, Google, Nvidia, to name a few) are survivors of the Dot-com bubble.

There are several similarities that commentators are seeing between AI now and economic bubbles of the past. There is rapid growth in building infrastructure, with most major companies investing in the massive data centres and power generation needed to support the technology. This has resulted in new AI companies (like Open AI) as well as tech stalwarts (like Microsoft, Meta and Google) to borrow billions of dollars from investors and governments to bring their data centres to fruition. Right now, Nvidia is – for all intents and purposes – the sole supplier of the processing hardware used in these data centres and as a result recently became the world’s first 5 trillion-dollar company. There is also a lot of AI-hype in the industry – something akin to investor euphoria.

The thing about bubbles, though, is that they aren’t definitively considered bubbles until they burst. With AI companies accounting for the entirety of the GDP growth in the US, and a lot of speculative investing in those companies, worry that AI is experiencing a bubble appears pretty valid. But AI doesn’t need to be a bubble to start causing economic trouble right now – and it already is.

Most electronic devices need a component called RAM (Random Access Memory). It’s a critical part of many consumer devices like PCs and smartphones and is often used in more than one way in a single device – without it, these devices won’t work. AI needs massive amounts of RAM. In the rush to build server-farms, AI companies have been buying massive quantities of RAM, and the few companies that manufacture it (there are only three: SK Hynix, Micron and Samsung) have started to shift away from making it for consumer electronics to concentrate on their more lucrative AI customers. In fact, Micron has completely shuttered their consumer memory division. This has resulted in a RAM shortage that is already driving up the price of consumer electronics. A memory upgrade for your PC that cost $120 last year now costs $600.

What this means for normal people like us is that the already expensive tech we use is going to become even more expensive for the foreseeable future and puts us in a catch-22; we can try and limp along with our older devices but can wind up in a situation where a device outside of its support life becomes hazardous to use (i.e.: susceptible to being hacked etc.)

Unfortunately, I have zero idea how any of this is going to wind up. Economic bubbles bursting can have a variety of effects, from only affecting the players involved in the speculation that created the bubble in the first place, to farther reaching consequences like mass unemployment and recessions. Regardless of there being a bubble or not, things aren’t great for consumers right now while tech companies chase their AI pipe dreams.

Learn more

AI is skyrocketing the price of RAM. Computers, phones and tablets could be next 2025. Rukhsar Ali (CBC) Last accessed 2026/01/13

Economic Bubble (Wikipedia) Last accessed 2026/01/13

Economic Bubble - The Decision Lab (The Decision Lab) Last accessed 2026/01/13

How Did RAM Get So Expensive? And How it’s Going to Get Worse… 2025. Linus Tech Tips (YouTube) Last accessed 2026/01/13

RAM: WTF? 2025. Gamers Nexus (YouTube) Last accessed 2026/01/13

WTF Just Happened? | The Corrupt Memory Industry & Micron 2025. Gamers Nexus (YouTube) Last accessed 2026/01/13

Correction log

Nothing here yet.