The numbers this week have been difficult to ignore.
Major tech stocks fell sharply after the new tariff announcements. Apple, NVIDIA, and a long list of hardware companies saw significant drops in a matter of days. Social media is full of predictions. Some people are calling it the beginning of a new era. Some are calling it an overreaction. Most people seem to be oscillating between both in the same afternoon.
I want to try to think about this more carefully, because I think the immediate reaction to events like this is almost always too loud and too fast to be useful.
What Is Actually Happening
The tariffs being discussed target goods manufactured in specific countries, with electronics and semiconductors among the most affected categories. For companies like Apple, whose hardware supply chain runs heavily through Asia, the implications are real. Manufacturing costs go up. Margins get squeezed. Either prices rise for consumers or profits fall for shareholders.
For the software and AI side of the industry, the story is different and worth separating out. Software does not have the same supply chain exposure that hardware does. A language model running on a server does not have a tariff attached to it. A SaaS subscription does not get held at customs.
So the first thing worth saying is: this hits hardware harder than software, and physical goods harder than digital ones.
The Semiconductor Question
The more complicated part is semiconductors. AI runs on chips. The best chips in the world are designed by a small number of companies and manufactured by an even smaller number of fabs, most of which are in Taiwan and South Korea.
Anything that disrupts the economics of that supply chain has downstream effects on AI infrastructure costs. Data centre expansion becomes more expensive. The cost of training large models goes up. Cloud providers face margin pressure.
This does not stop AI development. But it does change the economics of it, and it may slow the pace of certain kinds of infrastructure investment in the short term.
What I Think People Are Missing
The reaction I keep seeing is either "everything is fine, markets overreact" or "this is the end of the tech boom." Both feel like they are reaching for a simple story when the actual situation is more complicated.
What I think is genuinely worth paying attention to is what this reveals about how dependent the tech industry has become on a globally integrated supply chain that most people never thought about until it started creating problems. The assumption for years was that globalisation was a permanent condition. These events are a reminder that assumptions about the operating environment can change faster than companies can restructure.
The companies that come through this best will be the ones that either had more diversified manufacturing already, or can move faster to adapt than their competitors. The ones that will struggle are the ones whose entire cost model was built around a world that may no longer exist.
The Software Layer Is More Insulated Than People Think
If you work in software, in AI development, in data, or in digital products, the direct impact of tariffs is much smaller than the headlines suggest. Your work does not sit in a shipping container. The tools you use are mostly cloud-based and globally distributed. The skills you are developing are not subject to import duties.
That is not a reason to ignore what is happening. But it is a reason not to catastrophise if your day job is on the software side of tech.
My Conservative Take
I do not think this is a temporary blip that will be forgotten in three months. Supply chain assumptions that took decades to build are being revised in real time, and that kind of recalibration tends to take longer than initial timelines suggest.
I also do not think it ends the current wave of AI development. The demand for what AI can do is not going away because manufacturing costs have changed. What it might do is accelerate the shift toward software-defined everything, because software does not have a tariff problem.
The most honest thing I can say is: we are in the middle of this, not at the end. Anyone who tells you they know exactly how it resolves is guessing. What we can do is pay attention, think carefully, and avoid making permanent decisions based on a week of market movement.
That is the conservative position. And in situations like this, I think it is usually the right one.