AI investment dashboard: Understand the layers before entering the market

Recently helped a client look at their AI positioning. The most annoying thing is that everyone has an "AI company" on their lips. I opened this global market dashboard, and my judgment is: it depends. It's suitable for creating industry maps and hierarchical categorization, but not for making direct buy/sell decisions. It's free, requires no registration, and you can see multiple markets and themes on one screen, which is very user-friendly; but if you want a button that tells you exactly what to buy, it will disappoint you.

When I actually clicked in, the first thing I saw was dashboard logic like "every market on one screen," where quotes and themes are compressed onto the same page. Scrolling down, AI investment is broken into several layers: underlying compute, data centers & energy, models, applications, and companies using AI to transform traditional businesses. The page also breaks down 246 public and private AI companies according to the "five-layer cake" model. More accurately, AI is a chain. The core competitive moat differs at each layer of the chain: chips depend on process nodes and ecosystems, data centers on power and licenses, and applications on distribution and cash flow.

What tripped me up in the interface was the classification criteria. The same company gets repeatedly categorized under different themes. For example, a cloud provider might appear under AI infrastructure, or under model platforms, or get stuffed into applications because they sell enterprise tools. This is unfriendly for beginners. Let me explain some terms: an ETF is an Exchange-Traded Fund, allowing you to buy a basket of companies just like buying stocks; overseas there's iShares' ARTY, where one trade buys you a basket of AI companies. It's convenient, but it's a black box—you don't know who contributed to the gains or losses within the basket.

The surprise lies in the comparison. Putting the dashboard's layering, Britannica's, and SSGA's perspectives together, the contradiction is clear: AI is too important, everyone wants to invest; the boundaries are too wide, any classification carries subjectivity. It looks more like a strategic sandbox than a trading terminal. It can turn "I'm bullish on AI" into "I'm bullish on which layer, and does that layer have verifiable barriers to entry?" But without an industry framework, just staring at the red and green on the screen makes it easy to mistake noise for signal.

Suitable for mapping, not for trading buttons. Recommended for investment research, strategy, and industry scanning people; also suitable for beginners to establish a framework first; not suitable for short-term traders. In my testing, the most practical action is to drag a company back into the chain and see whether it sells compute, software, data, or is just telling an AI story.

---

📌 This article is compiled from Hacker News, original text: https://gmdmarkets.com/how-to-invest-in-ai

Copyright belongs to the original author. This article is a compilation and independent analysis based on public reports.