AI can talk. But can it think, see, move, and make money? Here's where retail investors should look before Wall Street's next robot revolution.

Imagine spending $745 billion building the world's smartest computers, only to discover they still cannot reliably pick up a coffee cup without dropping it. ☕😂

Welcome to the next great AI investment debate.

The big question isn't whether artificial intelligence (AI) will change the world. It already is. The question is whether today's enormous spending on language-based AI will generate enough profits to justify the bill.

And here's an interesting twist: The next big AI winners may not be the companies teaching machines to talk, but those teaching machines to do things.

1. A 19th-century story that challenges Silicon Valley

In 1892, psychologist William James published Thought Before Language, discussing Melville Ballard, a deaf man who reportedly developed abstract thoughts and wondered about the universe before acquiring conventional language.

The lesson? Thinking doesn't necessarily begin with words. Humans also reason through visual images, physical experiences and spatial relationships.

Now consider large language models (LLMs). They learn patterns from enormous amounts of text and generate responses by predicting likely sequences of words.

Critics argue that speaking intelligently isn't necessarily the same as understanding reality.

Call it the Ballard challenge: Can AI reason about the world without relying entirely on language?

This isn't proof that modern AI cannot reason. Today's systems can process images, use tools and interact with simulations. But it raises an important question about whether simply scaling language models will eventually deliver diminishing returns.

For investors, the distinction is crucial. Don't confuse impressive demonstrations with profitable businesses.

2. The $745 billion spending spree: Follow the money!

Amazon, Alphabet, Microsoft and Meta are collectively expected to spend approximately $720–745 billion on 2026 infrastructure investment, according to TMT Finance's September analysis.

CapEx means capital expenditure: money spent on long-term assets such as data centres, servers and equipment.

But spending isn't earning.

Alphabet's Q2 2026 capital expenditure reached $44.9 billion, while quarterly free cash flow turned negative $5.9 billion. Yet Google Cloud revenue grew strongly.

That's the paradox: AI demand is real, but the return on investment still needs proving.

Watch four things:

• Revenue: Are customers actually paying for AI?
• Margins: Is AI improving profitability?
• Cash flow: Is spending producing sustainable cash?
• Productivity: Does every additional dollar of infrastructure generate better returns?

A data centre isn't a gold mine just because it contains expensive GPUs. 🤑

3. The next frontier: Physical AI

If language-based AI becomes increasingly commoditized, another opportunity could emerge in machines that interact with the physical world.

Think sensors, cameras, robotic arms, autonomous vehicles and warehouse automation.

A robot must see an object, calculate its position, estimate its weight, avoid obstacles and manipulate it safely.

That's a different challenge from writing a convincing paragraph.

Here are six US-listed companies worth researching:

My favourite research angle is the picks-and-shovels layer: companies supplying the sensors, vision systems and components that multiple robot manufacturers need.

NVIDIA deserves special attention too. Its Isaac platform provides simulation and development tools for autonomous robots. It isn't merely selling chips; it's building infrastructure for physical AI.

But remember: a promising technology doesn't automatically make its stock cheap.

4. Three ETFs, three different investment strategies

Don't want to marry one robot? Consider exchange-traded funds (ETFs), which hold baskets of securities.

My interpretation:

• ROBO: Interesting for diversified exposure to industrial components, sensors and automation specialists.
• BOTZ: More concentrated exposure to established names such as Intuitive Surgical, Fanuc and Keyence. It also holds NVIDIA, so don't assume it eliminates mega-cap exposure.
• ARKQ: The aggressive option for investors comfortable with active management and speculative autonomous technology bets.

My preference for a diversified physical-AI satellite would be to investigate ROBO first, then decide whether BOTZ or ARKQ adds something genuinely different.

All three carry thematic risk and relatively high fees. Check current holdings rather than relying on marketing labels.

5. The retail investor's blueprint: 85/15, not 15/85!

Robots are exciting. Your retirement portfolio shouldn't require a fire extinguisher. 🔥

For a long-term investor with a five-to-ten-year-plus horizon, consider this illustrative $10,000 allocation:

Choose VTI for the broad US stock market or VOO for the S&P 500. QQQ is an alternative for investors deliberately seeking greater technology concentration, not a defensive substitute.

Your practical checklist:

1. Build the broad-market core before buying thematic ETFs.
2. Use dollar-cost averaging (DCA): invest fixed amounts regularly instead of guessing the perfect entry.
3. Check overlapping holdings, especially NVIDIA and Tesla.
4. Rebalance annually. If a 15% satellite grows to 25%, consider trimming it.
5. Monitor revenue growth, margins, debt, free cash flow and valuation.
6. Size positions so you can survive a potential 40% thematic-sector drawdown without panic selling.
7. Reassess the thesis when customer adoption or business economics deteriorate.

6. Red team: What could go wrong?

The biggest risk isn't necessarily that AI fails.

It may succeed, but generate less shareholder profit than investors expect.

Language models could become cheaper, competition could squeeze margins, robotics adoption could take longer, and expensive thematic ETFs could deliver disappointing returns.

Physical AI might also benefit cloud providers, semiconductor companies and logistics operators more than standalone robot manufacturers.

The 10th-man perspective: If everyone is buying the future, make sure you're not overpaying for the press release.

My verdict: Own the broad market first. Research the physical-AI supply chain second. Buy individual winners only when business quality, cash flow and valuation justify the risk.

Why Wealth Builder matters

Investors don't suffer from a shortage of ideas. They suffer from too many ideas, contradictory headlines and insufficient time to investigate them. Wealth Builder, passive-income and investing newsletters help turn information overload into practical decisions: what deserves research, how to evaluate an investment, how much risk to take and when to rebalance.

Instead of chasing every AI sensation or panicking at every market correction, readers can build repeatable wealth-building habits and focus on long-term compounding. You don't need to predict which robot will become the next superstar. You need a process that helps you participate in opportunities without putting your financial future at unnecessary risk.

Want more thoughtful investing perspectives? Explore other like-minded newsletters here.​

Own the hands. 🤖✋📈

Notes & Sources

1. William James, Thought Before Language: A Deaf-Mute's Recollections, The Philosophical Review, 1892. Historical account of Melville Ballard.
2. TMT Finance, 2026 Hyperscaler CapEx Tops US$700bn – Analysis, September 23, 2026. Read report.
3. Reuters, AI Investment Boom Puts Big Tech's Free Cash Flow Under Pressure, July 22, 2026.
4. NVIDIA Developer, Isaac Robotics Platform.
5. Official fund information: ROBO Global, Global X BOTZ, ARKQ.
6. Company investor relations: NVIDIA, Cognex, Teradyne, Symbotic, Intuitive Surgical and Tesla.

Abbreviations: AI = Artificial Intelligence; LLM = Large Language Model; AGI = Artificial General Intelligence; CapEx = Capital Expenditure; ETF = Exchange-Traded Fund; DCA = Dollar-Cost Averaging; GPU = Graphics Processing Unit; FCF = Free Cash Flow; cobot = Collaborative Robot.

Investment allocations are educational examples, not personalized financial advice. Company forecasts, ETF holdings and expenses change. Verify current information before investing.