To The Moon

The AI trade is facing another reality check.

The Nasdaq Composite fell 2.86%, as concerns emerged around the financing requirements of major AI companies. Reports that OpenAI’s fundraising commitments fell short of its original target, alongside a negative Wall Street assessment of Anthropic’s financial position, have added to investor anxiety.

For much of the past year, the market has rewarded companies connected to AI expansion. Semiconductor manufacturers, cloud providers and data-centre operators have benefited from expectations of enormous future demand.

But investors are beginning to ask a more difficult question: how much capital can the AI industry continue absorbing before the financial returns become visible?

I think this is an important turning point for the sector.

1. AI demand and AI profitability are two different things

There is little doubt that AI is transforming how businesses operate. Companies are investing in computing capacity, advanced chips and data centres to support increasingly sophisticated AI applications.

However, demand for AI technology does not automatically mean every company involved will make attractive returns.

Building infrastructure requires enormous upfront investment, while revenue from AI products and services may take years to mature.

If funding becomes more expensive or investors become less willing to finance aggressive expansion, some companies could face pressure to reconsider their spending plans.

This is where the distinction between AI demand and AI profitability becomes important.

2. Semiconductor companies could face two different outcomes

For companies such as $NVIDIA(NVDA)$ the biggest question is whether customers will continue spending heavily on advanced computing infrastructure.

NVIDIA benefits from strong demand for its GPUs, but its future growth is also connected to the investment plans of major cloud providers and AI developers.

If those customers slow their expansion, the market may begin reassessing semiconductor valuations even if demand remains structurally strong.

$Intel(INTC)$  presents a different situation. Its turnaround depends on manufacturing execution, competitive products and improving financial performance rather than simply riding the AI infrastructure spending cycle.

I would be careful about treating all semiconductor stocks as one investment opportunity. Their business models, competitive positions and financial risks are very different.

3. The biggest risk could be the companies funding the infrastructure

$Microsoft(MSFT)$ , $Alphabet(GOOGL)$ and $Meta Platforms, Inc.(META)$ are committing substantial capital to AI infrastructure.

These companies have established businesses and significant cash-generating capabilities, which gives them more flexibility than privately funded AI startups.

Nevertheless, increasing capital expenditure can eventually pressure free cash flow if the revenue generated does not keep pace.

I would pay particular attention to the relationship between capital expenditure growth, cloud revenue growth and operating cash flow.

The market may tolerate substantial spending while returns are improving. It could become much less forgiving if spending continues accelerating without corresponding financial benefits.

4. Could this create opportunities for stronger companies?

A correction does not necessarily mean the entire AI investment cycle is over.

In fact, a period of more disciplined spending could benefit companies with strong balance sheets, established customers and differentiated technology.

I would focus on businesses that can demonstrate:

• Sustainable revenue growth rather than relying on future projections.

• Strong free cash flow and manageable capital requirements.

• Competitive advantages that are difficult to replicate.

• Clear evidence that AI investment is translating into customer demand and earnings.

The next phase may reward companies that monetise AI rather than simply spend money building it.

5. My investment strategy

I would not rush to buy every AI stock after a correction, but I would not abandon the sector either.

My preference is to separate infrastructure providers, cloud platforms and AI application businesses rather than assuming they all carry the same risks.

For semiconductor exposure, I would continue monitoring NVIDIA’s demand outlook and the spending plans of its largest customers.

For broader technology exposure, I would also watch whether established cloud businesses can maintain growth while controlling infrastructure costs.

I would keep some cash available and use market weakness to identify companies where valuations have become more reasonable relative to their earnings potential.

My conclusion

I don’t think the AI opportunity has disappeared. However, the market may be moving away from rewarding AI spending alone.

The distinction between companies that can afford to invest, companies that can build useful technology and companies that can actually generate attractive returns from AI will become increasingly important.

The AI revolution could continue while some AI stocks experience substantial corrections. Both things can be true at the same time.

For discussion purposes only.