Cheap AI Is Opening the Next Technology Trade

The KraneShares China Internet ETF has gained more than 20% since its June 25 low, roughly twice the Hang Seng Index’s advance, while the Philadelphia Semiconductor Index has fallen about 11% over the same period.

The move does not signal an investor retreat from semiconductors. It is opening a second front in the AI trade as cheaper deployment improves the economics of companies positioned to turn intelligence into cloud usage, advertising, commerce and software revenue.

Silicon Data’s blended inference-price index has fallen 36% from its May peak even as customers move toward more capable reasoning models and AI agents.

Lower deployment costs are giving platforms more room to put AI across existing products without requiring every interaction to carry a premium price.

The AI trade is widening from the companies supplying compute toward businesses positioned to use it commercially.

Internet Stocks Are Capturing the Next AI Rotation

Alibaba’s cloud business is providing some of the clearest earnings evidence behind the rotation.

Cloud Intelligence external revenue grew 40% in the March quarter, while AI-related product revenue recorded triple-digit growth for an eleventh consecutive quarter and reached 30% of external cloud revenue.

Tencent and other Chinese internet platforms are also drawing more investor attention as cloud, advertising, commerce and software give AI spending several routes into existing revenue streams.

Goldman Sachs has pointed to stronger AI usage, higher capital expenditure and expanding domestic compute supply in its preference for Chinese cloud and data-centre exposure.

Firm GPU rental rates still point to sustained infrastructure demand, keeping the semiconductor trade intact even as another group of beneficiaries emerges.

AI infrastructure created the first technology premium. Internet platforms are beginning to compete for the next one.

Lower AI Costs Are Strengthening Platform Economics

DeepSeek currently prices its V4 Flash model at $0.14 per million uncached input tokens and $0.28 per million output tokens, even as the company prepares broader API price increases.

The wider market trend still points toward lower inference costs, giving Cheap AI a stronger commercial role across platforms already operating at scale.

Lower cost per interaction gives those businesses more room to embed AI into search, advertising, recommendations, customer support and enterprise workflows without turning every use into a separate paid product.

Cheap AI is improving the economics of repeated deployment, allowing established platforms to turn higher usage into stronger commercial growth.

Distribution Is Becoming the Next AI Advantage

Microsoft is already showing what cheaper AI can look like when distribution is built in. Microsoft 365 Copilot passed 30 million paid seats in the June quarter, while Azure annual revenue exceeded $100 billion for the first time.

Those figures put AI inside software, cloud and enterprise relationships that already generate recurring revenue.

Stanford HAI’s 2026 AI Index found that 88% of surveyed organisations were using AI, with generative AI present in at least one business function at 70% of organisations.

Wider adoption raises the value of platforms that already control workflows, software ecosystems and customer access.

The next AI advantage is moving closer to distribution, where cheaper intelligence can be deployed across customers already paying for digital services.

Cheap AI Is Broadening the Technology Trade

The investment test is moving from access to AI toward commercial execution. Compute demand remains strong, but falling deployment costs are opening another layer of value across cloud, software, advertising and commerce.

Investors can now judge a wider group of companies on whether greater AI usage produces stronger revenue, margins and customer engagement.

China’s internet-stock rotation is providing the first market signal. The next premium will not come from cheaper intelligence alone, but from the companies that can distribute it at scale and turn usage into measurable financial growth.

Cheap AI is widening the technology trade beyond the infrastructure that powers it.