The track of 2026

Recently, observing AI-related news reveals a clear trend: major companies are no longer just competing in models but are now aggressively building up AI infrastructure. For example, $Meta Platforms(META.US) spent about $2 billion acquiring a company specializing in general AI agents, aiming to integrate autonomous decision-making and complex task execution capabilities into its ecosystem. $NVIDIA(NVDA.US) is also in talks to acquire an Israeli AI company, focusing on natural language and AI platforms, clearly strengthening its software and services. $IBM(IBM.US) went even further by investing $11 billion in a real-time data platform to enhance enterprise AI capabilities in data processing and real-time analytics. Even AI companies in security and defense sectors are acquiring AI capabilities for controlled environments.

These moves all point to one thing: models are just the brain, but what truly determines scalable implementation and sustainable profitability is computing power, data, and underlying infrastructure. Additionally, the U.S. is pushing the SPEED-related bill, essentially providing policy support for AI infrastructure. From a trend perspective, 2026 is likely to be the year when AI infrastructure truly takes off. It’s only a matter of time before companies like $Coreweave(CRWV.US) and $Nebius(NBIS.US), which focus on foundational layers, experience a market surge.

Moving forward, I plan to add AI infrastructure players like $Nebius(NBIS.US), $Coreweave(CRWV.US), and nuclear energy sector stocks such as $Oklo(OKLO.US) and $NuScale Power(SMR.US) to my watchlist, preparing for entry after a pullback.

The above is just personal observation and does not constitute investment advice.