Ideas
Massive AI infra buildout benefits builders.
Big tech capex is running at roughly $400B annualized, mostly for AI infrastructure and data centers, and infrastructure is being built for future training and inference demand. The largest tech companies are bearing the buildout burden and can handle potential overbuild, which benefits companies building on top of this infrastructure.
Big Tech can fund AI capex.
Google, Facebook, Amazon, and Microsoft are arguably the best companies ever created, and their cash-generating ability lets them bear the massive AI capex and potential capacity overbuild, making them durable leaders in the AI buildout.
Cheaper better AI fuels app growth.
Input costs for AI models have declined over 99% in two years while frontier capabilities double roughly every seven months, and AI is built on internet/cloud infrastructure enabling immediate global distribution. This creates a massive market opportunity for AI applications, with AI becoming ubiquitous like electricity or Wi-Fi.
Consumer AI monetization has huge upside.
AI business models allow effective price discrimination, unlike prior internet platforms, and there is a huge monetization gap between roughly 2 billion AI users and only 30-40 million paying users. Upside should come from monetizing free users through subscriptions and advertising, with high consumer stickiness.
Nuclear is best AI power source.
Energy is a bottleneck for AI data centers, and nuclear power is the most promising solution. There is a renewed embrace of nuclear, existing plants like Three Mile Island may restart, big tech is siting data centers near nuclear plants, and the firm has invested in nuclear.
xAI executes AI data center buildout fast.
xAI is a large portfolio investment, and its ability to stand up the biggest data center at the time in a quarter of the normal time demonstrates exceptional execution and speed in AI infrastructure buildout.
Cooling is next AI bottleneck.
After energy, cooling is the next major bottleneck for AI data centers; most people have not yet focused on it, and a wave of innovation in cooling will be needed to prevent overheating chips and environmental impact.
OpenAI consumer stickiness funds research.
OpenAI has an advantaged position because its large consumer base is stickier than enterprise/developer API usage; consumers are unlikely to switch for a slightly better model, giving OpenAI a better position to fund continued research.
Back high-momentum AI app leaders.
The firm targets AI application companies with undeniable momentum that are 'flying off the page,' naming Cursor, Decagon, Abridge, and ElevenLabs as examples of this first bucket of AI strategy.
Databricks offers asymmetric attractive returns.
Databricks has an attractive return profile with downside protection and upside optionality: the firm is confident in 3-4x and sees 2x as safe, with potential for 5x over five years.
Salesforce faces eventual AI disruption threat.
Salesforce.com is a potentially vulnerable public software incumbent because it is mostly forms/checklists with a sticky backend database and seat-based pricing; AI startups could disrupt it if they reimagine UI/UX, access unstructured data, and innovate business models, but the firm has not yet found a killer entrant.
American Dynamism is next big theme.
After AI infra and apps, American Dynamism is the next largest opportunity, driven by pressing market/world need, talent from SpaceX and Palantir, and advances in autonomy and vision technologies.
AI healthcare is growing opportunity.
The firm is starting to see more interesting AI-enabled health companies and will continue to be active there, though it is likely a smaller area than AI infrastructure and applications.
Stablecoin enablement is exciting opportunity.
The firm is seeing exciting developments in the enablement of stablecoins, working with its crypto team, and would increase exposure if the market sees a massive takeoff.
This a16z video, published January 26, 2026,
features David George, Jen Kha
discussing AIQ, DTCR, GOOG, MSFT, META, AMZN, AI applications, AI-SECTOR, URA, XAI, Data center cooling, OPENAI, CURSOR, Decagon, Abridge, ElevenLabs, DATABRICKS, CRM, American Dynamism, AI-enabled health, STABLECOINS.
14 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
David George,
Jen Kha
· Tickers:
AIQ,
DTCR,
GOOG,
MSFT,
META,
AMZN,
AI applications,
AI-SECTOR,
URA,
XAI,
Data center cooling,
OPENAI,
CURSOR,
Decagon,
Abridge,
ElevenLabs,
DATABRICKS,
CRM,
American Dynamism,
AI-enabled health,
STABLECOINS
