🧠 Food for thought

Implications, context, and why it matters.

Meta is spending far more to push AI ads

• Meta expects capital spending to reach as much as US$145 billion in 2026, about twice its roughly US$72 billion in 2025 1.
• The company says it has kept underestimating how much AI computing it needs for foundation models, which are large AI systems trained on vast amounts of data 2.
• Meta says ad tools such as Lattice and GEM lifted landing-page-view conversion rates by more than 6%, which helps explain the cost 2.
• Its Value Optimization Suite, a set of tools that helps advertisers aim campaigns at higher-value customer actions, has more than doubled its revenue run rate to above US$20 billion over the past year 2.

Huge AI budgets are reshaping big tech

• Meta’s outlay suggests competition now hinges on who can pay for private AI infrastructure on the scale of a small country’s GDP 1.
• Only a few hyperscalers, the biggest cloud and internet companies with enormous computing infrastructure, can spend at this level, which could lock in their power and leave smaller AI rivals behind 3.
• Meta also plans layoffs while saying AI tools have raised output per engineer by 30% since the start of 2025, pairing bigger capital spending with a leaner workforce 4.
• Geopolitical tension has become a business risk, as China ordered Meta to unwind its acquisition of AI startup Manus for more than US$2 billion 5.