Meta stakes its future on personalized 'superintelligence' with a massive CapEx push

Meta announced a strategic pivot that channels unprecedented capital into building highly personalized AI assistants, signaling a shift from product experiments to large-scale investment. The company plans to boost capital expenditures sharply, allocating roughly $115–$135 billion to support AI research, infrastructure and productization after spending about $72 billion the prior year. Leadership frames the plan as an attempt to create AI systems that understand individual users’ histories, preferences and social connections and then tailor experiences across Facebook, Instagram and WhatsApp. Executing this vision will require tighter integration between large language models and the recommendation engines that curate social feeds, and it will depend on operational links between Meta’s core apps and emerging device platforms. That device angle is underlined by a renewed push into AI-enabled eyewear: Meta reported roughly 3x year-over-year unit sales for its glasses and has shifted Reality Labs toward lighter AR devices and software that can surface personalized agents. The Reality Labs reorientation included about 1,500 role reductions and a pullback from some longer‑term VR ambitions as the company reallocates capital toward nearer-term consumer hardware and services. Parallel to hardware shifts, Meta is preparing limited trials of paid subscription tiers across Instagram, Facebook and WhatsApp that would gate advanced AI tools and Manus agent integrations behind recurring fees while keeping baseline access free. Those pilots are designed to test willingness‑to‑pay for creator tools, AI-generated short-form video features and business-facing Manus capabilities, creating a potential recurring revenue offset to heavy infrastructure spending. The push follows a turbulent year inside Meta’s AI teams—high-profile hires, internal disagreements and a reorganization that included hundreds of cuts to AI groups—underscoring execution risk. If Meta succeeds, personalized AI could translate into new monetization channels, higher engagement and deeper platform lock‑in; if it fails, the company risks amplifying privacy exposures and provoking stricter oversight. The broader industry context heightens scrutiny: major cloud and consumer platforms are also increasing capex, and investors are pressing executives to show how new build-outs will convert into sustainable revenue and margins. Over the next 12–24 months, Meta must demonstrate that heavy spending, tighter hardware–software integration and nascent subscription pilots produce measurable returns without eroding user trust or inviting prohibitive regulation.