Antler CEO Magnus Grimeland says Silicon Valley doesn’t have a monopoly on tech: ‘People can innovate from almost anywhere’

Antler has grown to 27 offices across six continents, deployed more than $1 billion in AUM, and raised $510 million in 2025, including a $160 million U.S.-focused fund. The firm is leaning into AI applications and claims two recent unicorns, Airalo and Lovable, with Lovable reaching $100 million in ARR just eight months after founding. The piece is broadly positive for Antler’s global venture platform, but it is primarily a profile and is unlikely to move markets materially.

Analysis

The bigger signal is not that another VC is opening a Silicon Valley office; it’s that the sourcing advantage in venture is shifting from brand to geography-plus-coverage. A dense on-the-ground network in undercovered markets can create an information edge on founder quality before the U.S. consensus sees it, which should pressure later-stage incumbents that rely on fly-in diligence and warm intros. That dynamic is structurally negative for firms whose edge is access rather than selection, and it favors platforms that can monetize breadth with small check sizes and high portfolio counts.

The AI implication is more interesting than the VC narrative. If application-layer winners are forming faster than infrastructure winners can re-rate, the public-market beneficiaries are likely to be picks-and-shovels enablers with distribution into startups rather than the model labs themselves. That is modestly supportive for enterprise software and cloud-adjacent names, but it is also a warning that the next leg of AI alpha may come from vertical, workflow-specific tools that compress time-to-revenue dramatically, making TAM-based underwriting less reliable.

For the named public comps, the read-through to Meta is slightly positive: a more decentralized founder ecosystem increases the pool of builders targeting consumer attention and creator tooling, reinforcing the value of scaled distribution. Cisco is a more nuanced beneficiary because AI application growth still pulls incremental network, security, and data-center spend, but the article argues against a durable multiple rerate for infrastructure alone. Telecom names like T are the least compelling: more startup activity and AI adoption do not automatically translate into pricing power; they mostly increase bandwidth demand in a market where carriage remains commoditized.

The contrarian risk is that the market is overpaying for the narrative that “everywhere can be Silicon Valley.” In practice, only a subset of geographies produce repeatable, globally scalable founders, and the dispersion in outcomes is huge. If exit windows remain closed for another 12-24 months, broad-based VC expansion can become an underwriting and capital-efficiency problem rather than an advantage, especially for firms using aggressive portfolio construction before proof of liquidity.

Key Decisions for Investors

• Long META / short T over 3-6 months: benefit from AI-driven new startup formation and creator-tool adoption; pair keeps beta muted while expressing a tilt toward platform monetization over commoditized connectivity.
• Small tactical long CSCO into the next 1-2 quarters on AI application buildout as a demand-support story, but cap sizing: upside is real on capex intensity, yet multiple expansion should be limited because infrastructure remains a second-order winner.
• Avoid initiating new long exposure to GS on this theme; if anything, use rallies to fade exposure to late-stage venture/IPO advisory leverage, since a prolonged liquidity drought delays monetization for the VC ecosystem.
• Optionality trade: buy 6-12 month out-of-the-money call spreads on META rather than outright stock if seeking exposure to decentralized AI app creation, since upside can compound without requiring near-term earnings revisions.
• Do not chase telecom beta here; if T rallies on broad AI enthusiasm, use it as a short-term sell opportunity given weak pass-through from venture formation to regulated carrier economics.