Long Live the AI Tech Bubble (news.crunchbase.com)

Nearly half of global startup funding this year has flowed into AI, with 16 companies capturing a third of capital in a single quarter and a single firm taking a disproportionately large share. While critics warn of a hype-driven bubble — pre-seed “unicorns” and sky-high multiples do exist — the broader pattern looks more like the early stages of a generational platform shift (think Netscape/internet). Massive infrastructure rounds are building foundational layers — model training, inference, data management — that lower the cost of building and spark waves of applied startups across legal automation, robotics, drug discovery and more. That shift is already producing real businesses with measurable ROI: examples include EliseAI automating leasing workflows for major property owners, Valence scaling personalized employee coaching, and Exodigo using multisensor AI to map underground utilities. Enterprise AI companies are reaching meaningful revenue far faster than legacy SaaS (often hitting ~$20M ARR in ~2 years versus ~7), with top performers growing >200% annually. The upshot for the AI/ML community: while valuations warrant discipline, the concentration of capital is creating durable infrastructure, accelerating application-layer innovation, and enlarging opportunities for founders and investors who can convert today’s foundational spending into profitable, enduring companies.