5 deals gobbled 40% of Q4's $75B

VCs pumped $74.6 billion into U.S. startups in Q4, yet $32 billion, over 40 %, went to just five AI‑heavy unicorns. The median startup saw no uplift, underscoring a widening funding chasm that leaves most founders still hunting capital. Expect future rounds to stay concentrated among AI leaders.

The Investing 101 essay argues that the push for ever‑larger funding rounds and sky‑high valuations creates fragile businesses that burn cash and crumble. It shows how a lean, self‑sustaining model delivers lasting value and why founders should resist the hype of hypergrowth.

Cloudflare hit $1.7 B ARR with 28% growth, but 65% of that comes from just 3,200 large accounts. The company also hit $400 k per‑employee revenue, half its sales now lie outside the U.S., and it proved self‑serve can work at enterprise scale. Those metrics show how a security‑heavy SaaS can scale profitably.

Superhuman turned down $8 M a month in potential revenue, limiting new sign‑ups to 100 a week and vetting each with a 30‑minute call. That restraint preserved product quality, delivering over $100 M ARR and industry‑leading retention while faster‑growing rivals burned out. The lesson: controlled growth can outweigh chasing scale.

Your first marketing channel boils down to two hard constraints: where your target audience already spends time and which medium feels instinctive to you. By intersecting these, you instantly cut dozens of options, focus effort on high‑engagement touchpoints, and avoid the costly trial‑and‑error of unfamiliar platforms.

Most B2B product‑led growth teams assume a slick signup flow is enough, but without end‑to‑end user metrics the motion stalls. The post shows how legacy sales stacks block the data needed to track sign‑ups, conversions, and churn, and outlines the data foundation required to scale PLG.

Steph Ango, Obsidian’s CEO, argues that design isn’t about perfection but about deliberate compromises. Opinionated products win by embracing clear trade‑offs, letting you be exceptional where it counts. The lesson for leaders: own the weaknesses you discard and double‑down on the strengths your audience values.

Speed alone fuels burnout; the real AI advantage is sharpening focus so knowledge workers can finish what matters. By using AI to deepen, not widen, effort, companies can cut open‑loop projects, boost impact, and keep teams healthy.

Fastly fell from a 36.5x revenue multiple to the bottom of its peer set because its gross margins stayed under 60%, revenue growth stalled at 2% YoY, and free cash flow turned negative. Those unit‑economics flaws also triggered a net‑retention death spiral, showing that high‑growth cloud firms can’t survive on slim margins alone.

Dianne Penn, Anthropic’s first technical product manager, recounts how the team’s eval‑driven development loop turned Claude into a coding powerhouse and why “token maxing” became a competitive edge. She also outlines the next frontier beyond coding, tool use, reasoning, and human‑in‑the‑loop judgment, showing how Anthropic builds AI that scales with product rigor.

Lift House, a six‑story co‑living hacker house in East London, replaces 24‑hour hustle with shared chores, flexible stays and a focus on holistic wellbeing. Its founders claim the model can produce world‑class AI startups without the toxic ‘996’ culture that defines Silicon Valley houses.

Moonshot AI unveiled Kimi, a Chinese model that rivals frontier U.S. systems on benchmarks, sparking a wave of alarm on Wall Street and in Silicon Valley. Executives worry it could tilt the AI race toward Beijing or force restrictive policies that benefit a few domestic labs. The debate now shapes lobbying, regulation, and funding strategies.