Earning AI opinions , VC liquidity crisis , rebuilding custo
Earning AI opinions , VC liquidity crisis , rebuilding customer success
Hey guys, Mr. Technology here — three stories that don't look related but are. The AI discourse is overrun with opinions nobody earned. The VC money that funded the AI wave is stuck. And customer success teams are now being asked to do the work of three people with half the headcount.
What You Need to Know: A widely-shared essay argued that the loudest voices on AI are people who "absorb a position by proximity" rather than shipping with the tools; new data from a UK VC conference shows 70% of venture funds have extended past their 10-year term; and AI-native customer success playbooks are reshaping CS org charts in 2026.
Story 1: "Your opinion on AI is only worth something if you earned it"
That's the core line from an essay that made the rounds in product and engineering circles this week. The author — writing in the Lenny's Newsletter / Substack ecosystem — argues that the AI discourse has become a status game. VCs, executives, and pundits are publishing confident takes on agents, vibe coding, and the future of work without ever having shipped anything with the tools. They're absorbing a position by proximity: they read three threads, attended one dinner, and now they have a worldview.
The piece distinguishes between three kinds of AI opinion: people who have shipped (small group, opinions worth listening to), people who are running experiments (larger group, opinions worth taking seriously with a grain of salt), and people who have opinions because they're paid to (the largest group, the source of most of the noise). The takeaway: "earned" means you've actually built something, debugged something, watched a model fail at 2am, and you can describe what worked and what didn't with specifics.
The framing lines up with a related HBR piece from April by Daisy Auger-Domínguez on how burnout looks different across the org chart, and with Ravi Mehta's argument in TLDR Product that AI hasn't made product management easier — it just moves the bottleneck from production to judgment. The pattern across all three: the hard part of working with AI is human, and the people who claim otherwise are usually the ones who haven't done it.
Story 2: The Great LP Liquidity Crisis is real, and it's reshaping VC
At the Women's Venture Capital Summit Europe 2025, consultants cited research on 17,000 funds showing that 70% of venture funds have extended beyond their initial 10-year terms to give themselves more time to return capital. Other sources put the number worse. The distribution drought has created a vicious cycle: LPs haven't received meaningful cash returns in years, so they're reluctant to commit to new funds; without fresh capital, portfolio companies struggle to raise follow-ons, which depresses valuations, which makes exits even harder, which keeps distributions low.
The metric that matters now is DPI — distributions to paid-in capital. "You cannot raise another fund when you do not have DPI or a very good track record," one European fund manager told the conference. The response has been a flood of secondary transactions: GPs are actively encouraging portfolio companies to explore secondary sales, even at the cost of upside, just to generate some cash for LPs.
This is the structural reason AI startup valuations look the way they do in 2026. Late-stage funds need liquidity events. AI companies with revenue are the only category generating exits right now. So every Series C and later round is being priced to clear — and the founders who understand that dynamic are the ones getting term sheets.
Story 3: Rebuilding customer success for the AI era
ClientSuccess and Customer Success Compass ran a joint webinar in January laying out a 2026 CS playbook, and the changes are more practical than the hype suggests. The core argument: AI is no longer optional for modern CS teams, but implementation matters. The teams winning are the ones using AI to predict churn, personalize journeys, and amplify human CSM impact — without losing the relationship.
The concrete shifts: structured discussion prompts replacing dashboards (Lenny's Newsletter has been pushing this for two years), AI-driven adoption/retention/expansion metrics replacing manual QBR prep, and a hard reorg of the CSM role around outcomes rather than accounts. The 2026 CS org chart has fewer "account manager" titles, more "outcomes lead" titles, and a layer of AI tooling that takes 60-70% of the rote work off the human's plate.
The piece lines up with a parallel trend: product management itself is being asked to move from "prioritization" (what to build next) to "curation" (what to ship, what to kill, what to defend) as the cost of saying "yes" to a feature drops to near zero.
The Take
Read those three stories in this order: the people with AI opinions aren't the ones shipping, the money behind them is stuck, and the customer success teams meant to retain the resulting revenue are being asked to do more with less. The whole stack is under pressure at the same time. The people who come out the other side with careers intact are the ones who can do all three: ship with AI tools, understand the financial reality of their employer or their investor, and operate inside a CS or PM function that is being reorganized around outcomes. Talking about AI without doing AI is now a measurable liability.
Quick Summary
The AI-opinion discourse is full of people who haven't shipped. 70% of VC funds have extended past their 10-year term as LPs wait for distributions. And 2026 customer success playbooks are moving from account-management metrics to outcome-based AI-augmented teams.
Sources: