Meta cut 8,000 jobs. Same week, $45B flowed to Anthropic.

Welcome back Chronicler!

This week, Big Tech tightened its grip on frontier AI, Meta cut 8,000 jobs to fund the next leg of its AI capex, and a Sequoia partner argued the next $1T company will not sell software at all. The throughline: AI is reshaping who builds, who owns, and who gets paid.

Plus: DPI vs TVPI vs RVPI (the metric that decides your next re-up), and Sequoia's services-as-software thesis that's spreading across LinkedIn.

As always, thank you for subscribing and engaging week-in week out. Still a work in progress, but I appreciate all 1,890 of you.

With Love,
Kev

THIS WEEK IN VC & TECH
What matters this week

LEAD STORY

Meta to Cut 8,000 Jobs as AI Capex Forces a Workforce Reshape

Meta will eliminate roughly 10% of its workforce, about 8,000 roles, on May 20, while also closing 6,000 open requisitions. The cuts arrive alongside a guided 2026 capex of at least $115B, up from $72B in 2025, with Microsoft simultaneously offering buyouts to about 7% of its US workforce. The signal for VCs is that the most expensive talent on the market is now available again, and seed-stage compute and infra startups should expect a deep bench of senior ex-Meta and ex-Microsoft engineers ready to be founding hires within 60 days. The deeper signal for founders: Zuckerberg himself said projects that used to require big teams are now done by single talented people, which means investors should price every early-stage pitch against the team-of-three counterfactual.

QUICK HITS

House Foreign Affairs Committee advances bipartisan AI export control bills. The panel moved a slate of bills, including the AI Overwatch Act, that would give Congress a 30-day window to block any advanced chip export license to foreign adversaries and impose a hard denial for anything above H200-class. Cross-border AI funds should expect tighter outbound investment screening and longer compliance timelines on China-adjacent diligence by Q3.

Cohere acquires Germany's Aleph Alpha in $20B transatlantic sovereign AI deal. Schwarz Group is committing $600M to anchor the combined entity, which will target defense, energy, finance and public-sector contracts in the EU. European founders building on US foundation models should watch this closely, because sovereign procurement budgets will start flowing to vendors who can credibly claim a non-US compute footprint.

Tesla lifts 2026 capex to $25B for AI compute, Optimus and Cybercab. Tesla plans to grow from 120K to 280K H100-equivalent GPUs by mid-year and is building an Austin Terafab for chip design, even as free cash flow turns negative for the rest of the year. Robotics and humanoid component startups should expect a Tesla-shaped vacuum across their supplier and talent pipelines starting this summer.

Microsoft offers first buyouts to ~7% of US workforce as AI cost pressure spreads. The buyout offer follows UKG's 950-person AI-cited cut on April 21 and brings the running 2026 tech layoff count past 96,000 across 249 companies. The hidden VC implication is a flood of Big Tech operators with vested equity and severance runway, which historically compresses pre-seed pricing by 20 to 30 percent in the six months after a major Big Tech RIF wave.

PRESENTED BY BELAY

Most business owners have financials. Few have financial clarity. BELAY's outsourced accounting team manages your books, tracks key metrics, and delivers timely reporting so you always know where your business stands — and what to do next.

Download the Free Guide

MONEY IN MOTION
The largest deals that you should know about

Anthropic essentially priced itself this week through compute commitments, not cash. Cursor doubled to $50B in five months and a Sequoia bet on AI-native services quietly reframed what Series A means in 2026.

Anthropic  ·  Frontier AI  ·  $5B from Amazon plus up to $40B from Google, $350B post  ·  Led by Amazon and Alphabet  ·  Link

Cursor (Anysphere)  ·  AI Coding  ·  Talks for $2B at $50B+ pre  ·  Led by Andreessen Horowitz and Thrive Capital  ·  Link

Supabase  ·  Database  ·  $200M at $2B valuation ·  Led by Accel  ·  Link

Auctor  ·  AI-Native Services / Enterprise Implementation  ·  $20M Series A  ·  Led by Sequoia Capital  ·  Link

Ray Therapeutics  ·  Biotech / Vision Restoration  ·  $125M Series B  ·  Led by Janus Henderson Investors  ·  Link

Omni  ·  AI Analytics  ·  $120M Series C at $1.5B post  ·  Led by Iconiq Growth  ·  Link

VENTURE BITE: One Concept in One Minute
DPI vs TVPI vs RVPI

A 3.0x TVPI fund can still owe LPs most of their money back. Paper gains are not distributions.

CONCEPT

TVPI is total value (cash returned plus residual NAV) divided by paid-in capital, while DPI is only the cash actually distributed and RVPI is the unrealized portion still sitting in portfolio marks. LPs pay bills with DPI, not TVPI, which is why a fund showing strong markups can still be considered immature or risky until exits convert paper into wires.

LEARNING OBJECTIVE

Decompose a fund's reported multiple into realized and unrealized components and judge fund maturity accordingly.

WORKED EXAMPLE

Take a 2018 vintage $200M fund, now in year 8. LPs have contributed $190M of called capital. The GP has distributed $250M back to LPs (mostly from one fintech IPO and a strategic acquisition), and the remaining 9 portfolio companies are marked at a combined NAV of $320M. Compute: DPI = 250 / 190 = 1.32x, RVPI = 320 / 190 = 1.68x, TVPI = (250 + 320) / 190 = 570 / 190 = 3.00x. Headline reads "3x fund," but only 1.32x is locked in. If the remaining NAV gets a 40% haircut in a downturn, residual value drops to $192M, TVPI falls to (250 + 192) / 190 = 2.33x, and DPI is unchanged at 1.32x. To clear a 2.5x net DPI target, the GP still needs to convert roughly $225M more of NAV into cash distributions, which on $320M of marks implies executing exits at about a 30% discount to current carrying value and still hitting the goal.

Quiz (Difficulty: Hard)

AI TOOLKIT FOR VCs
Our curated stack for the modern investor

Wispr Flow turns your voice into clean, professional text you can send the moment you stop talking. Not rough transcription you have to clean up. Actual polished text — ready for email, Slack, or any app.

Speak the way you think. Go on tangents. Change your mind mid-sentence. Flow strips the filler, fixes the grammar, and gives you text that reads like you spent five minutes writing it.

89% of messages sent with zero edits. Millions of professionals use Flow daily, including teams at OpenAI, Vercel, and Clay. Works on Mac, Windows, and iPhone.

Start flowing free

THIS WEEK’S STANDOUT

🪉Harmonic

Harmonic is an AI-powered company intelligence platform that tracks tens of millions of startups in near real time, surfacing founder signals, hiring spikes and stealth companies before they hit traditional databases. For VCs, it has become a core deal sourcing layer, letting investors find pre-seed and seed companies weeks before they appear on PitchBook or Crunchbase. The stealth-detection workflow has quietly become the default outbound starting point at multiple top-tier seed firms.

🏹 Affinity  ·  Relationship intelligence CRM that auto-builds your network graph from email and calendar data, now widely used for warm intro routing and pipeline management across venture firms.

👓 Specter  ·  Data-driven sourcing platform that ranks startups by growth signals across web, app, hiring and social, useful for VCs running thesis-driven outbound.

💿 Synaptic  ·  Alternative data platform aggregating web traffic, app downloads, headcount and consumer signals for diligence and portfolio monitoring on private companies.

🔍 Tracxn  ·  Sector-mapped private market database with analyst-curated taxonomies, helpful for fast market mapping and competitive landscapes during diligence.

FROM X
Our favorite post from this week

Bek's services-as-software thesis caught fire on X this week and is already showing up in deal memos. Sequoia's $20M Auctor check the day before publication shows they are putting capital where the thesis points, and emerging managers should expect this framing to define the next 18 months of applied-AI deal flow.

PRESENTED BY MARKETBEAT

AI isn’t a tech trend – it’s a full-blown, multi-trillion dollar race, and 10 companies are already pulling ahead.

These are the innovators driving real revenue, attracting institutional attention, and positioning for massive growth.

Get all 10 tickers in The 10 Best AI Stocks to Own in 2026, free today.

Download Now

THE BOARD
Top 10 jobs, curated for you

Here's what you've been waiting for: exciting job opportunities! To maximize your chances, remember these key tips:

That’s it for today, hope you enjoyed this as much as I did curating it, see you next week!
-Kev