🌊 The AI money loop
When AI money chases itself.
I’m Ivan. This is a newsletter about startups and investing—for founders and investors. Each week I cover how top 1% startups use AI, raise, and grow.
Hello there,
Big week in ai and venture capital land, here’s what stood out:
🌊 AI
I pulled together the Top 30 European AI financing rounds of 2025, from Mistral’s $1.99B mega-round to fresh $40–100M bets across legal, health, and robotics. Most are AI-first, all are hiring, and together they show Europe’s AI wave is both broad (defense, healthcare, compliance) and deep (infra, vibecoding, new materials).
Eric Flaningam at Felicis wrote a piece on how to size the market opportunity for Vertical AI (no surprise here, its big). The best AI apps expand “small markets” by cutting costs + driving revenue. We’re already seeing it in legal and health with Abridge, Harvey, and OpenEvidence. Next up are life sciences (Bluenote), real estate (HouseWhisper), automotive (Mia), insurance (Strala), and home services (Netic).
Azeem Azhar wrote about wether or not we’re in an AI bubble, using a cool (and useful) heuristic. On his dashboard: AI spend is ~1% of GDP (green), revenues are doubling yearly (green), but capex is 6x revenues (yellow). Valuations are hot but not dot-com hot (green), funding is strong but a $1.5T gap looms (amber). Net: a boom, not yet a bubble, unless growth slows or debt fills the gap.
Sam Altman wrote that he wants a factory that cranks out a gigawatt of AI infrastructure every week. Meanwhile OpenAI commits $300B to Oracle, Oracle stock jumps 36% (best day since ’92), Oracle buys Nvidia GPUs, Nvidia turns around and puts $100B back into OpenAI, raising legitimate concerns (and funny memes).
We invested in Pensero.ai. Think Gong for software engineering teams. Instead of managers wasting hours digging through GitHub or tickets, it shows what’s working, what’s blocked, and where performance can improve. Say hello to Dave for me if you want to try the product.
Great piece on the origins of Forward Deployed Engineer. Like a lot of Silicon Valley lingo, FDE gets thrown around as hype. But, at Palantir it meant dropping engineers into the field with autonomy to solve problems and feed those learnings back into the product. TLDR: FDE = a product discovery loop, not an implementation service.
Lenny interviewed Brendan Foody, founder of Mercor, the fastest-growing startup in history, going from $1M to $500M in 17 months. They organize experts to write “evals,” tests that measure how well AI models actually work. Those evals become the feedback loop powering the breakout growth of today’s AI companies.
Math inc. dropped Gauss, showing that math once thought to take decades of expert work can now be verified by machines at scale.
Replit just launched Agent 3, which can now run fully on its own for ~3 hours straight testing, fixing, and even building new agents (wild).
Google is finally baking AI straight into Chrome, bringing native writing help, page summaries, and smart tab management right into the browser.
Meta and Oakley just rolled out $499 smart glasses for athletes. Basically the first real AI sports wearable. Real-time coaching, stats, and AI in your line of sight. (I worked at Facebook Reality Labs back in the day, fun to see this).
Other cool tools I’ve found this week / prompts I’m exploring:
One of my favorite prompts from one of Reddit’s founders on “Theory of Change”.
Shared my AI toolbox for writing on Substack
Google mixboard - idea canvas with AI image-model built in
Sings-ai - learn sign language with ai
Twitter CRM - AI that manages your X Network
This week’s sponsor is AI CRM Attio (raised $52M from Google Ventures):
Attio is the AI-native CRM for the next era of companies: Connect your email and calendar, and Attio instantly builds a CRM that matches your business model — with all your companies, contacts, and interactions enriched with actionable insights.
With Attio, AI isn’t just a feature - it’s the foundation. You can do things like:
Instantly prospect and route leads with research agents
Get real-time insights from AI during customer conversations
Build powerful AI automations for your most complex workflows
Fast-growing companies like Granola, Modal, Replicate and more are all experiencing the future of GTM with Attio.
🌊 Venture capital
Carta published AI fundraising benchmarks, and an interesting valuation segmentation between ai + non-ai startups from seed to series A.
Palle Broe wrote about founder dilution at seed. The data shows most founders part with ~20% at seed on $1.5–3.5M raises, usually at $10–15M valuations. The rule of thumb: avoid over-optimizing, raise enough for 18–24 months of progress.
Jon GP at a16z wrote about him being bullish on founders that understand the power of focus and constrains when it comes to sizing seed rounds.
VC / GTM Reports
a16z - New era of consumer software
CB Insights - Legal Tech Market Map
Bessemer – Cloud 100 Benchmarks
Highspot – Sales Enablement 2025
HubSpot – AI in GTM Report 2025
Maxio – 2025 SaaS Performance
Mercury – 2025 Startups GTM
Gong – State of Revenue 2025
Outreach – Sales 2025 Report
SaaStr – The SDR Downsizing
ICONIQ – State of GTM 2025
G2 – 2025 Buyer Behavior
Google - ROI of AI 2025
🔒 Enjoyed this edition? Join Startup Riders Pro:
🔒 How not to raise money - a checklist of fundraising mistakes.
🔒 Fundraising Frameworks — 3 frameworks to tighten your pitch.
🔒 GTM + AI — How founders are rebuilding sales in 2025.
🔒 HubSpot’s 0-$100M Playbook — $0 to $100M sales strategy.
🔒 Founder Compensation — salary data from top founders.
🔒 Founder Info-Diet - 40+ quality info resources to stay ahead.
🔒 Agentic Revolution — Deep dive into the next tech platform shift.
🔒 AI Voice Agents — Market map of where capital flows.
That’s it! I’d love to hear from you, just reply to this email (I reply to all of them) 👋.
Great one!
By the way I'm enjoying your Notes more than ever now.