Willy BRAUN

co-founder @galionexe / ❤️ books & social sciences / Prev: op partner @revaia_rap, co-founder @daphnivc / first gm @frdigitale / co-editor @CapitalCall_

[Willy BRAUN @WillyBraun]

@dhuynh95 Can’t wait to test it :)

[Daniel Huynh @dhuynh95]

Glad to announce Reduck MCP launch: the fastest and easiest way to connect your Agent to any website!
We enable agents to automate lead gen, GEO, or downloading invoices thanks to our library of scripts executable as tools on your Chrome Extension https://t.co/M1qsXz4yuu

[Pavel Prata @pavelprata]

Monthly VC/LP debrief.

What I actually saw in April, 2026:

1/ The market is basically collapsing into a single point. In Q1 2026, 5 firms captured 73.1% of all LP commitments (vs. 12 firms capturing 75% in 2025). Those same 5 firms are putting 75% of their capital into 5 companies. And those 5 companies are paying $1M+ comp packages and pulling the best researchers and operators out of the rest of the market. Capital, deals, and talent – all three concentrating into the same handful of names at the same time.

2/ Sovereign wealth funds owned 1.8% of venture's LP base in 2005. Today: 15.2% – $548B in committed capital. After 2022, distributions to traditional LPs collapsed from 29% of NAV to 11%. Endowments couldn't recycle. Pension funds hit allocation ceilings. Sovereign funds had no such constraints and kept writing checks while everyone else stopped answering the phone. (h/t @lananth)

3/ Large endowments are running SpaceX at 8-20% of NAV, often across 10-15+ funds simultaneously. When it IPOs, that capital has to go somewhere. The recycling effect into new VC commitments could be the single biggest LP liquidity event in years – and emerging managers fundraising in that window are going to be in a very different market than the one today. (h/t @MeghanKReynolds)

4/ The talent bifurcation is happening at the GP level too. Big platforms are pulling in the best partners – @lishali88, @mignano, @nikesharora, @Beezer232. At the same time, spinouts are accelerating – @kirbyman01, @suzannexie. The same dynamic compressing the LP market is playing out inside firms: strong platforms get stronger through talent, and the ones leaving are building emerging firms to play a different game entirely.

5/ Concentrated funds and diversified funds serve different moments in a company's life and the math bears it out. Concentrated funds are structurally risk-averse: total failure is too painful, so they come in after the fundamentals are derisked. Diversified funds take the moonshots early. The math: for every 100 concentrated funds in the market, there should be ~69 diversified ones. Most funds sit in the middle (neither particularly concentrated nor diversified) which is exactly where the structural alpha disappears. (h/t @credistick)

6/ @cartainc data: the 95th percentile seed-stage post-money valuation went from $65.6M in early 2022 to $173.6M in Q1 2026. The original angel investors in OpenAI put in ~$10M collectively and are sitting on ~$1.4B in paper at an $852B valuation – 140x on the most important company of the decade. If that's the ceiling on the biggest winner, the math on diversified seed portfolios is broken. The 1000x outcome is structurally disappearing. (h/t @lucasbagnocvaz)

7/ Alameda Research was Cursor's first check – $200K for half the company at pre-seed in 2022. Sold in FTX bankruptcy for $200K. Cursor is now tracking toward a $60B acquisition by SpaceX – zero to $60B in ~4 years. For @a16z and @ThriveCapital who led early rounds: ~150x on the Series A in ~2 years. A small group of firms, again, capturing one of the highest-quality outcomes. (h/t @daveclark85, @Jeffreyw5000)

8/ The quarterly update is no longer a one-way instrument. The 2020/2021 vintage funds made it obvious – inflated entry points, brutal macro turn, and suddenly a lot of managers weren't just navigating portfolio reality but managing how it was perceived. That's getting harder. A smaller FO can now open a $20 Claude and sanity-check the gap between what was promised 3 years ago and what actually got built – no analyst team required. The interpretation layer is no longer exclusively on the GP's side, and managers who relied more on narrative than substance are starting to feel it.

9/ Brand-name mega-funds and small founder-operator funds are not the same LP product and allocators who underwrite them with the same metrics are making the first mistake. Mega-funds are the "Blackstone-ification" of venture: scale, platform, brand. Small funds win on optionality – lead rights, co-invest, secondaries, faster DPI. Blending them into one "VC allocation" is where most LPs make the first mistake. (h/t @Noah_L)

10/ Nobody actually knows what makes a great founder or a great company – and that's exactly why seed funds will always exist. LP interest is lagging right now because of imminent IPOs in generational companies, but that's short-term. Multi-stage funds have too much capital to deploy small checks meaningfully. The bar is high, the market is brutal, and most people are wrong most of the time. Be a student to the job, don't get overconfident – and you'll likely thrive. (h/t @NWischoff)

Every month I track new fund launches, LP events, market reports, and what's actually moving in VC/LP.

All of it in the @murphcapital newsletter: https://t.co/IlZ3tMTtLx

[Willy BRAUN @WillyBraun]

Plus: weekly TechLunch insights from 300+ AI engineers, portfolio updates, and a zeitgeist section on what's shifting beneath the surface.

[Willy BRAUN @WillyBraun]

Newsletter.exe #20 is out.

This one started as our LP letter. We decided to make it public, full state of tech, AI, and VC markets going into 2026, including 91-slide from our LPAC deck, open-sourced for the ecosystem.

https://t.co/nhRcSj5E11

What you'll find inside:

[Willy BRAUN @WillyBraun]

🌱 Three deep dives Molecular farming disrupting the $65B crop protection market. First AI-designed molecule reaching Phase IIa. The crack in the $80B+ SEO market as discovery goes conversational.

[Willy BRAUN @WillyBraun]

The session will take place on the 8th of December at 7:pm CET on Contrast.

[Willy BRAUN @WillyBraun]

Join us for a one hour session to get his unfiltered learnings, the behind the scenes operational playbook including workflows, timelines, tooling, partners and pitfalls.

[Willy BRAUN @WillyBraun]

[SOTA GTM Series] How do you turn events into a scalable GTM channel?

@k7vin orchestrated twenty meetups in one month for @DustHQ across two continents and seven cities with top tier speakers.

🔗 Register here: https://t.co/UZSPA5QY0i

[Willy BRAUN @WillyBraun]

Our portfolio company @prior_labs just released a major new milestone: TabPFN-2.5, the most advanced foundation model for tabular data to date.

Instant predictions, zero tuning, and state-of-the-art accuracy. SOTA and impressive.

Kudos to the team!!

[Frank Hutter @FrankRHutter]

The data science revolution continues — TabPFN is now SOTA up to 50k data points and 2000 features 🚀
For the size limits of TabPFNv2, in a forward pass Real-TabPFN-2.5 outperforms AutoGluon 1.4 (complex ensemble including TabPFNv2 tuned for 4h) by 93 ELO points on TabArena.🧵1/

[Willy BRAUN @WillyBraun]

Because as foundation models converge in capability, deployment excellence becomes the durable advantage.

Full article: https://t.co/xcfaAbevSD

[Willy BRAUN @WillyBraun]

Most AI deployment strategies fail because teams attempt strategies their physics forbids. But how do you execute within your constraints? I've built two diagnostic frameworks to help adapt the deployment strategy:

[Willy BRAUN @WillyBraun]

𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸 #2. 𝗧𝗵𝗲 𝟱 𝗙𝗼𝗿𝗰𝗲𝘀 𝗼𝗳 𝗗𝗲𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁 𝗠𝗲𝘁𝗮𝗯𝗼𝗹𝗶𝘀𝗺 (feedback loops, iteration speed, interconnection depth, reversibility, scale)

[Willy BRAUN @WillyBraun]

Because as foundation models converge in capability, deployment excellence becomes the durable advantage.

Full article: https://t.co/xcfaAbevSD

[Willy BRAUN @WillyBraun]

𝗙𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸 𝟮: 𝗧𝗵𝗲 𝟯 𝗖𝗼𝗻𝘀𝘁𝗿𝗮𝗶𝗻𝘁 𝗗𝗶𝗺𝗲𝗻𝘀𝗶𝗼𝗻𝘀 (product physics × customer type × lifecycle stage, with constraint profiles mapped)

The principle: maximise learning velocity WITHIN your constraints, not despite them.