
Seed stage investor at @fcollective. Over 50 founding teams served & counting! Founder of 3 tech companies - Handshake, Brontes (acq by 3M) and Sample6/Corvium.
New York City
Joined June 2009
- Tweets9.9K
- Following1.2K
- Followers26K
- Likes13.8K
Pinned Tweet
In order to respond to the question "what do you invest in?" from founders, angels, scouts, VCs, LPs, and my mother... we've created this easy to use cheat-sheet. Only 26 categories - guess we don't follow that many themes ;)
Man v AI in videogames! (as a proxy for strengths and weaknesses of models)
I still think I can beat AI in Mike Tyson's Punch Out! (not much else)
Congrats to the team!
We launched a new benchmark: KradleBench
Yesterday morning, our human entry (@sethbling) was head-and-shoulders above every AI model.
Then Astra 6.2 launched and beat him. I'm still processing this
To see AIs improve so fast in multi-agent realtime games makes the curve real
1
3
699
There's a subtle difference between investing and deploying but I struggle to put my finger on what the characteristics and mindsets are of each. Thoughts?
Nvidia has become something like the central bank of the AI economy. It reminds me of running a dot-com in 2000, when the media companies we were buying ads from offered to invest in us. The money would have gone right back to them as marketing spend. Not saying it’s exactly the same. But it rhymes.
Massively growing fund size has become consensus which almost by definition means lower returns.
And from what I’ve seen from peers in Hedge funds or PE the strategies that worked at small scale don’t necessarily work at 3-10x the size.
Venture capital is dead (again).
More money is being raised by fewer investors. So far in 2026, funds over $1 billion have taken 72% of all venture money raised. Last year it was 34%. Ten years ago, it was 16%.
Even firms famous for staying small, like Benchmark and Union Square Ventures, have caved. Benchmark’s five partners raised $425 million in 2024. This June, they raised $2 billion.
When a fund has too much money per partner, they are forced to cope by either making bad deals or by taking on less uncertainty. The former is bad business, the latter is not venture capital.
The last time this happened was the 1980s. A decade when money per venture investor tripled, median returns fell from above 30% to 8%, and firms moved into buyouts and rollups. Even Sequoia dabbled in LBOs. By 1994, Marc Andreessen called Silicon Valley “dead in the water.”
In his latest essay, Colossus contributing editor @ganeumann explains what happened next, what’s likely to happen now, and which firms are best placed to win.
colossus.com/article/venture…
This may be the best management manifesto I’ve ever read. All founders should read this. All companies don’t need to act and look exactly like Ro but there are so many nuggets here to pull from.
Thank you Z for writing this.
We're at the point of the cycle - or our industry - where the meta discussion of the industry is bigger than the actual business itself. Sort of like when the business of sports became bigger than the sports themselves.
3
4
823
In the crypto age, I heard the word "canonical" multiple times per day. Now I hear the word "legible." Every cycle has a lingo.
3
16
1,125
It’s so cool to see all the marathoners striding around the city with their breakers and medals. I’m envious. But the parent in me worries that they are all limping;)
2
8
1,522
The beauty of You Can See Everything was that it was so non-obvious that there was room for another telling of the Theranos story.
Why do airlines show where you are on the upgrade list - feels like it only underscores devaluation of status - eg number 32 with 1 seat available.
Why not come up with a probability system or some other way to show it? Even red yellow green might be more
palatable.
3
8
1,897
Turns out the human parts of hiring, training and managing are still hard - fortunately the tech is getting better and smarter. Autonomy + Humanity FTW
Growing companies think their biggest problem is hiring, but they are wrong. It’s only half the story.
In reality even the best candidates will fail if they are not set up for success from the start.
I experienced this at @yext across a team of 2,500, and today, with the pull of our growing customer base, Windmill is launching the system I always wanted
An agentic onboarding system to give every new hire the best chance of being successful as quickly as possible.
@trywindmill now has over 250 customers and many of them growing faster than any companies before in human history. Our customers feel this pain viscerally. You spend millions on recruiting and closing candidates only for them to show up to an onboarding based off a couple slides and a Notion doc.
Onboarding is one of the most important things a company does. It’s a new hire’s introduction to the company culture and how it runs. And in today’s rapidly changing world, it’s harder than it’s ever been before.
Effective onboarding isn't about teaching people everything. It's about getting them to real contribution as fast as possible, with clear goals, regular feedback, and a manager who stays in the loop.
Windmill puts an AI agent in charge of making that happen for every new hire, and holds everyone accountable along the way.
We’re currently live with amazing customers like @Etched (among others) and it’s incredible to see the results. Every one of our customers I’ve shared this with has raised their hands and we’re including it as part of their subscription.
If this is something your team struggles with, we’d love to show you the system that’s helped customers around the globe onboard and ramp their teams 3x faster.
5
19
770
Micah Rosenbloom retweeted
ISO an amazing VC Partnerships Manager to work with us at @OpenAI
Ideal person? Founder, investor, or operator (ideally you've done it all) who loves getting in the weeds and truly building.
You love startups and startups love you.
openai.com/careers/vc-partne…
Some companies genuinely failed but others IMHO ended as zero to founder because the valuation was too high, the preference stack too big, and/or the founders and VCs didnt have the M&A convo often enough (guilty as charged;).
We built a business that only pays at the extremes. Both sides sign up for it, but the VC is diversified. Founder is not.
Micah Rosenbloom retweeted
OMFG… WHO DID THIS?! 💀🤣 nitter.cf/JTmvp0no/status/211316…
You spend years changing diapers, teaching, driving them around, helping with homework, teaching them right v wrong.
Then they go to college.
And you communicate by sharing funny memes all day long. This week its Bears - Patriots;)
10
32
2,011
So good- I wish I wrote this.
It’s no longer just founders optimizing for the next round. It’s VCs saying “will a Tier 1 want this at the A,” which is front-running a perceived mark, not partnering for business building.
When building something instrinsically valuable becomes a secondary goal, that's not a good thing. I fell for the same thing in 1999 and ended up empty handed.
100% and there aren't enough stories told like this - too much of the wins, not enough of the ups and downs which make up the lions share of start-up experiences - we all thank you @dens
In addition to everything you mentioned @dens, please factor in that you inspired a whole generation of entrepreneurs (including me) by showing them that:
- you can build a cool startup from NYC
- you can build something fun and that you care about
- you can persevere & stick to trying to build what you're excited about no matter what (it's the journey, not the destination!)
Thank you Dennis & team!
Yep the data tells the story. The future would have to be so dramatically different than the past 25 years. I believe the big outcomes will be bigger. But the median over 25 years of top 500 exits is "only" 2.7B. Hard to justify capital / valuations / fund sizes given the math. linkedin.com/pulse/unicorn-l…
There are only 130 US tech companies, public or private, valued at $10B or more (decacorns). That’s roughly 20 to 30 new companies per decade, for the last five decades. This year alone, with a quarter to go, there have been 42 decacorns minted privately. Either this is the most economically transformative event since the industrial revolution, or we’re getting a wee bit ahead of ourselves.















