the ocean does not care about your cap table.
neither does the bike.
that is exactly why I need both.
neither does the bike.
that is exactly why I need both.
4
0
20
1,758
It is an uncomfortable truth that 70% of pre-seed funded startups never raise another dollar.
Seneca: we are dying every day, little by little.
So is your company. Quietly. On a Tuesday. While you polish the deck.
Memento mori isn’t morbid. It’s runway math.
Seneca: we are dying every day, little by little.
So is your company. Quietly. On a Tuesday. While you polish the deck.
Memento mori isn’t morbid. It’s runway math.
13
4
48
4,654
cold water. early mornings. no alcohol. long rides.
performance is not a hack.
it is a series of small, boring, unglamorous decisions made every single day.
performance is not a hack.
it is a series of small, boring, unglamorous decisions made every single day.
2
0
25
1,926
I have passed on good decks with bad founders.
I have funded bad decks with great founders.
the deck is a moment in time.
the founder is the whole game.
I have funded bad decks with great founders.
the deck is a moment in time.
the founder is the whole game.
12
5
84
2,954
most founders want to grow faster.
few ask why they are growing at all.
understand the why before you pour fuel on the fire.
otherwise you are just burning things down faster.
few ask why they are growing at all.
understand the why before you pour fuel on the fire.
otherwise you are just burning things down faster.
7
3
37
2,134
old but still good guidance on a seed deck
ycombinator.com/library/2u-how…
ycombinator.com/library/2u-how…
7
6
103
9,355
yes some harnesses get so good that they swap out the general purpose brain for a specific purpose brain (Cursor, Harvey, etc.). This is one path and certainly has legs, but most niches are not large enough for their own trained brain.
2
1
17
1,898
Big multistage funds now decide which startups become consensus.
If you're a founder, that leaves you two kinds of investors.
Consensus investors show up once everyone already agrees on you.
Pre-consensus investors back you before anyone agrees.
At Incisive Ventures, we only do the second.
That means we're in before the data proves you right. We judge the founder, not the traction. And we're the ones who stay in your corner when the big funds are still saying "come back when you have more."
300+ pre-seed investments over 20 years.
If you're building something nobody believes in yet, pitch me. DMs are open.
If you're a founder, that leaves you two kinds of investors.
Consensus investors show up once everyone already agrees on you.
Pre-consensus investors back you before anyone agrees.
At Incisive Ventures, we only do the second.
That means we're in before the data proves you right. We judge the founder, not the traction. And we're the ones who stay in your corner when the big funds are still saying "come back when you have more."
300+ pre-seed investments over 20 years.
If you're building something nobody believes in yet, pitch me. DMs are open.
13
3
57
3,419
The Human brain is very good at harnessing.
The Computer brain is already better than Humans in some subset of harnessing tasks (writing code) and is getting better.
The next 10-20 years will refine and evolve this separation of work between the two brains.
The Computer brain is already better than Humans in some subset of harnessing tasks (writing code) and is getting better.
The next 10-20 years will refine and evolve this separation of work between the two brains.
3
0
24
1,924
OK, I passed on LLM Wrappers, but am going all in on Harnesses,
Why?
What is the difference?
Wrappers were basically just complex prompts to the LLM. Telling it how to act, what to use, etc. Basically, specific instructions to the brain on how to think. They are thin, and the final output relative to the contribution of the LLM.
Harnesses give the brain a full body and resources to complete more complex tasks, use tools, get context, etc. A Harness is a full system around a brain (LLM) to solve a certain set of problems well (Claude Code: Coding). The Harness has value separately from the brain in it's deep understanding of a specific workflow, context, or integrations into other systems to get work done. Their value is substantial in improving output quality relative to the LLM's contribution.
Why?
What is the difference?
Wrappers were basically just complex prompts to the LLM. Telling it how to act, what to use, etc. Basically, specific instructions to the brain on how to think. They are thin, and the final output relative to the contribution of the LLM.
Harnesses give the brain a full body and resources to complete more complex tasks, use tools, get context, etc. A Harness is a full system around a brain (LLM) to solve a certain set of problems well (Claude Code: Coding). The Harness has value separately from the brain in it's deep understanding of a specific workflow, context, or integrations into other systems to get work done. Their value is substantial in improving output quality relative to the LLM's contribution.
27
8
64
4,506
Founders: what's the one piece of fundraising advice you followed that turned out to be wrong?
Best answers get a coupon for a free campaign match of 20 vc investors and a deck tear down on investormatch.
Best answers get a coupon for a free campaign match of 20 vc investors and a deck tear down on investormatch.
31
4
31
8,834
✅ I'll forgive you for:
Missing a month
A pivot
A down round
❌ I won't forgive you for:
Hiding bad news
A cap table that doesn't match the data room
Learning nothing in 90 days
Pre-seed is a trust business.
Missing a month
A pivot
A down round
❌ I won't forgive you for:
Hiding bad news
A cap table that doesn't match the data room
Learning nothing in 90 days
Pre-seed is a trust business.
11
6
113
5,661
optimization isn't about doing more.
it's about removing everything that doesn't serve the mission.
subtraction is the ultimate life hack.
it's about removing everything that doesn't serve the mission.
subtraction is the ultimate life hack.
8
6
48
2,344
Anthropic's S-1 puts revenue per employee at $6.6M+.
The typical B2B company: $237K.
That gap is rewriting what I fund at pre-seed:
3-person teams that ship like 30
Pricing on outcomes, not seats
A first hire who's an operator, not a manager
Plans to hit $1M ARR before hire #10
Vertical AI over horizontal, every time
Headcount used to be a signal of momentum. Now it's a cost you have to justify.
The typical B2B company: $237K.
That gap is rewriting what I fund at pre-seed:
3-person teams that ship like 30
Pricing on outcomes, not seats
A first hire who's an operator, not a manager
Plans to hit $1M ARR before hire #10
Vertical AI over horizontal, every time
Headcount used to be a signal of momentum. Now it's a cost you have to justify.
8
3
27
3,352
✅ Your pre-seed investors will forgive you for:
Missing your revenue plan by 40%
A pivot in month 9
A bad month, if the update still shows up on time
A cofounder breakup, if we hear it from you first
❌ They will NOT forgive you for:
Going dark for 90 days
A bridge request 3 weeks before payroll
A fudged metric in the deck
Raising your next round without telling us
After 300+ pre-seed checks: it's never the miss. It's the surprise.
Missing your revenue plan by 40%
A pivot in month 9
A bad month, if the update still shows up on time
A cofounder breakup, if we hear it from you first
❌ They will NOT forgive you for:
Going dark for 90 days
A bridge request 3 weeks before payroll
A fudged metric in the deck
Raising your next round without telling us
After 300+ pre-seed checks: it's never the miss. It's the surprise.
13
7
146
10,616
Nailed it
My mom was a hippie. My dad was a beatnik.
I was so poor that at one point I worked two jobs and shared a bed with a friend. I got home at 8 AM from a night shift at a bakery, got into his bed, and slept while he went to his day job.
I bought my first computer with my own credit card at age 20 and taught myself HTML. At the time, I worked nights in a giant, city block-sized printing plant in Queens, NY. I took the 7 train home at 2 AM most nights and was spit on and beaten up for being Irish on that train. Like GD Gangs of New York. I got up at 7 AM and coded websites until 11 AM and went to my printing plant job until the wee hours of the morning.
My first website was featured in a book from Japan in 1997 as one of the 100 best-designed websites in the world. That opened doors and eventually brought me to San Francisco in 2004. And you know the rest of the story. It's on my website.
I never begged for money, expected handouts, or called for government intervention in markets because of my difficult existence when I was young.
I worked so hard my eyeballs fell out. And then I worked even more.
And I succeeded well beyond anyone’s expectations because THIS IS AMERICA.
So yes, I have a problem with politicians looking at what someone earns today and deciding they deserve a bigger piece of it. They see a fattened pig ready for the slaughter. I see the decades of sacrifice it took to build it. Even if it's modest.
Calling it a so-called “millionaires tax” doesn't settle the argument. They already admitted it will apply to everyone. But I want the person cleaning the bakery at 4 AM to build something worth a fortune and keep the rewards. That is what America is about.
Besides, we all pay plenty of tax as it is.
I am voting YES on Initiative 645 to repeal Washington’s new income tax.
I will never apologize for believing everyone should keep more of what they worked for and earned
1
0
9
2,339
Had coffee with one of my LPs yesterday. He said something that floored me:
“ I hired a ceo to run my $70m ecommerce company. Now I just text him and he runs it probably even better than i did. Since I am just texting to run the whole company, I expect to eventually replace the ceo with an agent.”
Man that guy got pilled. That last statement is a huge logic jump that discounts the massive variance between ceos at the end of the text chain. The value of the ceo is all the crap they do that you don’t have to text them about. Ai agents are barely able to do what I ask today it will be a very long time till they can do something better than me without Me even asking.
I love the optimism. The diffusion will take much longer than his current expectation though. That is true across the board. When you live in it you expect rapid diffusion. Ai is diffusing faster than any technology in the history of the world. But it will still take longer than you think
“ I hired a ceo to run my $70m ecommerce company. Now I just text him and he runs it probably even better than i did. Since I am just texting to run the whole company, I expect to eventually replace the ceo with an agent.”
Man that guy got pilled. That last statement is a huge logic jump that discounts the massive variance between ceos at the end of the text chain. The value of the ceo is all the crap they do that you don’t have to text them about. Ai agents are barely able to do what I ask today it will be a very long time till they can do something better than me without Me even asking.
I love the optimism. The diffusion will take much longer than his current expectation though. That is true across the board. When you live in it you expect rapid diffusion. Ai is diffusing faster than any technology in the history of the world. But it will still take longer than you think
37
5
98
11,717