building a way for strangers to prove they picked you, not sprayed a list. ex Oracle, 3x Ironman
šØš¦Canada š USšŗšø
Joined August 2010
- Tweets7.2K
- Following1.3K
- Followers433
- Likes8.1K
I want strangers in my house.
Open the door to everyone ā noise.
Filter by pedigree ā keep meeting the same kinds of people.
Soon their agents will be knocking, too. They'll need a way to make that first approach.
The way out šsabledima.substack.com/p/theā¦
š¤ Made with AI
about warm intro mechanism (will drop a full essay soon): an āintroducerā is a court built into the fact-delivery channel. We need to separate these two functions:
1)a Ā«notary" handles the delivery of facts, while the court remains with the recipient.
2)the Ā«notary" allows a stranger to bring their own time with them- a certified story is.
in fact, condensed years of observation, recorded not by the person it is about.
Getting angry at people is like yelling at a thunderstorm.
Itās silly. Because people are part of nature.
After all, we donāt get angry at rain or snow, do we?
so itās silly to get angry at people. It doesnāt change anything. And above all, youāre the one who looks silly.
save our energy for something more important.
And thereās a lot more thatās important than it seems at first glance.
Sometimes, finding yourself among the overwhelming majority also means to pick up a winning lottery ticket.
i believe, that future agent-2-agent intent layer systems should collect the data about actions done and their provenance, - not reputation.
because provenance means "what actually happenedĀ», meanwhile reputation is about what weāve decided it means.
i see many ppl started using Pangram. that means theyāre tired of ai slop. understand . But they dont solve the core problem.
ok,imagine if there was an ideal version of āpangram' with zero false positives (today itās not), it answers the question: āwho is likely to have WRITTEN (typed) these words?ā. And only.
!!!meanwhile it does not ask those Q-s needed:
*Did this person really CHOOSE me specifically?
**Did they make a special EFFORT to reach me?
***Could they REPEAT this action for e.g. 100 people at (almost) zero cost?
noticed one (paradox) thing:
people who are the 1st to intellectually recognize the breakdown of the old norm often initially try to restore the old norm, rather than change the mechanics of the system
any ideas - why?
Is missing an alpha-deal crucial for VC? Yes:
7k investments (1985 to 2014), 6% of deals made 60% of the total return.
So the cost of failure is determined by the TAIL, not the AVG
in a business governed by a power-law distribution, the cost of failure is the tail
Nothing worth comes to your open inbox for free today sabledima.substack.com/p/notā¦
this guy's A & B are bingo!
Iāll add one more - "C": whoās forcing people to say this lie like āI respond to everyoneā? Why do people lie on purpose? in fact,no one even asked them like, āWhy arenāt you open to everyone?ā or āAre you really responding to everyone?ā
I have no why's - do you have any ideas?
Iāve concluded that anyone more popular than me who claims to personally answer every DM is either:
A) Lying
B) Operating with more hours in the day than the rest of us plebs
So Iāve been getting ~150 DMs/day lately and reading/answering them properly takes about an hour
To be fair, itās been +EV
Several genuinely valuable conversations for the startup started in my inbox, and Iām very grateful for that
But the less elegant truth is that most messages, however well-intentioned, simply arenāt very useful to me personally
I know it sounds harsh, but I have to be honestā¦
Iāve still replied to almost everyone because ignoring someone who took the time to write felt disrespectful in a way
I always liked being the guy you could actually reach, and Iād like to think I was in part known for repeatedly doing so!
But all good things come to an end, I canāt keep up, the arithmetic simply no longer works
I donāt have the time, and I refuse to outsource my inbox to an assistant, agent, or AI pretending to be me
Iff I reply, it will always be me
If I donāt, please donāt take it personally
Send me something interesting enough and Iāll always try to make the time, it might take a while, but Iāll reply
I genuinely am sorry
That said, Iām certain there are many people far more capable than I am who are generous with their time and would be happy to engage with you!
Thank you for all the conversations weāve had over the years. Iāve genuinely cherished many of them, and quite a few of you have become people I now consider friends
So thank you <3
Ppl physically live today, moreover thinking they are living in the future, but in fact living by yesterday and the day before yesterday thinking models. We adapt too slow to how External conditions transform. E.g.:
1/accelerators are not that necessary today as they used to be 2-4 yrs ago, but we automatically count them so that valuable.
2/Or take subscription business model - it is becoming more obsolete vs pay for result model, but ppl continue using it.
3/Moreover, applying it to llm based products, which is incorrect, since tokens cost 80/20% ratio and the math doesn't work.
4/As a consequence - valuations are counted completely wrong for them. but we still multiply arr by 8-12x avg factor we used to get an estimation of SaaS products who had a flip economics of 20/80 cost/margin.
Human says (and think)they are fast but they're not.
ai filters can't beat ai slop and why the chess game is a good (bad) example? sabledima.substack.com/p/ai-ā¦
VCās intake form is the best ai slop generator but they cry about it(1 min to read this short post) sabledima.substack.com/p/vcsā¦
the research:
pnas.org/doi/10.1073/pnas.24ā¦
arxiv.org/abs/2509.00462
1) @paulg recommends not attacking a frighteningly idea head-on, directly- but rather approaching it from the side;
2) while solving how strangers meet online with my own product, I solved a wider version of a proof of human problem - human-owned - as a by-product.
