tim cook to nvidia is just a polymarket rumor rn but it makes weird sense
nvidia's problem was never vision. it's who gets chips first
and nobody alive runs a supply chain like cook
2
3
79
i was wrong
an hour before the wsj story dropped i replied that nobody was ever firing jensen
my prediction career lasted 60 minutes xitter.twitterapi.xyz/foundrceo/status/21206…
Replying to @theinformation
a symbolic pay cut is the most board move possible. nobody was ever firing jensen
1
2
157
google shipping a model that can't write is the most builder friendly launch this week
half of what an agent does is pick a tool or say yes or no. never needed an essay for that
36
everyone on my timeline tonight is hiring ai teammates
half of them still haven't landed their first customer
1
50
every vibe coding argument on here is seniors saying it cant do production and some kid replying with his stripe dashboard
🤖 Made with AI
2
63
a founder in my replies today shared their first month on google play
842 people saw the app. 29 downloaded it.
that's not a reach problem. 800+ people already found it, the listing just didn't sell them
fix the screenshots before paying for more eyeballs
2
6
284
the shops that ignored the first one spent years invisible on maps
same thing is happening again, just with a chat box instead of a pin
🤖 Made with AI
4
161
talked to a founder in my replies this week who can't code at all
ai wrote his whole swift app. it runs on his iphone, widgets and a live activity included
it's still not on the app store
ai made the building part easy. the $99 dev account, testflight, review and the privacy label are what's left, and nobody vibe codes those
8
8
363
signups who never answer your email usually aren't ignoring you. the email just looked like it came from a platform
the replies under my last post turned into a better playbook than the post, so writing it down:
1
1
80
1. plain text, no logo, no footer, one question they can answer in a single line
2. send it at a weird time. 9:00am looks scheduled, 4:47pm looks like a person who just thought of them
1
1
54
spent today in the replies of 32 founders building something
one did all the seo basics and still has zero customers. another has people signing up and never answering a single email
neither of those gets fixed by shipping another feature. and that's exactly the first thing most of us reach for
4
8
248
last week i posted about 24,664 lovable apps leaking their database and the replies taught me more than the post did.
two traps kept coming up:
- a policy written as using (true) means rls is "on" and still lets everyone read everything
- rls on with no policy returns nothing, the app looks broken, so people switch rls back off
so here's the whole check in one image. takes about 2 minutes, do it before you post your launch
7
3
167
last week i posted about 24,664 lovable apps leaking their database and the replies taught me more than the post did.
two traps kept coming up:
- a policy written as using (true) means rls is "on" and still lets everyone read everything
- rls on with no policy returns nothing, the app looks broken, so people switch rls back off
so here's the whole check in one image. takes about 2 minutes, do it before you post your launch
1
75
claude fable 5.2 is out, anthropic says 30% faster and 20% cheaper than 5.1, and the timeline is about to be 48 hours of benchmark screenshots.
the founders who come out ahead this week aren't testing it on riddles. they already have 10 prompts from their own product saved and can answer "do we switch?" before lunch.
if you don't have that list yet, today's the reason to make it
1
2
120








