🔥Telegram İfşa

is there a non lab ai company with better pmf than cursor? if anything i feel like cursor breaks the where are the young successful people that Patrick Collison has been talking about afaik even the last few yc batches have lower avg ages than years past

HOLY FUCK notebooklm’s interactive mode is fucking goated jesus christ how have i not heard about this before you have to try it now, it is genuinely incredible

we spent an hour plus with roommates talking to a paper with notebooklm’s interactive mode just deep diving every question we had very casual sf tuesday night

when people know you too well 🥲🥲

so sad that i’m getting on a 22hr flight and can’t play with what will definitely be new bestie immediately but so excited whatever i tried looks so good also love that we bullied them is 3.7 as the name and not 3.5 (new) (new) or some shit

mini conspiracy theory - yc really hates cursor it is a company they would've funded before or maybe they passed or something it is explains the 500 cursor competitors, explains the windsurf shilling - i don't they ever once talked about cursor this way

always fun to wake up and see everything broken cause your db’s server ran out RAM

i feel like i want to talk to this model so much more - blown away by the quality also honestly if we are going to linkedin spam this seems like a better way to do it

my friend got his o1 and now will actually move into my closet help Garry Tan

i have starlink on my plane!!! so excited for this to become normalized soon

he said ‘touch grass’ so bro brought his laptop to the sand.

While the Sam (from Wendover) and team do this in NY, Rahul Tarak and i are going to host a Jet Lag S15 Watch Party in San Francisco Who wants to attend?

our closet to o1 pipeline is strong

one of my angel investments decided he needs to build a new city to get the data he needs for his models i’m so excited about him founding a new city

one observation i was sharing with friends, the quality of software seems to have gone of a cliff compared to last cycle i don’t want to call out anywhere here but it just seems worse even in “high craft” spaces like devtools three potential theories - llm codegen - speed of