Sarah Guo

.@zackmziegler cofounder CTO @openevidence talks to @niki4conviction on how unhinged recruiting practices selected for folks who would push boundaries in Conviction's new podcast that explores building legendary teams -- "Good Company" https://t.co/WK5BKnutBu First 3 episodes at https://t.co/T0QubFwkEm

wow do coding agents show us how repetitive programming was before

Deadline for @conviction Embed ($250K grant, compute and tools credits, community and mentorship program for the 10 best AI startups) extended to Monday 11:59PM PT. As is tradition, here are some of startup ideas we've been pondering and cooking -- apply if you are too! https://t.co/kpoftevR6H https://t.co/jIKHXfHK68 more ideas at https://t.co/tXfsp49ism

the one in which @eladgil and I discuss the SaaSpocalypse, who should have all that market cap, growth vs. durability, software abundance, and slop-in-production https://t.co/BBYyGZmWrU full episode of @NoPriorsPod on YouTube or anywhere you get your podcasts https://t.co/f5T4nKJMF7

One of the strongest predictors of startup success is the ability to attract exceptional talent. At @Conviction, our Talent Team helps the next gen of category-defining companies build the talent density that wins markets. Excited for @niki4conviction to launch "Good Company" -- podcast where we talk to founders and operators about what it takes to recruit, scale and outperform.

finally settled the question of what @harvey's named for turns out you can just recruit the partners you want

AI Naive: use agents to try to solve the problem. AI Native: use agents to fix the missing data and scattered context that make the problem hard.

I have so many assistants I need a general manager

friends become trolls in the age of ai https://t.co/hQFETSbN7k

hi! if you want to work at some of the most interesting startups in AI, come to our next mixture-of-experts event at @conviction and hear from the founders of Harmonic, Worktrace, Sunday, Treehouse and Trajectory https://t.co/ZL7mdnszUl more about it https://t.co/ALq9mZY19O

“The compute used in our post-training of Composer 1.5 even surpasses the amount used to pretrain the base model.”

Very hard without enough observation to have intuition for when something is directionally true but temporally false I am sure the models will improve on this dimension, as every other

🧑‍🍳 💋

the divergence of opinion in how robotics plays out is one of the biggest money making (and career making) opportunities in AI

In order to have a sense of AI adoption right now, you really need to have one foot at the bleeding edge with the kids Wispr-ing through a mic at Devin all day and one foot in the 99% enterprise world where they’re still debating the merits of leaving “bundled-MSFT-ELA-Copilot”

fan of mamoon hamid. great investor, human, leader -- look forward to finding another board with him

Code is suddenly cheap! What’s scarce is deciding what software should do—and under what constraints. I wrote down some thoughts about why engineering leverage is shifting away from typing code and toward making intent scalable

Congrats to Embed team Phylo! Really impressive to see how quickly domain specific agents are advancing also, apply to Embed https://t.co/dlbc5yjpXS

@auralix4 Finally

watching tiny startup teams voice prompt their way to wild and viral launch videos makes it clear the returns to good ideas and taste are higher than ever https://t.co/I38Mzm1799

CRM Was Built for Humans. Day was Built for AI. Why We Invested In Day.AI Why CRM Stops Working as Companies Scale Software markets are in a prolonged drawdown. Public valuations increasingly reflect the belief that even deeply embedded Systems of Record may be vulnerable to disruption from AI. We at @conviction agree with that premise: meaningful disruption will come from companies that architect for AI, and companies bolting chatbots onto legacy designs risk missing the mark. CRM is a clear place where this matters. Ask most CEOs, CROs, or CFOs what their CRM does and the answer is usually some version of "it holds the pipeline." In pra…

Most “AGI” panic isn’t about capability. It’s ego-scaled evaluation: if a model does *my* hard thing, it must be AGI and the world has changed. Intelligence feels shocking when it trespasses on identity.

“good enough but faster” models are going to eat much of the existing token share this year

👀 probably nothing

what do you think has better results for contribution to a country’s scientific base: enhanced support for 100K kids who show signs of genius or… killing basic math / gifted programs at the altar of “equity” https://t.co/BIZOEjOjMl

Ok well there are quite a few signs of takeoff

in the end the only approval that matters is that of your customers 💫

when you text a founder you work with and you just get a clawdbot autoresponse

the best deal and community in AI

battleground shifting in many domains from intelligence --> inference