I recently created a fully #opensource SDK for autonomously organizing data in hierarchical structures.
https://t.co/ge3nrdRTfo
It is designed to be fully modular with the ability to have human/programmer in the loop at any stage of the process.
#AI#OpenAI#MachineLearning
@nic_detommaso this flips the timeline i was operating on. i'd assumed you start talking to investors once you're actually raising. so what does picking them early look like in practice, before there's a round on the table or anything you need to ask for yet?
@hpierrejacques the 8-year arc is the part i keep thinking about. you only really know who someone becomes by watching them over time. as a founder starting to raise, i'm curious how you compress that at seed, when you're betting on a person years before the arc is visible.
@blakeir work stripped of friction and paired with visible progress is also a pretty good description of the products people actually keep using. as a seed investor do you find you can feel that quality in a founder's early build, or does it only show up once real users are in it?
@AmberIllig@noxmetals the "they wanted the failures" bit is the interesting part. most founders ask for the playbook and get the highlight reel. curious what actually transfers when you share a failure vs what just stays your story. that gap is on my mind as i start talking to funds.
@alanaagoyal@ninklefitz@AmDroste@tacitlabsco@OpenAI the "missing verification loop for biology" framing is the part i keep rereading. as a founder early on a raise, i'm curious how you get conviction on a research-lab bet that early, when the loop is the whole thesis and there's no product to point at yet.
@fintechjunkie when speed is abundant the scarce thing is judgment, that line keeps me thinking. founder early on a first raise here, and it reframes it all: the machine makes everyone fast, so what's left to underwrite is taste. curious how you read that this early, before any track record.
@polynoamial@OpenAI I think the point was to keep the thinking private, that being said reasoning was going to emerge soon with our without OpenAI everyone was already doing chain of throught so it was just natureal for the models to absorb it.
@ashleymayer the who wins if you win bit is the one i underestimated. i'm early on a first raise, kept building the deck around my own growth story. curious, when a founder gets that wider narrative right in a meeting, can you feel it in the room or does it only show up later?
@jordanrcrook@betaworks the agentic economy framing is the part i keep circling. i'm a founder early on a first raise, genuinely curious what the camp actually screens for this early, when half the applicants will say agentic and the product barely exists yet. what separates a real one for you?
@LeeJacobs@2pasc the "well before it was cool" part is the whole job, no? i'm a founder early on a first raise and that pre-consensus conviction is what i keep trying to understand. curious what told you it was real that early, before the market caught up.
@joshbuckley the "makes the old way feel broken" test stuck with me. funny it took ai to make research, the thing meant to de-risk everything else, look broken. founder early on a first raise, building near here. how do you tell who actually sees a shift this new from who's just riding it?
@jordanodinsky the hard part of selling the outcome is proving it before the customer trusts the agent with the workflow. been building toward exactly that gap. curious whether the proof has to show in usage data or whether founder conviction carries it at pre-seed.
@nikipez the "finally technically feasible due to ai" bar is the part i keep chewing on. the harder call now feels less whether something can be built and more whether it stays durable once everyone can build it too. as a founder early on a raise, curious how you weigh that at rfs stage.
@shreyansalecha funny, the founder feels the same in the other chair. you walk in sure you have to build this, then one sharp question makes you wonder if you see it clearly or just want to. maybe the doubt on both sides is the tell it's real. what separates the founders you bet on?
@eladgil the "research" part of that shift is the one i sit closest to. founder early on a first raise, building in how product research gets re-run instead of done once. genuinely curious: when the ground moves this fast, what do you anchor an early bet on that doesn't shift?
@kevinhartz the "led the seed, doubling down in the A" arc is the conviction loop i keep trying to understand. building noemica (synthetic users for product research), early on a first raise. when you back someone this early, how much of the bet is the person vs the wedge being legible yet?
@lennypruss composable primitives applies to how products get validated too. teams still test in one monolithic research round, not clean re-runnable signal. founder here, early on a first raise in that gap. curious what makes an infra bet legible to you this early?
@michuk the cockroach side is the one i keep thinking about. you can read whether a founder inspires people in one meeting, but resilience only shows after the first thing breaks. as a founder before a first raise, curious how you spot the never-die trait before it's tested.
@rexsalisbury clarifying, thanks. i'm a founder early on a first raise and have been overthinking the wrapper. if personalization reads like slop, what actually earns a reply from you in a cold note, the traction, the problem itself, or how the founder thinks?
@anamitra agents that scale a team without headcount is the cleanest pre-seed story right now, and also the easiest to demo and hardest to verify early. as a founder before a first raise im trying to learn how you tell a real one from a good demo at pre-seed. what do you look at?