Darshan sk

Co-Founder @wishtalesai | Always up for a conversation!

[Darshan sk @Darshan_skumar]

@ola_supports Will do!

[Darshan sk @Darshan_skumar]

Anyone else noticing cab apps are reusing the same OTP for you every single time??
@Olacabs @Uber

[Darshan sk @Darshan_skumar]

I bet I can walk to the restaurant have dinner and walk back before @Swiggy delivers my order.

[Darshan sk @Darshan_skumar]

@jonathanbrnd Both of them working as co founders.

[Isabel🌻 @isabelunraveled]

The older i’ve gotten, the more I realize every lesson is just a derivative of ‘trust your intuition’

[gaurav @gaxrav]

all your heroes were misfits, and here you are, trying to fit in.

[Rohit Mittal @rohitdotmittal]

Sam Altman said something this about YC once "if your hit rate as a vc is high, you're doing it wrong"

in vc and startups, you're not supposed to be right most of the time, you're supposed to take big swings on seemingly impossible companies - the crazy ideas that could change the world if they work

yes, you can make money investing in b2b saas that does slightly better expense tracking. nothing wrong with that

but the best vcs are backing founders who sound delusional to most people. rockets, brain interfaces, fusion energy, whatever

the math works because one crazy bet that hits pays for 50 that don't

a high hit rate with a small number is a lot of luck, but with a high number, it's not venture risk/return

portfolio construction people talk about it like they are investing in public markets

they've turned venture capital into something it was never supposed to be. venture was created to fund the unfundable. the stuff banks wouldn't touch. the ideas that seemed insane until they weren't

now we have vcs bragging about their "disciplined approach" and "portfolio construction" like they're running a hedge fund.

this results in founders with genuinely world-changing ideas get passed over because they can't show a clear path to 3x in 5 years. meanwhile the safe bets get funded, return 2-3x, and everyone pats themselves on the back

this is what most of the venture industry that's long on returns and short on actual innovation

the vcs who changed the world - the ones who backed apple, google, spacex - they looked stupid at the time. that's the point. if it was obvious, it wouldn't need venture capital

if you want a high hit rate, go buy index funds

[Darshan sk @Darshan_skumar]

The most accurate chart on coffee.

[Simon Sarris @simonsarris]

tell them the truth

[gaurav @gaxrav]

it’s funny how wisdom works. you aren’t always ready to receive it. it sounds stupid, borderline retarded, so you laugh at it and ignore it.

until one day, when you've lived enough life, seen enough things, and suffered enough pain, you finally remember it. and then it clicks.

[Darshan sk @Darshan_skumar]

@kmr_dilip I would pay 75k-1Lac one time. Would be happy to give regular feedback.

[Wholesome Side of 𝕏 @itsme_urstruly]

Adult money hits different 😂👏🔥

[Darshan sk @Darshan_skumar]

@_adishj mosaic

[Darshan sk @Darshan_skumar]

@elonmusk Step 1: Buy a sh*tload of GPUs
Step 2: Fund Wishtales AI for GSD
Step 3: Profit

[Darshan sk @Darshan_skumar]

@techhalla You can do all of this on one platform.

Introducing @Wishtalesai

[All day Astronomy @forallcurious]

Amazing NEWS🚨: Scientists have discovered a fungus in the Amazon rainforest that naturally consumes plastic and could help save the planet

[Saurabh Kumar @drummatick]

i don’t think you understand

you train a small model on scikit-learn

it gets 80% accuracy

you fine-tune a huggingface transformer

it gets 90% accuracy

you build a custom data pipeline

it gets 93% accuracy

you ensemble three models with voting

it gets 95% accuracy

you throw in feature engineering and SHAP

it gets 96% accuracy

you replace it with a gradient boosting tree

it gets 97% accuracy

you build a custom loss function for your niche problem

it gets 98% accuracy

you realize the labels were wrong

fix them, retrain

it gets 99% accuracy

you make a tiny architecture change

it beats human baseline

you remove half the parameters

it runs in real-time

you quantize the model

it runs on a potato

you distill it

it fits in a JavaScript bundle

you deploy it to edge

it autocompletes before users even type

you realize

you haven’t written a paper

but you built god

i don’t think you understand

[Rowan Cheung @rowancheung]

Elon Musk posted on X that Grok 4 is planned for ‘'just after July 4"

xAI engineer Tim Li said the intelligence of the new model (previously dubbed Grok 3.5) will be “unmatched”.

Musk previously said it will reason from first principles

[Elon Musk @elonmusk]

Generate charts with @Grok!

[Darshan sk @Darshan_skumar]

@primal_brainer One of our VCs confidently stated, “Distribution eats product for breakfast. So build distribution first.”

This changed the way I look at entrepreneurship.