HopperInput's profile picture. Learning about cryptography/ML/AI/robotics/energy/physics. 
I use this channel to listen. Not much to say beyond occasional sh*tpost. Meowf.

input hopper

@HopperInput

Learning about cryptography/ML/AI/robotics/energy/physics. I use this channel to listen. Not much to say beyond occasional sh*tpost. Meowf.

Pinned

Being precedes naming.

HopperInput's tweet image. Being precedes naming.

input hopper reposted

EMV effectively locks down NFC on all point of sales terminals to traditional card payment rails. Terminal by @useburner upends this with true, native tap to pay for stablecoins. Excited to have @CastleIslandVC @placeholdervc @inflectionxyz @1kxnetwork @seedclub onboard.

Arx Research raises $6.1 million seed round to launch 'Burner Terminal' stablecoin and fiat point-of-sale device theblock.co/post/377876/ar…



input hopper reposted

Channa Horwitz, 1 Canon Twelve, Moiré Number 2. 1985. Casein on mylar.

Pixel0Symphony's tweet image. Channa Horwitz, 1 Canon Twelve, Moiré Number 2. 1985.

Casein on mylar.

goblin town


input hopper reposted

A decent summary of the 4xPRO6000 vs 8x5090 challenge. We definitely need a betting market so you can play along at home.


market overview, 2025


feline brain slowly realizing we're in a Pareto cascade not 80/20, rather 20 of the 20 i.e. 96/4 wylde tymes await (thanks to @nearcyan for the image--did you clean up the dd mound?)

HopperInput's tweet image. feline brain slowly realizing we're in a Pareto cascade

not 80/20, rather 20 of the 20 i.e. 96/4

wylde tymes await

(thanks to @nearcyan for the image--did you clean up the dd mound?)

"ew, you live across the street from a primary school and have to hear children all the time?!" yes, I live across the street from a primary school and get to hear children all the time


input hopper reposted

"Mirror Mirror" is the best ST episode to ever light up a CRT.

HopperInput's tweet image. "Mirror Mirror" is the best ST episode to ever light up a CRT.
HopperInput's tweet image. "Mirror Mirror" is the best ST episode to ever light up a CRT.
HopperInput's tweet image. "Mirror Mirror" is the best ST episode to ever light up a CRT.
HopperInput's tweet image. "Mirror Mirror" is the best ST episode to ever light up a CRT.

ahhhh I'm de-cringing more free bookshelf space!

HopperInput's tweet image. ahhhh I'm de-cringing

more free bookshelf space!

more cringe recycling feels like redemption

HopperInput's tweet image. more cringe

recycling feels like redemption

retroactive cringe-audit of my bookshelf the bin awaits

HopperInput's tweet image. retroactive cringe-audit of my bookshelf

the bin awaits

input hopper reposted

LLMs are recommending products every day, shouldn’t you know if you’re one of them? If you want to see how often LLMs mention your product (vs competitors) in real queries, check out my new distraction. It runs prompts like “best X tools”, parses model answers, and tracks…

nibzard's tweet image. LLMs are recommending products every day, shouldn’t you know if you’re one of them?

If you want to see how often LLMs mention your product (vs competitors) in real queries, check out my new distraction.

It runs prompts like “best X tools”, parses model answers, and tracks…
nibzard's tweet image. LLMs are recommending products every day, shouldn’t you know if you’re one of them?

If you want to see how often LLMs mention your product (vs competitors) in real queries, check out my new distraction.

It runs prompts like “best X tools”, parses model answers, and tracks…

input hopper reposted

btw if you liked DeepWiki, you're gonna love what's coming next for repos like this. trying to see what we can open source too.

swyx's tweet image. btw if you liked DeepWiki, you're gonna love what's coming next for repos like this. trying to see what we can open source too.

Excited to release new repo: nanochat! (it's among the most unhinged I've written). Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single,…

karpathy's tweet image. Excited to release new repo: nanochat!
(it's among the most unhinged I've written).

Unlike my earlier similar repo nanoGPT which only covered pretraining, nanochat is a minimal, from scratch, full-stack training/inference pipeline of a simple ChatGPT clone in a single,…


input hopper reposted

Wow, language models can talk without words. A new framework, Cache-to-Cache (C2C), lets multiple LLMs communicate directly through their KV-caches instead of text, transferring deep semantics without token-by-token generation. It fuses cache representations via a neural…

jiqizhixin's tweet image. Wow, language models can talk without words.

A new framework, Cache-to-Cache (C2C), lets multiple LLMs communicate directly through their KV-caches instead of text, transferring deep semantics without token-by-token generation. 

It fuses cache representations via a neural…

الرزق على الله، والمؤمن لا يُحرم نصيبه.


input hopper reposted

"it happened again, sir"

DipWheeler's tweet image. "it happened again, sir"

"good morning sir, you got liquidated again"

DipWheeler's tweet image. "good morning sir, you got liquidated again"


What's the problem fam? NVDA is fine.

HopperInput's tweet image. What's the problem fam? NVDA is fine.

bulls die in front of everyone bears die alone


Loading...

Something went wrong.


Something went wrong.