patrick

178 posts

patrick banner
patrick

patrick

@patrikkml

🇵🇭

Katılım Temmuz 2014
799 Takip Edilen118 Takipçiler
patrick retweetledi
m
m@skitzocat·
if you're so smart why are you retarded
English
136
652
4.3K
115.7K
patrick
patrick@patrikkml·
last thing you see before you get laid off
patrick tweet media
English
1
0
3
98
Colossus
Colossus@colossusmag·
This is the story of Hyperliquid, the most profitable startup per employee on earth, told from a guarded office in Singapore. Last year, its team of 11 generated $900 million in profit. It's 3 years old, has never taken a dollar of venture capital, and is beginning to change how century-old markets work. Its founder, Jeffrey Yan (@chameleon_jeff), had never taken a physics class when he picked up a textbook at 16. Two years later, he won gold at the International Physics Olympiad. In 2019, he started trading with $10,000 from a living room in Puerto Rico—working off a television because he didn't own a monitor. Within 3 years, he was running one of the largest anonymous crypto trading firms. Then he shut it down. Yan was rich and free, but he had spent years inside crypto, watching it betray itself. Bitcoin's central premise was decentralization. Yet the biggest exchanges were centralized. Crypto kept reintroducing the dependence on trust it was built to eliminate. He set out to create what should have existed. Hyperliquid is a blockchain with a trading exchange on top, and anyone can build on it. Yan's vision is to house all of finance. In 3 years, it has done over $4 trillion in volume. And in the past few months, it has begun to outgrow crypto. Markets for oil, silver, and the S&P 500 now trade on Hyperliquid around the clock, weekends included, and are growing roughly 40% week on week. When the US and Israel bombed Iran on a Saturday in February, Hyperliquid was the venue traders turned to. Hyperliquid's success has cost Yan his freedom. He works out of a secret office in Singapore and cannot travel without two bodyguards. Even the team's housekeeper doesn't know what they do. In January, @domcooke spent a week at their office. Read his profile on Yan and @HyperliquidX below.
Colossus tweet media
English
412
1.1K
5.2K
3.5M
arman
arman@ksw_arman·
how am i supposed to get work done when it looks like this outside in sf
arman tweet media
English
5
1
19
1.5K
patrick
patrick@patrikkml·
this tax stuff way too hard I think i’m just gonna go to jail
English
2
0
6
105
patrick
patrick@patrikkml·
@kylecompute @yimothysu clear visibility into the traces of exploration and decision lineage from the agent
English
0
0
5
96
kyle
kyle@kylecompute·
Auto-Hypothesis beats Karpathy’s baseline auto-research in 10x less time Auto-research is a repository that gives an AI agent an H100 SXM and tells it to optimize LLM training autonomously. The system ran 125 experiments to reach its best result; ours took 13.
kyle tweet media
English
4
2
17
1.2K
patrick
patrick@patrikkml·
we are quickly moving towards a future where agentic systems enable scientific breakthroughs. we have this intuition that real innovation comes from nowhere. that a genuinely new idea has no precedent. but that's not how scientific progress has ever worked. most breakthroughs come from taking existing knowledge in one domain and applying it to another. john maynard smith took game theory - originally developed by von neumann for economics - and applied it to animal behavior, creating evolutionary game theory. it completely changed how biologists model competition and cooperation between species. agentic systems work the same way. and as they get better at searching, iterating, and running experiments autonomously, the rate of those recombinations accelerates. that's why this matters. not because AI will replace the scientist, but because it will find the connections the scientist didn't have time to look for.
kyle@kylecompute

Auto-Hypothesis beats Karpathy’s baseline auto-research in 10x less time Auto-research is a repository that gives an AI agent an H100 SXM and tells it to optimize LLM training autonomously. The system ran 125 experiments to reach its best result; ours took 13.

English
0
0
7
214
SpaceSquid
SpaceSquid@sfspacesquid·
just saw a leaked interview of my boy failing 2sum in the mercor leaks
English
31
64
2.6K
141.1K
patrick
patrick@patrikkml·
@vussyviz pure uncut obsession is the sweetest drug
English
0
0
1
21
Vis (not in SF :((( )
Vis (not in SF :((( )@vussyviz·
there was a period of time where i'd starve myself to work like, i'd set certain goals and I wasn't allowed to eat until then it really got the pressure up.. lowkey might be the move again
English
2
0
3
147
patrick
patrick@patrikkml·
"who needs an idea, i have a domain" - @yimothysu
patrick tweet media
English
2
1
10
548
kyle
kyle@kylecompute·
AGI
kyle tweet media
5
0
12
384
patrick retweetledi
patrick
patrick@patrikkml·
hyperliquid
Dansk
2
1
6
240