Char K
837 posts

Char K
@char_helps
Stealth AI data infra startup 1 exit, 1 unicorn
Silicon Valley Entrou em Temmuz 2008
32 Seguindo291 Seguidores
Char K retweetou

@albrgr False - Ant's real revenue is ~20-30% of that.
Ant reports *gross* revenue (vs. OAI: *net* revenue). For example, for every $100 token sales, they give ~80% to AWS, so net revenue is actually $20.
This is before deducting any costs of good sold.
theinformation.com/newsletters/de…
English
Char K retweetou

@realEstateTrent Good one - clearly people don’t realize what today is :)
English

@warikoo @bryan_johnson X users aren't higher net worth -- a subset of X users are (tech+finance) HNW, but most X users are conspiracy theorists, fringe politics, etc.
English

My take:
1.X is a niche that feels like a mass platform
It’s heavily skewed toward tech + finance. Step outside that bubble and its real-world footprint drops sharply.
2.X users are higher net worth and lower intent to engage
The higher the perceived status, the lower the likelihood of approaching a public figure. Ego plays a role.
3.X is cognitively intense, not widely pervasive
It demands more opinion, attention, and emotional energy. That creates the illusion of scale, when in reality it’s depth over breadth.
English
Char K retweetou
Char K retweetou

model+harness is now over model only. agent perf can be significantly influenced by the design and quality of harness. i do believe this is a right direction, nice work!
Yoonho Lee@yoonholeee
How can we autonomously improve LLM harnesses on problems humans are actively working on? Doing so requires solving a hard, long-horizon credit-assignment problem over all prior code, traces, and scores. Announcing Meta-Harness: a method for optimizing harnesses end-to-end
English
Char K retweetou

Char K retweetou
Char K retweetou
Char K retweetou
Char K retweetou



















