The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 6 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Ramaswamy: LLMs Can Replace Massive Click Data for Core Search Tasks
“We realized that a lot of problems that Google solved with massive scale and user data, they could, in fact, solve with LLMs. So we use a lot of them for things like query rewriting.”
Ramaswamy: Flash Let Neeva Solve Index Replacement Faster Than Google
“Asim, for example was just brilliant at engineering a system that ran completely on Flash in which we could do things like super rapid iteration, replace the entire index or the space of two days or put in arbitrary amounts of information for experimentation i…”
Ramaswamy: Small Language Models Generalize HTML Parsing and Scraping
“All of us have nightmares about writing beautiful soup code in order to parse web pages. It's basically regular expression parsing over ever-changing websites. It's horrible. We have done a bunch of it in the first two-ish years of Neva. That stuff is also eas…”
Ramaswamy: European Consumers Care More About Privacy, Aiding Neeva
“For us, for example, we were surprised that we did so much better in Europe compared to the United States, and you don't really think of them as being that different, but in practice, in terms of how many people care, it is actually very different.”
Ramaswamy: Neeva Built In-House Summarization Over OpenAI for Scale
“We decided that we didn't really want to be beholden to say like using OpenAI's API. For doing things like summarizing a four billion page index. We built a lot of the technology in-house”
Ramaswamy: Neeva's First Three Years Taught Harsh Lessons on Consumer Readiness
“So the first three years of Neva were really about building a better private search engine. And honestly, it also taught us a lot of pretty harsh lessons about consumers and you know, whether they were ready for change or not.”