Aru Labs' Lossless-memory is a local, file-based long-term memory layer for a personal AI that deliberately refuses to summarize: every conversation turn is stored verbatim with an ISO-8601 timestamp in per-day JSONL logs, and all indexes are rebuildable from those raw logs. The design rests on three pillars: a lossless raw log (seven-field records: ts, actor, role, type, text, model, session), a Temporal Backbone that treats time as the primary axis (time-aware query parsing narrows ranges before ranking, returning chronologically ordered verbatim lines), and LLL, a tiny human-maintained "where are we now" topic index injected into the model each turn so context survives window compaction. Search prefers exact SQLite FTS5 matches (bigram tokenization for Japanese/English) and uses sqlite-vec semantic fallback only when needed.
Implementation specifics: ingestion produces indexes incrementally with a background re-index daemon, and the system ships examples, docs, and a quickstart; it has run daily for one user since mid-2026. Operational metrics show major improvements after fixes (exact-index rebuild 40s → 1.24s; vector store 2.54GB → 337MB). Limitations are explicit: single-user/single-machine, Japanese-first relative time parsing, no benchmarking, and local embedding model requirements. The stated goal is a companion AI that remembers exact words and timing on hardware you own.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.
AI-generated writing often appears polished initially but becomes repetitive and unnatural over time. Human-authored content, especially concise prompts, retains character and clarity, making it more engaging to read. (ssp.sh)
OpenAI is likely to replicate Jev's innovative classification capabilities by leveraging its existing large language models. This could enable OpenAI to integrate advanced classification features into its models, potentially surpassing Jev's offerings. (arcturus-labs.com)
N-Operator Correlators (NOC) are introduced as a generalized method for measuring quantum information spreading, extending beyond traditional two-operator techniques. These concepts are demonstrated through simple quantum systems at finite temperature, providing a foundational understanding of higher-order operator correlators in quantum mechanics. (arxiv.org)
OpenAI has fired multiple contractors for using AI to train its models, despite the company's focus on AI-assisted training. Some contractors admit to frequently using AI tools, risking model degradation and highlighting internal contradictions in OpenAI's approach. (404media.co)
Large language models encourage dependency by automating coding and information retrieval, reducing active learning and engagement. Companies design these tools to maximize reliance, which hampers human understanding and learning processes. (jola.dev)
JevBench is a benchmark developed by Benchmark Heaven to evaluate Jev-class decision models based on typed answers to 534 decisions. It ranks 52 systems using the JEVBENCH SCORE, focusing on reproducibility and detailed performance metrics. (benchmarkheaven.com)
Today's best Hacker News stories, summarized and screenshotted, one email a day.