A developer built ai-slop-linter, a zero-dependency Node 20+ CLI and browser tool that flags stylistic tells of AI-generated writing in READMEs and commit messages. The linter ships twenty rules — eig…
A training-free framework combining reasoning and retrieval over Wikipedia improves multilingual multimodal entity linking by 6.9% overall and up to 23.3% on rare-entity slices on the MERLIN benchmark…
Recursive self-improvement (RSI) is a hypothesized process in which artificial general intelligence (AGI) systems rewrite their own computer code, potentially triggering an intelligence explosion that…
The fx2-cmix-transformer program achieved the top ranking on the Large Text Compression Benchmark, compressing the first 10^9 bytes of the English Wikipedia XML dump (enwik9) to 96,996,198 bytes, with…
Polign, a new vector database that separates index storage from compute, stores all 12.5 million English Wikipedia passages (81.7 GiB) in an S3 bucket for $2.28 per month, and a $55 spot instance hand…
A developer built P.U.C.K., an AI agent that measures uncertainty in language models using logprobs and self-reported confidence, and tested it against Wikipedia ground truth with tools for weather an…
A developer's benchmark comparing RAG against long-context prompting on 12 Wikipedia articles about space missions found that long-context answered 96% of questions correctly (23/24) using about 14.5%…
A developer spent two days adding a source-location axis to Vodou's memory store, enabling it to track where each memory came from via source_url, source_host, and source_ref columns, plus a four-tier…
A new AI memory system achieved #1 on the BABILong benchmark at 10M tokens, with a gaming audit showing 24/24 accuracy at 1M tokens and 16/16 at 2M, 5M, and 10M tokens on RULER variable tracking. The …
A Wikipedia editor added an 'Irrelevant' tag to the article 'X-kom', citing that it is an LLM-generated article with self-contained Polish references that limit verification by non-Polish editors. The…
A programmer spent about $260 in AI tokens to jailbreak an abandoned Amazon Fire tablet, and AI is being used to repurpose off-patent drugs for new diseases, illustrating a new 'cognition surplus' tha…
Cosine similarity, the default metric in most RAG pipelines and vector databases, discards vector magnitude, which can carry critical semantic information such as confidence, specificity, or negation,…
Open-source and commercial 'unslop' tools that rewrite AI-generated text to sound human are emerging as a new software category, with at least three MIT-licensed projects on GitHub, a pip install unsl…
A developer built a full-duplex real-time voice assistant that runs entirely in the browser on WebGPU, using jax-js, with all stages (speech, ASR, LLM, TTS, and optional vision) processed locally. The…
China's massive data scale gives it a fundamental advantage in AI development, according to a new analysis, as the sheer volume of digital interactions, sensor data, and industrial logs from its integ…
IPullRank's Q3 2026 zero-click and paid-click analysis found that ChatGPT had the highest share of paid clicks among leading Google destinations, with about 4.75% of its Google traffic coming from pai…
Jimmy Wales discussed his new book, 'The Seven Rules of Trust', at the Edinburgh Book Festival on August 18, 2026, sharing stories from Wikipedia's early days and his views on AI's impact on the platf…
AI companies are quietly buying up large quantities of physical books in the UK, a trend that signals the exhaustion of public internet data and a shift toward 'dark data' to improve large language mo…
A new daily tracker, 'Who is using AI?', shows ChatGPT leads in 84 of 85 tracked countries, capturing 74.4% of global AI-tool search interest as of July 19, 2026, with Gemini second at 8.7% and Kimi t…
WikiPaw, an interactive Wikipedia-hopping game centered on dog breeds, uses Gemini AI to guide players from a starting Wikipedia page to a target breed page. The game calculates semantic proximity bet…