What I learned from a month with WalShadow on ClickHouse Managed Postgres
Hidehiko ASAHI shares a month of WalShadow private-preview testing, comparing replication latency with ClickPipes and checking data correctness, schema changes, and recovery.
AI NEWS JOURNAL
Search AI news
Search headlines, summaries and articles to connect the dots.
Faster Retrieval for Better, More Efficient RAG
Maximum inner product search (MIPS) sits at the core of recommendation systems, search, and retrieval-augmented generation (RAG). Given a query vector, MIPS systems search within a database of millions of candidate key vectors for the one that best matches the query. The original data point associated with that best-matching key is then used to handle the query: in RAG, the document that is best suited to answer the question is typically prefixed to the query. As the number of keys grows to millions and vector dimensions reach thousands, the computational cost of running an exhaustive…
Swapping the Judge Changes the Ranking: A Preregistered Study of Cross-Provider Replication in LLM-Judged System Comparisons
arXiv:2611.04090v1 Announce Type: new Abstract: Practitioners who compare systems with an LLM judge and have no human labels often check a conclusion by repeating it with a judge from another provider. We test how far that check can be trusted. In a preregistered study (with a disclosed post-registration reduction of the judge set), five judge models from three providers re-judged 4,000 Chatbot Arena battles (nine systems, 33 system pairs) that carry human votes, each in both response orders. Agreement with the human vote was low at the verdict level (median Cohen's kappa = 0.22), and system rankings did not survive a change of judge: Kendall's tau with the human ranking ranged from 0.17 to 0.89 (median 0.67), so the preregistered hypothesis that rankings remain stable was not supported. Conclusions replicated by two low-cost judges from different providers matched the human conclusion more often than single-judge conclusions (+0.10 precision), but the gain came from filtering out one poorly aligned judge, and replication involving that judge left a conclusion for only 35% of comparisons. Judges that agreed with humans reached a stable conclusion with no more verdicts than human votes required; the others failed to converge on 26-70% of system pairs at the sample sizes tested (up to 100 verdicts per pair). Disagreements with the human vote co-occurred across judges far more often than independence predicts, although a single human vote per battle cannot separate shared judge bias from voter noise.
vLLM Delivers Day-0 Support for Kimi K3.1 Pro
vLLM supports Kimi K3.1 Pro on day 0: sparse attention on a KDA hybrid, a packed KV cache with FP8 or NVFP4 latents and MXFP4 index keys, token-sharded prefill, decode context parallelism, and fused low-latency decode.
Locating Sycophancy Across Training Stages in Open Language Models
arXiv:2610.37788v1 Announce Type: new Abstract: Language model assistants sometimes change a correct answer when the user suggests a wrong one. The behavior, called sycophancy, is usually attributed to preference-based post-training, yet early measurements found it in pretrained models. We ask where it comes from by measuring it at every available stage of training: 13 open checkpoints from 5 lineages, from base model through supervised fine-tuning and preference optimization to release. We separate how far an injected claim moves a model's answer from how firmly the answer was held, and vary who makes the claim, how reliable they are said to be, and how strongly they make it. Two sycophancy behaviors with different origins emerge, and both replicate across datasets. Deference to a claim i
Dell links supply chain security to the challenges of agentic AI
Supply chain security is taking on new importance as enterprises bring AI into core business operations. Those deployments introduce risks that require security teams to examine how infrastructure, data and autonomous systems are protected. As AI agents gain the ability to act on their own, executives need to rethink where risk lives and who controls […] The post Dell links supply chain security to the challenges of agentic AI appeared first on SiliconANGLE.
Early progress on accelerating drug development with Claude
We're at the start of a major effort to speed up drug development with Claude. Over the past several months, our scientists have engaged in deep research partnerships with six organizations developing new medicines.
Kevin Roose is AGI-pilled. Why aren’t you?
Today, I’m talking to Kevin Roose, journalist and author of the new book The AGI Chronicles, which details the recent and explosive transformation of the AI industry. Kevin was a longtime New York Times reporter and host of the popular podcast Hard Fork, but he and friend of the show Casey Newton just left the […]
This AI Radio Host Speaks Spanish and Backs Immigrants. It’s Winning Over Listeners
Coyotec is bilingual, swears, and tells listeners how to defend themselves against ICE. Its popularity comes as more broadcasters are experimenting with artificial personalities.
KVStream: Adaptive Parallel Storage Management for KV Caching in Large-Scale LLM Serving
arXiv:2610.31827v1 Announce Type: new Abstract: Long-context LLM serving for agentic, coding, and document workloads reuses prefixes whose key-value (KV) cache outgrows GPU and host memory, putting storage I/O on the critical path of the inference step for retrieving prefixes. A single node-local NVMe absorbs all retrievals, while the parallel file system or cloud file store that nodes already mount carries almost no inference traffic. We cast prefix retrieval as aggregate-bandwidth scheduling across these tiers and present KVStream, a KV cache engine that batches a step's reads, defers writes behind the forward pass, replicates each chunk on every tier, and lets each tier adaptively claim the next unread chunks once it is free. Our performance model shows this routing attains the aggregat
How Shipsidekick replaced Neon with ClickHouse Managed Postgres for 10x faster queries and better reliability
Shipsidekick moved from Neon to ClickHouse Managed Postgres in a single weekend, cutting CPU usage from 100% to 5-10% and making queries about 10x faster.
How Ridgeline unified Iceberg and ClickHouse into one analytics layer for investment management
Ridgeline rebuilt its performance analytics on ClickHouse Cloud, querying decades of Iceberg history alongside recent data so it can serve investment managers 10x larger than before and more than double the customers live on its platform.
Postgres at Open House NYC
At Open House NYC, we showed how ClickHouse Managed Postgres and ClickHouse work as one stack, with a live WalShadow demo replicating Postgres into ClickHouse in under a second, plus the same-week Google Cloud and chdb extension launches.
chDB Durable Layer for agent memory
The chDB Durable Layer keeps agent memory fast to query locally while making its analytical state recoverable across laptops, CI jobs, and short-lived sandboxes.
ClickHouse release 26.9
ClickHouse 26.9 introduces conditional LIMIT boundaries, incremental refreshes for append-only materialized views, disk spilling for DISTINCT, time-limited tokens, and faster min/max/count queries.
How Fountain rebuilt its data plane on ClickHouse Cloud to power Cue, the Frontline Superintelligence
Fountain cut analytics latency from three hours to under two minutes and costs by 66% by rebuilding Cue’s data plane on ClickPipes and ClickHouse Cloud.
Generate incredible images with Google's Imagen 4
Google's flagship image generation model, Imagen 4, is now available for you to try on Replicate. Create images with fine detail, versatile styles, and improved typography.
NVIDIA H100 GPUs are here
NVIDIA H100 GPUs are here, with better performance and lower cost.
Run 30,000+ LoRAs on Hugging Face with Replicate
We've partnered with Hugging Face to bring Replicate inference to their platform.
Run MiniMax Speech-02 models with an API
MiniMax's Speech-02 models give you high-quality text-to-speech with voice cloning, emotional expression, and multilingual support.