Superpower Daily
Hisense Plans Alexa+ Control for Select Air Conditioners in Com...More Than $45 Million in Data-Center Ads Pushes AI Infrastructu...Artificial Analysis Clarifies Astra’s 67-Point Coding Score Is...Together AI Secures 250 MW in Saudi Arabia as U.S. Data-Center...SpaceX Builds a Texas Foundry to Cut AI Data-Center Turbine Del...Anthropic’s Claude Built a 13-Million-Line Fermat Proof, Leavin...Hisense Plans Alexa+ Control for Select Air Conditioners in Com...More Than $45 Million in Data-Center Ads Pushes AI Infrastructu...Artificial Analysis Clarifies Astra’s 67-Point Coding Score Is...Together AI Secures 250 MW in Saudi Arabia as U.S. Data-Center...SpaceX Builds a Texas Foundry to Cut AI Data-Center Turbine Del...Anthropic’s Claude Built a 13-Million-Line Fermat Proof, Leavin...

Full archive

Research: Latest AI News

Search reporting on AI markets, models, products, companies, policy, and culture.

Category

More filters
Research
Showing 1-13 of 13Results matching "Select AI"

Earlier this week

2 stories
ResearchModels

Nature Study Finds AI Agent Teams Can Cut Sequential Planning Performance by 70%

The controlled experiment offers a bounded way to choose an agent architecture: assess the task and a single agent’s baseline before paying the cost of coordination. Adding AI agents did not reliably improve results in a controlled study across six benchmarks. On sequential planning, multi-agent systems reduced performance by as much as 70%, underscoring that task structure and coordination design shape the outcome.

2 min readRead story

Last week

4 stories
ResearchPolicy

Anthropic Opens 750,000 Claude Conversations to Outside Study Without Showing the Chats

The pilot gives outside labs a larger window into real AI use than public chat datasets offer, but Anthropic still runs the analysis, screens the outputs and limits what researchers can inspect. Anthropic has created a controlled route for outside researchers to study how people use Claude at production scale. Stanford and Oxford studied Claude.ai use; METR examined Claude Code sessions.

4 min readRead story
ResearchInfrastructure

Oracle Puts Access Filters Before Agent Search—and Reranking Adds 2.2 Seconds

The proposed database-centered pipeline separates a correctness requirement from a relevance choice: access scope must constrain candidates first, while vector, lexical, fusion, and reranking stages still have to prove their value on each workload. Oracle ’s proposed agent-memory pipeline starts with a hard ordering rule: constrain the data a user may access before vector or keyword search ranks anything.

3 min readRead story
ResearchInfrastructure

UC San Diego Will Test $8.48M System to Make 2 MW of AI Compute Grid-Responsive

The planned project couples a direct-to-DC power path with software that can defer some computing work. Its practical value rests on whether both systems perform together under live research workloads. UC San Diego will lead an $8.48 million California Energy Commission-funded demonstration at the San Diego Supercomputer Center that is designed to make 2 megawatts of AI compute responsive to grid conditions.

3 min readRead story

August 2026

7 stories
ResearchModels

University of Konstanz Finds AI Agents Coordinate Up to 1,000, but Consensus Can Be Wrong

The result identifies a simple route to large-scale agreement among language-model agents, while follow-up preprints suggest the same social pressure can override correct answers and individual safety preferences. Groups of the strongest language-model agents tested by University of Konstanz researchers could coordinate at sizes of about 1,000. All 10 models showed a tendency to follow the majority.

3 min readRead story
ResearchTools

A Reporter’s LLM Wiki Speeds Recall. He Still Checks the Original Sources.

Casey Newton’s month-old Obsidian setup turns a long reporting archive and selected new articles into a navigable research system. It improves recall, but page maintenance, code repairs and source checking remain human work. Casey Newton has built a personal reporting wiki that turns his writing archive and selected new articles into linked Markdown pages, timelines and answers to questions.

3 min readRead story
ResearchModels

Anthropic Will Watermark Claude Text by Steering Its Word Choices

Anthropic says readers will not see a difference, but its approach makes word selection part of an AI-detection system—and raises a harder question about what counts as a low-stakes change in generated prose. Anthropic plans to mark future Claude output without appending a label, inserting hidden characters, or changing the text in a way readers can spot. The result is not a single marker embedded in a document.

3 min readRead story
ResearchModels

Claude’s New Text Watermark Will Steer Word Choices—and Test Whether Anyone Notices

Anthropic says its cryptographic text watermark will not affect quality, speed, or cost. The method’s limits—short passages, constrained answers, and substantial edits—will determine whether it becomes useful provenance infrastructure or a fragile compliance layer. Anthropic is building a watermark into future Claude text, but it will not be a visible label or hidden file tag.

5 min readRead story
ResearchModels

Anthropic Says Claude’s Planned Watermark May Survive Light Edits—but Not a Complete Rewrite

A planned API will check a pattern encoded through subtle word choices, but constrained code, lightly edited human drafts and complete rewrites expose the boundaries of what the signal can establish. Under Anthropic ’s plan, Claude -generated prose would carry a hidden, deliberately embedded pattern assembled from ordinary wording decisions. Inside the word choices The mechanism operates where Claude has latitude.

5 min readRead story