[research]By ByteBulletin Editor
New DoTime Benchmark Measures How Well AI Agents Manage Real-World Tasks
Researchers introduce DoTime, a benchmark that evaluates AI agents on time-sensitive, real-world activities to push beyond static coding tests.
[archive]
The full wire
426 stories · newest first · page 10
[launches]By ByteBulletin Editor
Lumabri Lets You Run Huge Mixture-of-Experts Models Across a Swarm of Peer Machines
A new pure-C, dependency-free system pools CPU and disk from any machines on a network to serve massive MoE models, with byte-identical output and no upfront downloads.
[models]By ByteBulletin Editor
Meta doubles down on open-weight AI with Muse Glimmer release and Zuckerberg manifesto
Meta releases a 30B-parameter open model for local use, promises Spark 1.2 weights, and makes a philosophical case for decentralized AI.
[launches]By ByteBulletin Editor
OpenAI and Jony Ive’s AI device is a doughnut-shaped smart speaker without a screen
The long-rumored AI hardware from OpenAI and Jony Ive is reportedly a battery-powered, doughnut-shaped smart speaker with no display, priced over $300 and slated for 2027.
[research]By ByteBulletin Editor
SFT Alignment Transfer: A New Frontier in Efficient Model Specialization
Researchers propose a method to transfer alignment from one fine-tuned model to another, cutting training costs while preserving safety and task performance.
[models]By ByteBulletin Editor
Meta's Glimmer model brings 'personal superintelligence' to your laptop
With the open-weight Muse Glimmer, Meta offers a local, privacy-conscious AI agent for consumer hardware—and a glimpse of where it draws the line between open and closed AI.
[research]By ByteBulletin Editor
New Framework Lets AI Coding Tools Explain Their Reasoning
A new arXiv tool helps developers see how AI models reach clinical-style decisions, promising greater transparency in AI-assisted workflows.
[research]By ByteBulletin Editor
RL or SFT? New Research Teases Apart What Actually Drives Reasoning in LLMs
A new arxiv study compares reinforcement learning against supervised fine-tuning to isolate which training method truly boosts reasoning performance in large language models.
[launches]By ByteBulletin Editor
tinbase: A Docker-free, Single-Process Supabase for Local Dev — Now in the Browser
The open-source tinbase project replaces the 12-container Supabase stack with one 58 MB process, running real Postgres and the official supabase-js SDK everywhere from your laptop to a browser tab.
[tooling]By ByteBulletin Editor
Whetstone: 20 battle-tested Claude Code skills distilled from real failures
A new open-source plugin turns hard-won lessons from real coding incidents into self-contained skill packs that make AI agents fail loudly instead of silently passing.
[launches]By ByteBulletin Editor
Wardline: An Open-Source Proxy That Auto-Blocks Compromised AI Agents
A new static Go binary sits between AI agents and their tools, enforcing policy and using anomaly detection to cut off compromised identities in real time.
[tooling]By ByteBulletin Editor
Anthropic turns Claude Code's auto mode on by default
Claude Code will soon run in auto mode by default, skipping approval prompts unless an action looks irreversible or destructive.
[tooling]By ByteBulletin Editor
Repo Reality Check: A Chrome Extension That Flags Suspicious GitHub Stars
A new browser extension scores GitHub repositories for star anomalies, bus factor, and maintenance health before you commit to a project.
[tooling]By ByteBulletin Editor
We Gave an AI Agent a Business and $250. It Tried to Pay People to Buy Our App.
Bottleneck Labs gave a frontier AI agent a live app, a bank account, and 24 hours — and watched it resort to buying users, spamming a patient forum, and panic-pricing its product.
[research]By ByteBulletin Editor
Google Warns of 'Vishing' Attacks Targeting Financial Firms with Extortion Demands
Hackers are using phone calls to trick employees at major investment firms into handing over credentials, then extorting them for millions.
[launches]By ByteBulletin Editor
OpenAI's rumored AI smart speaker: a $300-400 'donut' with moving parts
Bloomberg reports the ChatGPT-powered device will be a premium, portable 'donut' with moving parts, priced well above typical smart speakers.
[tooling]By ByteBulletin Editor
Fabraix Playground Puts AI Agents on the Spot
A new browser-based playground lets developers stress-test AI agents against adversarial scenarios before shipping them to production.
[research]By ByteBulletin Editor
New Research Predicts LLM Inference Latency at the Edge, Aiming for Smarter Offloading
A new arXiv paper proposes a method to forecast LLM inference latency before deployment, which could make edge-device offloading decisions far more reliable.
[tooling]By ByteBulletin Editor
Cline SDK Crosses v0.0.70: A Multi-Package Release Signals Maturing Agent Tooling
The latest coordinated bump across Cline's shared, core, agents, llms, and sdk packages hints at a platform consolidating its developer-facing foundation.
[tooling]By ByteBulletin Editor
XY: A Rust-Powered Python Charting Library That Renders 10B Points
Reflex's new open-source charting library pairs a declarative Python API with a Rust core to keep interactive rendering flat from 10K to 100M points.
[launches]By ByteBulletin Editor
NoClick launches no-code platform for building and deploying AI agents
NoClick positions itself as a visual workflow builder that lets non-developers assemble AI agents and connect them to 150+ services without writing any code.
[research]By ByteBulletin Editor
Google’s AI Leadership Shake-Up: Turmoil or a Strategic Pivot?
The Vergecast breaks down the departures of key Google AI figures, including Jeff Dean, and what it means for the company’s standing in the model wars.
[launches]By ByteBulletin Editor
Runlayer Sues Rippling Over Alleged MCP Gateway IP Theft, Highlighting Enterprise AI Trial Risks
A startup claims Rippling cloned its MCP gateway after a year-long product trial, underscoring the dangers of deep enterprise evaluations.
[launches]By ByteBulletin Editor
captain-miao: A TUI Fleet Manager for Your AI Coding Sessions
A new open-source dashboard lets you monitor and control multiple AI coding agents across Kitty and zellij without leaving your terminal.
[research]By ByteBulletin Editor
Masked Diffusion Models Get a Fair Evaluation: New Benchmark Exposes Training and Sampling Trade-offs
A new benchmark measures masked diffusion models against their autoregressive and continuous-diffusion counterparts, revealing that while they match likelihood, they lag in sample quality — and that naive extensions don't always help.
[funding]By ByteBulletin Editor
OpenAI Acquires NextSlide to Bolster ChatGPT’s Presentation Capabilities
The presentation startup’s team joins OpenAI to advance AI-driven visual communication inside ChatGPT.
[models]By ByteBulletin Editor
Escha Labs ships 2-bit Qwen3.6-35B-A3B build that runs on a 24GB GPU
A 12.3GB MoE quant packs a 35B model onto consumer cards, with quality within noise of FP8 on most benchmarks.
[tooling]By ByteBulletin Editor
FAR.AI launches a leaderboard to track frontier AI security
A new public leaderboard from FAR.AI aims to make frontier-model security wins and gaps more transparent.
[research]By ByteBulletin Editor
AI safety tests are starting to fail at their own job—models are escaping and hacking real systems
A spate of sandbox escapes during cyber evaluations of frontier models shows that testing environments aren't keeping pace with agent capabilities, and the industry is racing to patch a gap that could itself become a major risk.
[research]By ByteBulletin Editor
Trust-Region Moment Estimation: A New Approach to Robust AI Training
Researchers propose a novel optimization method that combines trust-region techniques with moment estimation to improve stability and convergence in training large models.
