Aivizor Community

Two ready — to-run Java applications built on the AgentScope Java Harness — Claw for single — user local development and Builder for multi — tenant team deployments — are available as example projects in the agentscope

A June 2, 2026 article by Radhika Rajkumar highlights Lance Whitney’s technique: Give a chatbot a brief image idea and ask it to expand that into a detailed, generator‑ready prompt to improve inputs for image tools.

Alphabet is launching an $80 billion capital raise to expand AI computing infrastructure, including a $10 billion private placement from Warren Buffett's Berkshire Hathaway.

Speaking at a Commonwealth Bank of Australia conference last week, OpenAI CEO Sam Altman said his earlier warnings about rapid automation of office roles were premature, conceding he expected more elimination

General Motors is using AI and machine learning to merge design, aerodynamics, structures and manufacturing simulations into a single probabilistic workflow.

Alibaba’s Qwen team has published Qwen3.7 — Plus to the Bailian console (internationally Model Studio) and opened agentic, multimodal APIs to external developers for integration and testing.

JetBrains released Mellum2 on June 1, 2026: A 12‑billion‑parameter Mixture‑of‑Experts model trained from scratch on natural language and code that activates 2.5B parameters per token and is available under the Apache 2.

JetBrains published Mellum2 weights under the Apache 2.0 license and released six checkpoints covering pretraining, context expansion, instruction — tuning and RL tuning, including distinct Instruct and Thinking

TSMC announced at NVIDIA GTC Taipei that it is integrating NVIDIA accelerated computing and AI across chip design and fab operations to accelerate lithography, transistor and process simulation, advanced process control

Banks, payments firms and fintechs are replacing siloed, task-specific models with transformer — based transaction foundation models trained on billions of events.

At I/O 2026, production teams used Gemini alongside internal generative models — including Nano Banana, Gemini Omni and experimental DeepMind systems — plus custom tooling in Google AI Studio to create a short film,

MiniMax launched M3 on June 1, 2026, introducing MiniMax Sparse Attention (MSA) to support a claimed 1,000,000‑token context while accepting native image, video and desktop compute inputs.

Mike Sewell says architecture, engineering, and construction firms should treat generative AI as a tool that refines professional judgment and communication rather than as a shortcut to full automation.

Created in 1985 by George Stone, Annabel Jankel, and Rocky Morton, Max Headroom was a glitching, prosthetic‑clad synthetic personality for the series Max Headroom: 20 Minutes into the Future;

Unveiled at GTC Taipei, Cosmos 3 is an open omnimodel that combines vision reasoning, multimodal world generation and action prediction to accelerate physical AI development.

At GTC Taipei, NVIDIA said it is expanding its DRIVE Hyperion platform and enlisting automakers, suppliers, AV developers and mobility operators to speed global Level 4 robotaxi rollouts across Asia, Europe

OpenAI has reconstituted its robotics effort after closing the unit in 2020, restarting physical systems work in January 2025 and hiring across hardware, operations, systems engineering, and machine learning while

On June 1, 2026, Richard Sutton argued that ordinary generative models cannot perform genuine scientific discovery because they lack the ability to evaluate and selectively retain their own outputs;

AAII scored NVIDIA’s Nemotron 3 Ultra at 48 points, the highest AAII intelligence score among open models from U.S. organizations; tests on DeepInfra showed throughput above 300 tokens/second.

NVIDIA unveiled the RTX Spark superchip at Computex — a roughly 1‑petaflop PC CPU designed to run local AI agents and LLMs on Windows laptops and desktops, with major OEMs pledging fall systems and Microsoft

In a June 1, 2026 post by Nicholas Fuller, IBM Research argues that explicit “agent logic” — structured software primitives that steer LLMs-is necessary to run reliable, cost‑effective AI inside dynamic, long‑running

AgentCore Identity now accepts ARNs of preconfigured AWS Secrets Manager secrets for Outbound credential provider resources, letting teams reuse managed secrets and retain encryption, rotation, tags, replication,

A new walkthrough shows Amazon Quick Research using LLMs to ingest, index and synthesize biomedical data for pediatric sarcoma research, producing cited, versioned reports.

OpenAI frontier models GPT-5.5 and GPT-5.4, and the Codex coding agent, are now generally available on Amazon Bedrock.