Tech News Radio

Tech News Radio Show Notes - #180 (2026-09-27)

エピソード情報

• エピソード: #180

• 日付: 2026-09-27

• 尺: 約 11 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Former Ukrainian Defense Minister Fedorov pitches a private-sector robot army (https://the-decoder.com/former-ukrainian-defense-minister-fedorov-pitches-a-private-sector-robot-army/)

• DeepSeek Elastic Compute (DSec): A Sandbox Infrastructure for Effective Agentic Training at Scale (https://arxiv.org/abs/2609.22978)

• NarrateAI: production-ready LLM quality assurance on Amazon Bedrock (https://aws.amazon.com/blogs/machine-learning/narrateai-production-ready-llm-quality-assurance-on-amazon-bedrock/)

注目の深掘り

• Deploying real-time personalized speech with Qwen3-TTS on Amazon SageMaker AI (https://aws.amazon.com/blogs/machine-learning/deploying-real-time-personalized-speech-with-qwen3-tts-on-amazon-sagemaker-ai/)

小ネタ・その他

• When chat is the wrong UI (https://github.blog/ai-and-ml/github-copilot/when-chat-is-the-wrong-ui/)

• Two years of OpenAI Academy (https://openai.com/index/two-years-of-openai-academy)

• Two years of OpenAI Academy (https://openai.com/index/expanding-openai-academy-with-new-learning-paths)

Tech News Radio Show Notes - #179 (2026-09-26)

エピソード情報

• エピソード: #179

• 日付: 2026-09-26

• 尺: 約 11 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Scaling MoE reinforcement learning on Amazon EKS with EFA and DeepEP with 40% more throughput (https://aws.amazon.com/blogs/machine-learning/scaling-moe-reinforcement-learning-on-amazon-eks-with-efa-and-deepep-with-40-more-throughput/)

• Accelerate multimodal RL training with SkyRL on Amazon SageMaker HyperPod (https://aws.amazon.com/blogs/machine-learning/accelerate-multimodal-rl-training-with-skyrl-on-amazon-sagemaker-hyperpod/)

• Accelerate multimodal RL training with SkyRL on Amazon SageMaker HyperPod (https://aws.amazon.com/blogs/machine-learning/multi-region-training-with-amazon-sagemaker-hyperpod-and-qumulo/)

• GitHub Copilot app for Beginners: How to build custom workflows with canvases (https://github.blog/ai-and-ml/github-copilot/github-copilot-app-for-beginners-how-to-build-custom-workflows-with-canvases/)

• GitHub Copilot app for Beginners: How to build custom workflows with canvases (https://github.blog/ai-and-ml/github-copilot/github-copilot-app-for-beginners-using-the-diff-terminal-and-browser/)

• GitHub Copilot app for Beginners: How to build custom workflows with canvases (https://github.blog/ai-and-ml/github-copilot/github-copilot-app-for-beginners-run-several-agents-at-once/)

注目の深掘り

• Agentic autofix now uses Copilot Memory (https://github.blog/changelog/2026-09-25-agentic-autofix-now-uses-copilot-memory)

小ネタ・その他

• Another Google Deepmind researcher quits, says building superintelligent AI soon is "inherently irresponsible" (https://the-decoder.com/another-google-deepmind-researcher-quits-says-building-superintelligent-ai-soon-is-inherently-irresponsible/)

• AI was supposed to hit new grads hard. So far, unemployment data says otherwise. (https://arstechnica.com/ai/2026/09/ai-was-supposed-to-hit-new-grads-hard-so-far-unemployment-data-says-otherwise/)

Tech News Radio Show Notes - #178 (2026-09-25)

エピソード情報

• エピソード: #178

• 日付: 2026-09-25

• 尺: 約 12 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• COMED: The Missing Middle Between Routing and Collaboration in Multi-LLM Inference (https://arxiv.org/abs/2609.26913)

• Nexterity wants to automate the hard, dangerous part of pipefitting (https://techcrunch.com/2026/09/24/nexterity-wants-to-automate-the-hard-dangerous-part-of-pipefitting/)

• AI-powered fuzzing with the GitHub Security Lab Taskflow Agent (https://github.blog/security/application-security/ai-powered-fuzzing-with-the-github-security-lab-taskflow-agent/)

注目の深掘り

• Silent Failures in Agent-Tool Interaction: An Audit of ToolUniverse (https://arxiv.org/abs/2609.26836)

小ネタ・その他

• Build a multi-account AI agent with AgentCore Gateway and MCP (https://aws.amazon.com/blogs/machine-learning/build-a-multi-account-ai-agent-with-agentcore-gateway-and-mcp/)

• Claude discovers a novel enzyme system with CRISPR-like repeats (https://www.anthropic.com/news/claude-discovers-novel-enzyme-system)

• Claude discovers a novel enzyme system with CRISPR-like repeats (https://the-decoder.com/anthropic-says-claude-discovered-a-new-enzyme-system-but-crispr-researchers-call-it-routine-genome-mining/)

Tech News Radio Show Notes - #177 (2026-09-24)

エピソード情報

• エピソード: #177

• 日付: 2026-09-24

• 尺: 約 12 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Anthropic says its biology lab has already found something big (https://techcrunch.com/2026/09/23/anthropic-says-its-biology-lab-has-already-found-something-big/)

• Didactic knowledge or Clinical Cases? How Data Types Shape Medical Large Language Models (https://arxiv.org/abs/2609.22161)

• Didactic knowledge or Clinical Cases? How Data Types Shape Medical Large Language Models (https://arxiv.org/abs/2609.25134)

• Startups Modal, Baseten in Funding Talks to Help Businesses Run AI (https://www.bloomberg.com/news/articles/2026-09-23/startups-modal-baseten-in-funding-talks-to-help-businesses-run-ai)

注目の深掘り

• VSCode’s SSH Agent Is Bananas (https://fly.io/blog/vscode-ssh-wtf/)

小ネタ・その他

• Binance Invests $100 Million in Circle (https://www.theinformation.com/briefings/binance-invests-100-million-circle)

• "As a Language Model...": Chat Template Switches LLM Self-Referential Voice and Activation Steering Reproduces It (https://arxiv.org/abs/2609.25021)

Tech News Radio Show Notes - #176 (2026-09-23)

エピソード情報

• エピソード: #176

• 日付: 2026-09-23

• 尺: 約 12 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Snorkel AI triples valuation to $3.5B as demand for AI training data booms (https://techcrunch.com/2026/09/22/snorkel-ai-triples-valuation-to-3-5b-as-demand-for-ai-training-data-booms/)

• Better prompt caching for GPT-6 (https://openai.com/index/better-prompt-caching-for-gpt-6)

• PRQuant: Permutation Residual Quantization for Low-Overhead Inference (https://arxiv.org/abs/2609.22106)

注目の深掘り

• Google Expands Gemini Notebook With Syncing, Interactive Tools and a New Name (https://dev.to/alifar/google-expands-gemini-notebook-with-syncing-interactive-tools-and-a-new-name-1dm)

小ネタ・その他

• Claude Opus 5.5 is now available in GitHub Copilot (https://github.blog/changelog/2026-09-22-claude-opus-5-5-is-now-available-in-github-copilot)

• Parallel cut research time and cost in half with GPT‑6 Astra (https://openai.com/index/parallel-cuts-time-and-cost-with-astra)

Tech News Radio Show Notes - #175 (2026-09-22)

エピソード情報

• エピソード: #175

• 日付: 2026-09-22

• 尺: 約 14 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• xAI’s Grok 4.6 is now available in Amazon Bedrock (https://aws.amazon.com/blogs/machine-learning/xais-grok-4-6-is-now-available-in-amazon-bedrock/)

• ByteDance launches Dramagic, a full-pipeline AI platform for producing short dramas from script to screen (https://the-decoder.com/bytedance-launches-dramagic-a-full-pipeline-ai-platform-for-producing-short-dramas-from-script-to-screen/)

• RBS-Attention: Radius-Bounded Sparse Prefill for Long-Context Large Language Models (https://arxiv.org/abs/2609.20971)

注目の深掘り

• Nscale IPO Files to Go Public, Shows Huge Revenue Jump And Steep Losses (https://www.theinformation.com/briefings/nscale-ipo-files-go-public-shows-huge-revenue-jump-steep-losses)

小ネタ・その他

• How BMW Group detects cost anomalies across 14,000 cloud accounts (https://aws.amazon.com/blogs/machine-learning/how-bmw-group-detects-cost-anomalies-across-14000-cloud-accounts/)

• Sparse Priors for Efficient Distribution Learning (https://arxiv.org/abs/2609.20883)

Tech News Radio Show Notes - #174 (2026-09-21)

エピソード情報

• エピソード: #174

• 日付: 2026-09-21

• 尺: 約 12 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Alibaba's open-weight Qwen-Image-2.1 claims to beat closed models in image generation with just 7 billion parameters (https://the-decoder.com/alibabas-open-weight-qwen-image-2-1-claims-to-beat-closed-models-in-image-generation-with-just-7-billion-parameters/)

• The new AgentCore runtime: Elastic, optimized, and consistently fast starts (https://aws.amazon.com/blogs/machine-learning/the-new-agentcore-runtime-elastic-optimized-and-consistently-fast-starts/)

• Tencent's Gander aims to keep talking while it works in the background (https://the-decoder.com/tencents-gander-aims-to-keep-talking-while-it-works-in-the-background/)

注目の深掘り

• Deploy Hugging Face models on Amazon SageMaker AI with coding agents (https://aws.amazon.com/blogs/machine-learning/deploy-hugging-face-models-on-amazon-sagemaker-ai-with-coding-agents/)

小ネタ・その他

• Why MCP Was Always a Bad Idea (https://maharship.com/blog/why-mcp-was-always-a-bad-idea/)

• Jev TypeSafe AI + LangChain + Vercel AI Gateway (https://dev.to/dcastillogi/jev-typesafe-ai-langchain-vercel-ai-gateway-427c)

Tech News Radio Show Notes - #173 (2026-09-20)

エピソード情報

• エピソード: #173

• 日付: 2026-09-20

• 尺: 約 12 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Unity launches official plugins for Claude Code and OpenAI Codex to stop AI agents from using outdated tutorials (https://the-decoder.com/unity-launches-official-plugins-for-claude-code-and-openai-codex-to-stop-ai-agents-from-using-outdated-tutorials/)

• Qwen3.8-Omni-Flash undercuts Google's Gemini Flash pricing while matching its multimodal benchmarks (https://the-decoder.com/qwen3-8-omni-flash-undercuts-gemini-flash-pricing-while-matching-its-multimodal-benchmarks/)

• Amazon SageMaker Inference: 2026 year-to-date launches in review (https://aws.amazon.com/blogs/machine-learning/amazon-sagemaker-inference-2026-year-to-date-launches-in-review/)

• Amazon SageMaker Inference: 2026 year-to-date launches in review (https://aws.amazon.com/blogs/machine-learning/introducing-amazon-sagemaker-hyperpod-inference-gateway/)

注目の深掘り

• Migrating multi-model AI agents to Amazon Bedrock AgentCore runtime (https://aws.amazon.com/blogs/machine-learning/migrating-multi-model-ai-agents-to-amazon-bedrock-agentcore-runtime/)

小ネタ・その他

• Should you read the code, is RAG dead, and did Skills kill MCP? (https://github.blog/ai-and-ml/should-you-read-the-code-is-rag-dead-and-did-skills-kill-mcp/)

• AI bots "Timmy," "Ren," and "Jackie" are flooding social media with slop (https://arstechnica.com/ai/2026/09/ai-agents-flood-the-internet-with-slop-infused-spam/)

Tech News Radio Show Notes - #172 (2026-09-19)

エピソード情報

• エピソード: #172

• 日付: 2026-09-19

• 尺: 約 13 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Modality Discrepancy Transformer for Ambivalence and Hesitancy Recognition (https://arxiv.org/abs/2609.19148)

• Claude Code changelog - Claude Code Docs (https://code.claude.com/docs/en/changelog)

• Claude Code changelog - Claude Code Docs (https://the-decoder.com/anthropic-keeps-pushing-claude-code-toward-autonomous-coding-with-new-parallel-agent-workflows/)

• Claude Code changelog - Claude Code Docs (https://www.theinformation.com/articles/flaw-found-claude-code-codex-gemini-cli-github-copilot)

• Introducing Kimi K3 on Amazon Bedrock (https://aws.amazon.com/blogs/machine-learning/introducing-kimi-k3-on-amazon-bedrock/)

注目の深掘り

• 既存の LLM が CPU なら、 Jev はその GPU 版みたいなやつ (https://zenn.dev/mizchi/articles/jev-is-gpu-for-llms)

小ネタ・その他

• Introducing Astra for Law (https://openai.com/index/astra-for-law)

• Introducing Astra for Law (https://the-decoder.com/openai-takes-aim-at-the-legal-market-with-astra-for-law/)

• AI hallucination of Chinese nuclear components almost led to US military attack (https://arstechnica.com/ai/2026/09/report-us-almost-boarded-chinese-ship-over-hallucinated-ai-arms-report/)

Tech News Radio Show Notes - #171 (2026-09-18)

エピソード情報

• エピソード: #171

• 日付: 2026-09-18

• 尺: 約 13 分

免責事項

当番組の内容は各情報源の要約です。正確な情報は必ず下記の原文リンクよりご確認ください。誤りがある場合はお知らせください。

出典リンク一覧

本日のトップニュース

• Migrating the GitHub Copilot runtime to Rust, using Copilot (https://github.blog/ai-and-ml/generative-ai/migrating-the-github-copilot-runtime-to-rust-using-copilot/)

• Migrating the GitHub Copilot runtime to Rust, using Copilot (https://github.blog/ai-and-ml/github-copilot/github-copilot-app-for-beginners-using-the-diff-terminal-and-browser/)

• OpenAI caught its models leaving notes to successors to hide bad behavior (https://techcrunch.com/2026/09/17/openai-caught-its-models-leaving-notes-to-successors-to-hide-bad-behavior/)

• OpenAI caught its models leaving notes to successors to hide bad behavior (https://openai.com/index/model-misalignment-reporting-framework)

• OpenAI caught its models leaving notes to successors to hide bad behavior (https://www.theinformation.com/briefings/openai-discloses-safety-incidents-adopts-new-reporting-framework)

• LLMs respond differently to harmful prompts when AI watermarking is used (https://arstechnica.com/security/2026/09/ai-text-watermarking-can-make-models-more-vulnerable-to-adversarial-prompts/)

注目の深掘り

• Enhancing Extubation Failure Prediction with LLM-Derived Features from Respiratory Therapy Clinical Notes (https://arxiv.org/abs/2609.17532)

小ネタ・その他

• The Download: mice with part-human brains and climate tech innovators (https://www.technologyreview.com/2026/09/17/1144314/the-download-mice-part-human-brains-climate-tech-innovators/)

• PrismML — Introducing Bonsai 2 27B: Near-Lossless Compression in a 9x Smaller Footprint (https://prismml.com/news/bonsai-2-27b)