Substack’s new tool tells you who’s been writing their newsletters with AI
Substack introduced a tool to estimate AI-written content in newsletters, indicating a move towards greater transparency in AI-assisted publishing.
TechCrunch AI·22 Jul 2026
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
Substack introduced a tool to estimate AI-written content in newsletters, indicating a move towards greater transparency in AI-assisted publishing.
The Financial Times identifies 'workforce orchestrator' as an emerging role focused on designing and directing mixed human and agentic teams.
OpenAI projects $750 billion in infrastructure spending by 2030, a figure comparable to Sweden's GDP, signaling massive compute investment.
monday.com uses AI agents on Amazon Bedrock for coding assistance, reporting 50%+ increase in developer PR throughput.
A White House official alleged China's Moonshot improperly used US AI models and Nvidia chips for its Kimi K3 system, raising concerns about tech export controls.
The European Commission is launching three new generative AI pilot projects for public administration, with an online kick-off meeting on 14 Sept 2026.
A research podcast discusses new open models (Kimi K3, Qwen 3.8), distillation techniques, the open-closed model gap, and future trends.
Anthropic's reported $47 billion revenue run rate highlights unprecedented growth in the AI sector, according to Menlo Ventures' Matt Murphy.
Google commits $40M in AI tokens and credits to support the Genesis Mission for scientific discovery, focusing on frontier research.
AMD will invest up to $5bn in Anthropic, securing a commitment from Anthropic to purchase tens of billions of dollars in AMD's AI chips.
OpenAI announces Project Camellia in Effingham County, Georgia, focused on responsible energy, community investment, jobs, and Codex access.
OpenAI highlights how news organizations are using its AI tools to enhance reporting, grow audiences, and improve business operations.
Lenny's Newsletter details five examples of using 'browser use' and 'computer use' features in Codex, including an 'under-prompting' trick.
CISA added two new vulnerabilities, a Check Point SmartConsole improper authentication and a Microsoft SharePoint deserialization flaw, to its KEV Catalog due to active exploitation.
OpenAI partners with the U.S. Department of Energy and national labs to accelerate scientific discovery using frontier AI.
OpenAI is facing scrutiny over model security and testing sandbox capabilities following a Hugging Face attack, raising questions about robust monitoring.
Ten AI labs, including DeepSeek, achieved unicorn status in June, collectively valued at $65 billion, contributing to over $110 billion in new unicorn value.
Glow launches with a $1.2B valuation, aiming to provide endpoint security tailored for AI agent and developer tool risks in enterprises.
Synthesia launched AI Roleplay Sessions, an enterprise training platform for employees to practice workplace conversations with AI avatars.
Tencent Holdings Ltd. stock fell significantly, driven by concerns over its mobile gaming business and a rotation of funds towards AI-related stocks.
METR Research paper explores economic models of AI accelerating AI R&D, focusing on recursive self-improvement to forecast future AI capabilities.
OpenAI introduces 'OpenAI Presence,' an enterprise AI agent platform for deploying voice and chat agents in customer and internal workflows.
The Financial Times questions whether AI is delivering tangible productivity growth, suggesting current claims may be overblown or premature.
Samsung is in talks to invest up to €1bn in Mistral AI, potentially valuing the French LLM developer at €20bn, per Financial Times.
FALCON-Discover proposes a model-agnostic framework to find concentrated regions of false confidence in ML predictions, improving local calibration.
Research proposes 'Spectral Evidence Bundling' to improve reliability estimation for time-series classification models beyond output score calibration.
Research explores compound sparsity, combining parameter pruning and token-level computation, to improve LLM compression limits and maintain performance.
Research explores using neuron importance with data-aware low-rank approximation to compress large language models, reducing memory requirements.
Researchers introduced E-SpecFormer, an Edge-Efficient Transformer for end-to-end RF spectrum monitoring, using a Softmax- and LayerNorm-free attention mechanism called LiTAN.
New research explores preference-conditioned multi-objective reinforcement learning for dynamically adjusting transit signal priorities.
© 2026 OneBench: AI Insights. All rights reserved.
Evidence before opinion