AI Briefing
TLDR AI news
The latest AI and developer news from TLDR AI, with short summaries.
Visit website →Nvidia Forecasts $700 Billion FY28 Revenue, Yet Market Concerns Persist
Nvidia forecasts $700 billion in FY28 revenue growth, but its stock has underperformed the market due to supply constraints and Neocloud uncertainty.
RLVR-based ReViSQL-K2.6 Achieves Human-Level Text-to-SQL Performance Without Scaffolding
The ReViSQL-K2.6 model achieved human-level performance on Text-to-SQL benchmarks without scaffolding through RLVR and reward shaping.
Halo NeuroAI Releases Sopro V2 Turbo, a 120M-Parameter Open-Source Voice Cloning TTS with Native European Portuguese Support
Halo NeuroAI has released Sopro V2 Turbo, a 120M-parameter open-source voice cloning model that natively supports European Portuguese and enables real-time streaming on laptop CPUs.
OpenAI Codex Adds Support for 'Persistent Reasoning Effort' Feature
OpenAI Codex adds a 'persistent' option to the reasoning-effort protocol and supports it in the TUI and SDK.
Stripe Data Analysis: In the AI Era, New Startups Are Dispersing to Smaller Cities, but Frontier Tech Remains Concentrated in Major Metros
According to Stripe data, new startups in the AI era are dispersing to cities with populations under 1 million, yet frontier tech sectors continue to maintain the agglomeration effects of major metropolitan areas.
LEAP Panel Releases AI Industry Outlook: Data Center Investment Surges, Semiconductor Growth Expected to Slow
The LEAP panel forecasts a surge in data center investment and a slowdown in semiconductor growth, while predicting continued revenue growth for OpenAI and Anthropic.
Token usage for Terra and Luna surges up to 13.8x during OpenAI GPT-5.6 discount period
OpenAI's GPT-5.6 discount policy increased token usage for the Terra and Luna models by 5.6x and 13.8x respectively, eroding competitors' market share.
SGLang Achieves Up to 6.24x Speedup in MiniMax-H3 Video Generation Using Cache-DiT and SubBlock
The SGLang Diffusion team accelerated the MiniMax-H3 video generation model by up to 6.24x on eight NVIDIA H200 GPUs using Cache-DiT and SubBlock sparse attention.
Bill Gates Warns AI Could Be 'Worst Driver of Inequality' in 6,000-Character Essay
Bill Gates published an essay stating that the AI transition will be the most disruptive in history, including specific policy proposals such as a robot tax.
Google in Talks for $1.5 Billion Technology License and Talent Acquisition Deal with AI Coding Startup Mechanize
Google is in negotiations to secure technology and talent from AI coding startup Mechanize in a deal valued at $1.5 billion.
Alibaba Previews Qwen4 Architecture: Only 6B of 125B Parameters Active
Alibaba previews the Qwen4 architecture, emphasizing reduced inference costs.
Anthropic Signs $45 Billion Cloud Infrastructure Deal with Nscale
Anthropic has signed a $45 billion cloud contract with Nscale to significantly expand its computing capacity.
OpenAI Adds WebMCP Support to ChatGPT Desktop App's Built-in Browser
OpenAI has added WebMCP support to the ChatGPT desktop app and Sites, enabling agents to directly utilize website tools.
Microsoft Releases AutoSaddler, an Automated Optimization System Based on Agent Execution Trace Analysis
Microsoft has released AutoSaddler, a system that analyzes agent execution traces to automatically update prompts and tools.
Salesforce Expands Anthropic Partnership and Unveils 'Claudeforce'
Salesforce and Anthropic have expanded their strategic partnership and unveiled 'Claudeforce'.
Z.ai Releases GLM-5.3-Flash: Approaching Claude Opus 4.8 with 18B Active Parameters
Z.ai released GLM-5.3-Flash with 18B active parameters, achieving performance close to Claude Opus 4.8 at one-tenth the cost.
NVIDIA Achieves $96 Billion Quarterly Revenue; New Revenue Leadership Shifts from Hyperscalers to Neoclouds
NVIDIA reported $96 billion in quarterly revenue, confirming that the drivers of new revenue are shifting from hyperscalers to neoclouds and enterprises.
Applied Compute Releases AC2, an AI Model Training Platform for In-House Research
Applied Compute has released AC2, an AI model training and deployment platform previously used by its internal research team.
Apple Unveils 2nm M6 and Quad-Die M5 Ultra
Apple unveiled the 2nm-based M6 and the quad-die M5 Ultra, significantly enhancing local AI performance.
Anthropic Integrates and Enables by Default Memory for Claude Chat and Cowork
Anthropic has integrated the memory systems of Claude Chat and Cowork, enabling them by default and introducing real-time recording capabilities.