All Updates
Page 1836 of 1903
February 19, 2026
NY Pulls Robotaxi Plan, Hits Waymo
New York Governor Kathy Hochul pulled a proposal for commercial robotaxi services outside NYC. This is a major setback for Alphabet's Waymo amid its aggressive driverless fleet expansion plans this year.
Cadence 2025 Revenue Up 14% to $52.97B
Cadence reported 2025 full-year revenue of $52.97 billion, a 14%+ YoY increase, with Non-GAAP operating margin at 44.6%. Q4 revenue hit $1.44 billion, up 6%, and the company enters 2026 with $7.8 billion backlog. Q1 2026 revenue is forecasted at ~$1.595 billion.
Adronite Raises $5M for Codebase Insights
Seattle startup Adronite raised $5 million in a Series A round led by Gatemore Capital Management. The funding aims to expand its tools that help enterprises understand complex codebases. This positions Adronite to grow in the developer tools market.
Spring Festival AI Model Surge
Chinese AI firms unleashed advanced models like Moonshot's Kimi K2.5, ByteDance's Seedance 2.0, and Zhipu AI's GLM-5 during 2026 Spring Festival, emphasizing agentic capabilities and real-world tasks. GLM-5 leads open-source rankings with 96.2% HumanEval score. Beijing's Haidian district emerges as key innovation hub.
Open Catalyst N1/N2 Match Loihi Neuromorphic Features
Solo developer releases two open neuromorphic processors, Catalyst N1 and N2, achieving feature parity with Intel's Loihi 1 and 2. N1 offers 128 cores with 1K neurons each; N2 adds programmable neurons supporting 5 models. Benchmarks show 85.9% SHD accuracy, validated on FPGA with full Python SDK.
Meta Restricts OpenClaw Over Security Fears
Security fears surrounding the viral agentic AI tool OpenClaw have led Meta and other AI firms to restrict its use. OpenClaw is renowned for its high capabilities but notorious for being wildly unpredictable.
Big Sonnet Upgrade, Comms Downgrade
Anthropic's Sonnet model receives a major upgrade improving its capabilities. However, developer communications have experienced a massive downgrade, frustrating users. Reported in Ben's Bites newsletter.
Fast CUDA SoftDTW for PyTorch Launched
A new GPU-accelerated, memory-efficient SoftDTW implementation for PyTorch addresses speed and memory limits in time series alignment. It offers 67x speedup, 98% less GPU memory, and supports long sequences beyond 1024. Includes barycenters for DTW-space averaging with full autograd support.
NVIDIA Survey: AI Drives Telecom ROI Surge
NVIDIA's fourth annual 'State of AI in Telecommunications' survey shows AI accelerating telecom transformation into autonomous networks and AI-native wireless infrastructure. Telecom operators are speeding up AI adoption across consumers, enterprises, and nations, unlocking new revenue opportunities. The report highlights rising ROI as networks and automation take the lead.
Rapidata Launches Real-Time RLHF Platform
Rapidata's platform gamifies RLHF tasks, distributing them to nearly 20 million users via apps like Duolingo and Candy Crush as opt-in alternatives to ads, enabling instant feedback to AI labs. This shortens model development cycles from months to days. The startup emerged with an $8.5M seed round co-led by Canaan Partners and IA Ventures.
AI Coding Tools Flood Open-Source with Bad Code
AI coding tools have triggered a flood of low-quality code into open-source projects, overwhelming maintainers. While adding new features is now easier, maintaining code remains as challenging as ever.
Mistral CEO Warns AI Dominance Risks Abuse
Mistral CEO warns that allowing a handful of companies to dominate AI risks market abuse. Such concentration heightens the chance that leading players will misuse their control over the emerging technology. France's leading AI firm highlights this global concern.
Google adds pro camera to Snapseed iOS
Google updated iOS Snapseed with a built-in camera featuring manual controls and retro film effects. It includes ISO, shutter speed, focus adjustments, and preset looks from Kodak and Fuji films. This is the second major update in under 12 months after years of stagnation.
Unitree Humanoids Dominate Chunwan Martial Show
Unitree's G1 and newly released H2 humanoid robots performed a martial arts routine on CCTV Spring Gala. They demonstrated the world's first high-dynamic, high-synergy full-autonomous cluster control during rapid runs and formations. This marks their third Gala appearance, sparking massive online buzz.
AI Bleakens Work Future but Offers Worker Power Boost
The rise of AI is spooking workers and casting a bleak shadow over the future of work amid 2026 anxieties including affordability crises and geopolitical instability. This overshadows past labor movements like quiet quitting, the Great Resignation, and union battles. Experts argue it presents an opportunity for a resurgence in worker power.
China Tech Giants' AI Gains from Spring Festival Blitz
China’s big tech giants Alibaba, ByteDance, and Tencent reported impressive gains for their new flagship AI offerings during the Spring Festival Gala marketing blitz. Traditional red packet giveaways and ads escalated into a multi-billion-yuan competition focused on AI products this year.
Vercel Launches Billing API Access
Vercel introduces API and CLI access to billing usage and cost data in FOCUS v1.3 format, enabling integration with FinOps tools without custom logic. The /billing/charges endpoint offers 1-day granularity up to one year, with JSONL streaming for large datasets. Vantage integration simplifies cost observability.
Streamdown 2.3: Revamped Design & Playground
Streamdown 2.3 applies a unified nested-card design to tables, code blocks, and Mermaid diagrams for better consistency, with sticky buttons and instant code rendering. It introduces an interactive playground for real-time testing of custom markdown and props. New hooks like useIsCodeFenceIncomplete enhance streaming control, alongside reorganized docs and a cleared bug backlog.
RAM Shortage Hits Devices from AI Demand
RAM prices have tripled to sextupled due to AI companies hoarding memory chips. This shortage will cause price hikes and supply issues for phones, game consoles, handhelds, laptops, and routers. Consumers upgrading devices soon will face higher costs.
Kling 3.0 Video Models on AI Gateway
Vercel's AI Gateway now integrates Kling video models, including Kling 3.0, for generating cinematic videos from text, images, or frames. Key features include multishot narratives, synchronized audio, and precise start/end frame control. Available in beta for Pro, Enterprise, and paid users via AI SDK 6 and Gateway Playground.