All Updates
Page 321 of 912
April 3, 2026
SpinQ Raises 600M CNY in C+ Round
Shenzhen's SpinQ completed C+ financing of 600 million CNY, totaling nearly 1 billion in three months, with scaled profitability. Offers NMR quantum computers for education and superconducting ones for research, with first overseas supercon deliveries. Q1 orders surged 80%, overseas business expanding.
Cloudflare Launches EmDash WordPress Alternative
Cloudflare launched EmDash, an open-source MIT-licensed CMS as a secure alternative to WordPress, built from scratch without WP code. It targets modern web hosting on global edge networks and appeals to new developers and AI agents avoiding WordPress vulnerabilities. Analysts note AI agents rarely choose WordPress unless instructed.
Google Launches Strongest Small Model for Phones
Google has released its strongest small AI model, capable of outperforming larger models. The model is optimized to run efficiently on smartphones. This marks a breakthrough in on-device AI capabilities.
iPhone 17 Scratches Spark Complaints Amid Record Sales
Apple's iPhone 17 series faces backlash for soft aluminum frames that scratch easily, slow 40W charging, weak signal, and 8GB RAM, yet iPhone 16 dominates global sales and Apple holds 65% premium market share in China. Users criticize conservatism but stay loyal due to ecosystem stability and resale value. Apple likened to Toyota for reliability over innovation.
Qwen-code v0.14.0 Adds Qwen3.6-Plus Support
Qwen-code version 0.14.0 introduces support for the Qwen3.6-Plus model alongside an extensible Channels platform for Telegram, WeChat, and DingTalk. It adds in-session cron scheduling, cross-provider model selection for subagents, and numerous fixes for stability issues like PTY leaks and subprocess management. This update enhances agent reliability and extensibility for coding workflows.
Geely WoFei ChangKong Launches eVTOL IPO Push
WoFei ChangKong, Geely's subsidiary, secured Sci-Tech Innovation Board IPO counseling on April 2. The firm develops pure-electric passenger eVTOL for low-altitude smart traffic. AE200-100, its debut product, rolled off the line in Chengdu on Sept 29, 2025, boasting 200km range and key flight test milestones.
Shulong Cup 2 Adds AI Agent Track
Second Shulong Cup Global AI Innovation Competition launches in early April with a new AI agent track to discover practical solutions. It expands to include multimodal apps, vertical industries, and more, offering media exposure and resources. Developer salon planned in Hangzhou.
CITIC eyes 2026 Agent boom & evolution
CITIC report bullish on AI Agents landing rapidly by 2026, driven by structured application via Agent Harness beyond model improvements. OpenRouter token calls expected up 11% WoW. Focus on self-iterating Agents.
Doubao LLM Surpasses 120T Daily Tokens
ByteDance's Doubao large model achieved over 120 trillion daily Token usage. Broadcasting industry issued strict statement against AI face-swapping. Other updates include auto industry leadership changes and L2 ADAS advancements.
Thiel: Asperger's Fuels SV Success
Peter Thiel claims neurotypical people fail in tech due to social conformity killing ideas, praising Asperger's traits like hyper-focus seen in Musk, Gates, Zuckerberg. Neurodiversity offers superior pattern recognition and independence amid imitation trends like LLMs. Critiques business schools for breeding conformists.
Musk on OpenAI secondary market chill
OpenAI shares are struggling in secondary markets, becoming nearly unsellable as investors shift to rival Anthropic. Elon Musk commented 'not surprising' on the report. This reflects cooling investor sentiment toward OpenAI.
Arcee Launches 399B Open-Source Reasoning Model
Arcee AI released Trinity-Large-Thinking, a powerful 399-billion parameter text-only reasoning model under Apache 2.0 license. The San Francisco-based 30-person team trained it in 33 days on 2048 NVIDIA B300 GPUs for $20M. It offers enterprises a customizable U.S.-made alternative amid concerns over Chinese models.
Wan 2.7 Video Suite Launches on Together AI
Together AI has released Wan 2.7, a four-model video suite for generation, continuation, reference-driven workflows, and editing. The rollout starts with text-to-video capabilities. This expands access to advanced video AI tools on their platform.
LLMs Boost DB Queries 4.78x
New research from Together AI shows LLMs can optimize database query execution plans. They deliver up to 4.78x speedups by correcting cardinality estimation errors overlooked by statistical heuristics. This advances AI applications in core systems.
April 2, 2026
Uncensored Gemma 4 E4B/E2B Multimodal Launch
New aggressive uncensored variants of Gemma 4 E4B (4B) and E2B (2B) released as fully multimodal models supporting text, image, video, and audio. Available in high-quality GGUF quants on Hugging Face, compatible with llama.cpp. Larger E31B and E26B models coming soon.
Broadcom Hires Alphabet Exec as CFO
Broadcom announced the appointment of Amy Turner as CFO, effective June 12. Turner, previously Alphabet's VP, Corporate Controller, and CAO since 2018, succeeds retiring CFO Mark Spears. Spears will advise for nine months to ensure smooth transition; Turner was formerly at PwC.
Tesla Ends Model S/X Era for Robots
Tesla halts custom orders for Model S and X, leaving ~600 inventory vehicles as it repurposes Fremont factory for Optimus humanoid robots. Amid 2025 sales drop and competition, company pivots to AI via robotaxis and robotics. Models delivered 610k+ units but now <3% of volume.
Gemini Generates Editable Slides in Google Slides
Gemini in Google Slides can now generate editable slides. The feature launches initially in the US and select regions. Japanese language support is planned soon.
Daihatsu Trains Factory Workers into AI Experts in 2 Months
Daihatsu transforms factory line workers who rarely touch PCs into key AI utilization personnel through targeted DX talent development. The company promotes 'people-friendly, everyone's digital' reforms originating from the field, yielding significant results. An interview highlights efforts led by Taigo Mugen.
AMD R9700 96GB VRAM Beats 5090 Price
Reddit post hails AMD R9700 for 96GB VRAM, delivering RTX 5080-level inference speed and quality. Claims it's cheaper than RTX 5090, ideal for future AI nodes. User apologizes for blurry photo.