All Updates
Page 798 of 1665
April 29, 2026
Big Tech AI Capex Faces Earnings Test
Google, Microsoft, Amazon, and Meta's Q1 earnings on April 30 will scrutinize AI investments exceeding $650B by 2026. Investors demand proof of ROI from cloud AI demand, not just spending scale. OpenAI growth concerns add pressure amid market rebound.
Hinton: From 'Fraud' to Nobel, Turing, WDFT Winner
Geoffrey Hinton evolved from AI winter skeptic to 'AI Godfather' with Turing Award, Nobel Physics Prize, and WDFT Breakthrough Award. He drove deep learning's rise but now warns of AI risks and safety at its peak.
Dex Raises $5.3M Seed for AI Talent Matching
AI recruiting startup Dex raised $5.3M in seed funding to connect AI engineers with high-demand companies. Founded by ex-Atomico adviser Paddy Lambros, it charges fees only on successful hires and achieved $1.8M ARR from zero in under six months.
AI Hardware: GPU to TPU Evolution
CPU inefficiency for AI matrix ops drives specialized hardware like GPUs (SIMT parallelism), TPUs (systolic arrays), and ASICs (dataflow optimization). Tradeoffs in flexibility vs efficiency explained for training/inference. Covers quantization, sparsity, and distributed strategies.
GitHub Fixes Critical RCE Vuln in <6 Hours
Wiz Research used AI models to uncover a critical remote code execution vulnerability in GitHub's internal git infrastructure, potentially allowing access to millions of public and private repositories. GitHub validated the issue in 40 minutes and deployed a fix within six hours.
Eka Robots' ChatGPT Moment: Pincers
Ekaโs robots perform eerily lifelike tasks like sorting chicken nuggets and screwing in light bulbs. The article compares this to ChatGPT's breakthrough but questions their genuine physical intelligence.
Qwen-Code v0.15.5: CLI Fixes & Shell Enhancements
Qwen-Code v0.15.5 introduces MCP config via CLI, fixes CLI header refresh on model switch, and integrates background shells into task_stop and dialog tools. New tmux skill improves TUI test logs readability, plus DeepSeek tool-call fixes. First contribution from @eliird.
Scout AI Raises $100M for Military AI Agents
Scout AI, founded by Coby Adcock, secured $100 million to train AI models for warfare. The startup is building AI agents enabling individual soldiers to control fleets of autonomous vehicles. TechCrunch visited their bootcamp training ground.
GM Rolls Out Gemini to 4M Vehicles
GM is deploying Google Gemini via OTA update to 4 million US vehicles (2022+ Cadillac, Chevrolet, Buick, GMC), replacing Google Assistant. This marks one of the largest in-car AI rollouts, amid data-sharing controversies and FTC scrutiny.
Hypothetical DeepSeek-Kimi Merger Brainstorm
This opinion piece from ้ๅชไฝ speculates on a hypothetical merger between DeepSeek and Kimi AI offerings. It poses the question of what might happen if these prominent Chinese LLMs combined forces. The teaser invites readers to imagine the potential impacts on the AI landscape.
Why AI Phones Still Elude Us
True AI phones demand new NPUs for on-device multimodal models, high memory bandwidth, and OS permission overhauls unachievable via software updates. Ecosystem giants like WeChat and Taobao resist radical AI agents that bypass app data monetization. Doubao phone's GUI Agent innovation faced industry-wide boycott.
Lovable Mobile Apps for Vibe Coding
Lovable launched iOS and Android mobile apps. These bring its AI-powered vibe coding platform to mobile devices. The update expands access to a wider audience.
SAP AI API Policy Sparks Lock-in Fears
SAP's new API policy prohibits using its APIs to integrate with AI systems outside its endorsed architectures. Partners are concerned this creates vendor lock-in, blocking third-party AI tools from SAP data. Experts warn it may push users toward undocumented APIs.
GM Adds Gemini to 4M Cars
General Motors will integrate Google's Gemini AI into around four million US vehicles from Cadillac, Chevrolet, Buick, and GMC (2022+ models with Google built-in). The upgrade from Google Assistant will roll out via OTA updates over several months. GM calls it one of the largest Gemini deployments.
Cognizant Buys Astreya for $600M AI Infra
Cognizant is acquiring San Jose-based Astreya for $600 million to bolster its AI builder strategy. The deal targets physical data center design, build, and operations for enterprise AI. This marks Cognizant's fourth major acquisition in 18 months.
Delete 90% AI Plugins by 2026
Rapid iterations of large models will make 90% of AI plugins obsolete by 2026. AI practitioners should master 'ๆญ่็ฆป' decluttering from hundreds of tools to stay efficient.
AI Entrepreneurs Become Zhipu & Kimi 'Workers'
Chinese AI entrepreneurs using Zhipu AI and Kimi models have become de facto 'workers' for these providers due to heavy dependency. Computing power shortages act as a major stumbling block, preventing true independence. The article critiques this dynamic in the AI startup ecosystem.
Banmo Yixing T6 AV Debuts Commercially
Banmo Yixing captivated Silicon Valley at The Magic X global embodied intelligence conference. Their T6 autonomous vehicle signals the dawn of a new commercial era for unmanned driving. The event garnered worldwide attention.
AI Fuels Google-Meta Ad Revenue Boom
Google and Meta are experiencing a digital advertising boom. Artificial intelligence is automating marketing processes. This automation is driving record sales for these online ad giants.
Fujian Compute-Power Consortium Forms
The 'Suan Dian Lian He Ti' compute-power consortium was established in Fujian province. Taichu Yuangi joins as one of the inaugural members. This alliance aims to integrate computing and power resources.