⚙️ Google + Intel double down on AI infrastructure

💸 OpenAI’s $100 plan targets power users (and Claude)

🔐 Perplexity is quietly building an AI security lab

⚡ Microsoft partners to scale massive AI data centers

🎥 Google cuts video AI costs by 50%

💰 OpenAI raises $122B (yes, billion)

🚀 SpaceX may be heading for an IPO

Google Cloud will continue deploying Intel Xeon chips, including the latest Xeon 6, across AI, inference, and general workloads 

The companies are expanding co-development of custom ASIC-based IPUs to handle networking, storage, and security tasks 

IPUs offload infrastructure workloads from CPUs, improving efficiency and freeing compute for core AI tasks 

The partnership builds on a collaboration that started in 2021 and will span multiple future chip generations 

Move comes amid rising demand for CPUs as AI shifts from training models to running them in real-world systems 

1,000+ ready-to-use prompts to solve problems in minutes instead of hours—tested & used by 1M+ professionals

Superhuman AI newsletter (3 min daily) so you keep learning new AI tools & tutorials to stay ahead in your career—the prompts are just the beginning

New $100 Pro tier sits between $20 Plus and $200 Pro, filling the gap for serious users 

Offers 5× more Codex usage than Plus, with even higher temporary limits until May 31 

Designed for heavy coding workflows where rate limits matter most during active sessions 

No plan offers unlimited usage, while the $200 tier still provides ~20× higher limits than Plus 

Codex adoption is surging with 3M+ weekly users and rapid growth, driving demand for higher tiers 

Claude Managed Agents provides a fully managed cloud environment to run AI agents with built-in orchestration 

Agents can browse the web, run code, read files, and execute tasks securely inside sandboxed environments 

Supports long-running sessions that can operate for minutes or hours with persistent memory and checkpoints 

Developers no longer need to build agent loops, tool execution, or infrastructure, reducing build time significantly 

Designed for production use with scoped permissions, monitoring, and multi-agent coordination 

Google introduced Veo 3.1 Lite via the Gemini API, cutting video generation costs by over 50% compared to Veo 3.1 Fast while maintaining speed. It also announced price drops for Veo 3.1 Fast from April 7 to drive broader adoption.

Anthropic confirmed a major leak of Claude Code source files after an npm release error exposed nearly 2,000 internal files. The breach revealed features and vulnerabilities, prompting warnings about supply chain attacks, typosquatting, and urgent security measures.

OpenAI secured $122 billion in funding at an $852 billion valuation, backed by Amazon, NVIDIA, SoftBank, and Microsoft. With nearly 1 billion weekly ChatGPT users and $2B monthly revenue, the company is accelerating infrastructure and global expansion.

Google unveiled Gemma 4, its latest open-weight AI models optimized for local use across GPUs and mobile devices. With improved speed, lower memory requirements, and relaxed licensing, the update targets developers seeking more control and efficient on-device AI deployment.

Apple blocked updates to AI coding apps like Replit, citing safety rules, but critics argue the move limits innovation. The decision could discourage new developers and shift momentum toward open platforms as AI-driven app creation gains traction.

SpaceX has secretly filed for an IPO with a potential June listing, according to reports. The confidential filing keeps financial details private, but the move signals Elon Musk’s company may soon open to public investors, fueling market anticipation.