- Future Tools
- Posts
- Anthropic’s $50B power play
Anthropic’s $50B power play
Plus: Zuck loses a legend
Welcome back! AI is showing up in places none of us expected. The latest? Churches around the world started experimenting with “chatbot Jesus,” using AI to handle prayers, sermons, counseling…and yes, even miracles-on-demand.
Pastors say it saves time. Congregations say it’s strangely comforting. And critics? Well, they’re not thrilled about outsourcing spirituality to a server rack.


Anthropic Bets $50B on America’s AI Backbone

Anthropic is building its own AI super-grid. The company announced a sweeping $50 billion investment to build US data centers across Texas and New York. The first sites will go live in 2026, with more planned in a multi-year expansion.
Massive build-out plans: Anthropic’s rollout includes custom GPU facilities developed with cloud partner Fluidstack, creating thousands of construction jobs and hundreds of technical roles.
Racing against OpenAI’s scale: The push comes as OpenAI continues its unprecedented $1.4 trillion infrastructure spree. Both companies now see physical compute (land, power, chips, and grid capacity) as the real battleground.
Why it matters: Anthropic’s big bet signals a change in the equation—infrastructure dominance > model quality. Whoever controls the compute layer controls what’s possible at the frontier.
Microsoft Unveils its First AI “Superfactory”
Microsoft is linking data centers into a single AI machine. Its new “superfactory” concept connects two geographically separate facilities and treats them as one unified computational system.
Smart data centers: The superfactory pools power, cooling, networking, and memory across the pair of facilities, allowing AI training workloads to run as if the entire system were a single giant processor—something traditional data centers can’t do.
The next wave of gigantic models: The design is aimed at supporting trillion-parameter systems, high-intensity inference loads, and rapid scaling without redesigning each site from scratch.
The bigger picture: Just like Anthropic’s investment, Microsoft’s move shows the AI race has shifted into physical infrastructure. Cloud giants are quickly moving to build the industrial grid beneath them.
Yann LeCun Prepares to Exit Meta to Launch New AI Startup
Meta’s chief AI scientist Yann LeCun is planning his next act. LeCun, one of the founding fathers of modern AI and a leading voice for open systems, is preparing to leave Meta and build a startup of his own.
A research-driven venture: Early reports suggest the company will pursue foundational AI research, likely aligned with LeCun’s long-held belief in autonomous reasoning systems rather than massive LLM scaling.
Why it matters: The AI talent reshuffle continues. As top researchers break off to pursue independent visions, the winners of the next AI wave may be determined less by capital and more by who attracts—and keeps—the best minds behind the machines.


AI integrations without the headaches

Via Klavis AI
Klavis AI is an open-source integration platform that lets AI models and agents connect to thousands of external business tools without comp
lex authentication or custom infrastructure.
How you can use it:
Build agent workflows that pull real-time CRM, comms, or SaaS data
Let LLMs trigger actions across apps with structured, secure access
Test prompts and integrations through AI-friendly protocols
Deploy connectors self-hosted or fully hosted to fit enterprise needs
Pricing: Free and paid plans available

Your AI-powered data engineering assistant

Via Yorph AI
Yorph AI is a conversational data engineering platform that connects multiple data sources and automates pipelines using AI agents—no code, full governance, with human-in-the-loop controls.
How you can use it:
Ask natural-language questions across datasets
Generate and modify pipelines automatically
Clean, merge, and validate data on demand
Speed up analytics while maintaining transparency and compliance
Pricing: Free and paid plans available


Jobs, announcements, and big ideas
OpenAI rolls out GPT-5.1 with sharper conversation and customizable ChatGPT controls.
Firefox launches AI Window, an opt-in feature letting users chat with an AI assistant while browsing.
Disney+ will allow user-generated content via AI.
Baidu releases open-source multimodal model, claiming performance above GPT-5 and Gemini.
Epidemic Sound debuts AI Studio for auto-generated, beat-matched soundtracks.
World Labs introduces Marble, a commercial world model for persistent 3D environment.
Waymo extends robotaxi coverage to highways and San Jose International Airport.
Deezer survey shows listeners often mistake AI music for real artists.


The best AI tool nobody expected? NotebookLM. And I’ll tell you exactly why ➡️

That’s a wrap! Enjoy your weekend. See you next time.
—Matt (FutureTools.io)
