🔲 Amazon’s Trainium Chips Gain Big Traction
Amazon says its Trainium2 AI chip business is already generating multibillion-dollar revenue, with over a million chips in production and 100,000+ customers. Trainium3, unveiled at re:Invent, is four times faster and more efficient than its predecessor. Anthropic alone uses more than 500,000 Trainium2 chips for Project Rainier. Amazon argues Trainium wins on price-performance, helping it chip away at NVIDIA’s dominance.

⚖️ EU Probes Meta Over WhatsApp AI Ban
EU regulators opened an antitrust investigation into Meta after WhatsApp moved to ban third-party AI chatbots from its Business API starting in January. The policy keeps Meta AI available while blocking general-purpose bots like ChatGPT and Perplexity, though customer-service bots remain allowed. The Commission worries Meta could shut rivals out of a key channel in the EEA. Meta calls the claims baseless, arguing users can access competing bots elsewhere.

🧑‍⚖ Chicago Tribune Sues Perplexity
The Chicago Tribune has filed a federal lawsuit against Perplexity AI alleging the company illegally scraped and reproduced its news content — sometimes verbatim — via its RAG‑powered systems and paywall‑bypassing Comet browser. The suit claims Perplexity’s prior denial (“no training on Tribune work, only non‑verbatim summaries”) is contradicted by its actual output. The case adds fresh pressure to how generative‑AI tools handle copyrighted journalism. 

🗣️ Anthropic Debuts Large-Scale AI Interview Tool
Anthropic Interviewer ran 1,250 automated interviews to study how professionals use and feel about AI. Most participants expressed optimism, though concerns surfaced around education, creative displacement, and security. Workers want AI to handle routine tasks while they keep identity-defining work. Creatives face stigma yet rely on AI for productivity. Scientists want AI-driven research partners but don’t trust current tools for core experiments.

Which everyday technology began as a failed attempt to detect evaporating black holes only to end up powering your laptop, your phone, and probably the device you’re reading this on right now?

Stick around to find out! 👇

AWS reveals massive AgentCore upgrades with built-in guardrails, real-time evaluations, and episodic memory to fix why most AI agents fail.

👉 Watch on X or YouTube.

Meta shares rose 4–6% after Bloomberg reported the company is considering cutting up to 30% of its metaverse budget for 2026, shifting resources toward faster-growing AI and advertising initiatives. The reductions—discussed during recent budget meetings and potentially accompanied by layoffs early next year—would impact Reality Labs, including Quest hardware and Horizon Worlds.

Analysts say the move could meaningfully improve earnings, with Mizuho estimating metaverse cuts might add ~$2 per share to its 2026 EPS forecast. The report follows Meta’s $29 billion bond raise in October to fund AI infrastructure and broader capital spending. Investor reaction was positive, viewing the potential cuts as a sign of tighter capital discipline and stronger focus on revenue-generating AI products. → Continue reading here.

Nevada’s Tahoe–Reno Industrial Center has become one of the world’s fastest-growing datacenter clusters as tech giants including Microsoft, Amazon, Google, Meta and Oracle accelerate massive AI-driven buildouts. The expansion—fueled by multibillion-dollar investments and expectations of global AI datacenter spending reaching $7tn by 2030—demands enormous amounts of land, water and electricity in one of the driest regions in the US.

Local tribes and environmental advocates warn that escalating water withdrawals and rising power consumption threaten the Truckee River watershed and strain an already fragile grid. Companies tout reclaimed-water systems and renewable-energy projects, but the growth rate is outpacing regional resources. The result is a modern gold-rush economy carrying equally modern risks. → Read the full article here.

AWS introduced “AI Factories,” a new offering that installs dedicated AI infrastructure—including NVIDIA’s latest accelerators, AWS Trainium chips, and high-speed networking—directly inside customers’ existing data centers. The model functions like a private AWS Region, giving enterprises and governments low-latency access to compute, storage, and foundation models while meeting strict data-sovereignty and regulatory requirements.

By leveraging customers’ existing space and power, AWS says it can cut multi-year AI buildout timelines and eliminate the capital burden of acquiring GPUs and integrating complex systems. The launch extends AWS’s long-running partnership with NVIDIA and includes a major deployment with HUMAIN in Saudi Arabia featuring up to 150,000 AI chips. Why it matters: organizations struggling to scale AI securely can now outsource the infrastructure without giving up physical control of their data. → Read the full article here.

📉 Microsoft Halves AI Sales Targets: The company cut Foundry quotas after most sales teams missed steep goals, showing enterprises aren’t paying premium for brittle agent tools.

🤔 AI Surge Drains Memory Supply: A global chip crunch is doubling prices and squeezing DRAM inventories, delaying data-center plans and pushing phone makers toward price hikes.

🤝 Anthropic, Snowflake Seal $200M AI Deal: Multi-year pact brings Claude models into Snowflake, enabling secure multimodal analysis and custom agent building for enterprise clients.

💼 Hinton Warns of AI-Driven Job Loss: The “Godfather of AI” says mass unemployment is likely as companies invest trillions in automation, urging caution amid uncertain long-term forecasts.

In the late 20th century, engineers trying to build fast indoor wireless networks were running into a stubborn physics problem called multipath distortion: radio waves ricocheted off walls, furniture, and unsuspecting houseplants, arriving at different times and turning clean data into scrambled static.

Enter an unlikely hero: Dr. John O’Sullivan and a team of Australian radio astronomers. Years earlier, they had engineered a sophisticated mathematical technique to isolate faint signals from “evaporating” black holes—hypothetical bursts predicted by Stephen Hawking. The cosmos never coughed up those black holes, but the algorithm did prove remarkably good at sifting through noise.

O’Sullivan noticed that the messy echoes inside a living room looked suspiciously like the messy echoes of deep space. So the team retooled their cosmic filter, packed it into a microchip, and—without meaning to—solved one of wireless networking’s biggest headaches.

That chip became the backbone of modern Wi-Fi standards.

📣 Want to advertise with Forward Future? Reach 600K+ AI enthusiasts, tech leaders, and decision-makers. Let’s talk—just reply to this email.

🛰️ Want more Forward Future? Follow us on X for quick updates, subscribe to our YouTube for deep dives, or add us to your RSS Feed for seamless reading.

Thanks for reading today’s newsletter—see you next time!
Matthew Berman & The Forward Future Team

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀