🗞️ YOUR DAILY ROLLUP

Top Stories of the Day

🧠 Zuckerberg Splits Superintelligence Labs into Four Teams
Meta CEO Mark Zuckerberg has approved a sweeping reorganization of the company’s AI division, Meta Superintelligence Labs (MSL), dividing it into four focused units: a TBD Lab for foundational LLaMA‑style models, a Fundamental AI Research (FAIR) team, a Products & Applied Research group, and an Infrastructure (MSL Infra) team. The move reflects growing frustration with past performance and an urgent push toward artificial general intelligence.

🔲 NVIDIA Plans Faster China AI Chip Despite Limits
NVIDIA is building a new Blackwell-based AI chip for China, designed to beat the H20 while staying within U.S. export limits. Named B30A, it may ship as early as next month. Regulatory approval remains uncertain, as Washington weighs security concerns. The chip could help NVIDIA retain ground against rising local rivals like Huawei.

🤖 DeepSeek Unveils V3.1 With Longer Context Window
China’s DeepSeek has released version 3.1 of its AI model, offering improved long-context handling for more coherent and sustained interactions. Announced via WeChat, the update builds on V3 with better memory and response capabilities. Technical documentation remains sparse, but the Hangzhou-based firm is inviting testing as it expands its LLM footprint.

⚡ China Solves AI Energy Needs as U.S. Grid Strains
While U.S. AI growth is hampered by power constraints, China treats data center energy as a non-issue, says Tech Buzz China’s Rui Ma. Touring the country’s AI hubs, Ma found energy availability considered “solved,” unlike in the U.S., where it’s a critical bottleneck. With AI infrastructure now rivaling consumer spending, the disparity could reshape global tech dominance.

🪧 POWERED BY KORBIT AI

Code Faster, Ship Smarter

Korbit’s AI powered code review platform helps engineering teams catch bugs, boost code quality, and keep projects moving — all with instant, context-aware feedback on every pull request.

Korbit delivers:

• Automated PR summaries so reviewers spend time fixing, not guessing

• High signal reviews tailored to your codebase and standards

• Actionable team insights on review velocity, compliance, and quality trends

→ Get your free code audit

Enjoying our newsletter? Forward it to a colleague—
it’s one of the best ways to support us.

🪫 ENERGY

America’s AI Surge Fuels Growth, But Drains Power, Capital, and Balance

America’s economic growth is increasingly powered by artificial intelligence infrastructure. Investments in data centers, chips, and grid upgrades now account for up to 40% of recent GDP growth, driven by tech giants betting on AI’s long-term payoff. Ashburn, Virginia, has become a focal point, with data centers consuming a staggering share of regional electricity.

But the boom is distorting the broader economy. Traditional sectors like housing and manufacturing are being squeezed by rising interest rates and energy costs. As AI consumes more capital and power, other industries stall. If the AI build-out slows, America could face a hard landing. → Read the full article here.

💰 INVESTMENT

Billions In, but Not Much Out: Corporate America’s A.I. Bet Hasn’t Paid Off—Yet

Despite pouring nearly $62 billion into generative AI this year, most companies report little to no impact on profits. While tech suppliers like Microsoft and NVIDIA are reaping early gains, a striking 42% of firms have abandoned AI pilot projects, citing technical snags and human pushback. Analysts now say AI is entering the ‘trough of disillusionment,’ echoing past tech cycles where hype outpaced results. In a recent New York Times article, Steve Lohr noted that this cycle of overpromising and underdelivering has long defined the adoption of emerging technologies.

Some firms, like JPMorgan and USAA, are making incremental progress—saving time or boosting productivity in select areas—but sweeping efficiency gains remain elusive. Executives insist AI’s transformation will come, just not on Wall Street’s timeline. → Read the full article here.

⚛️ SCIENCE

New Brain Implant Decodes Inner Speech, Offering Hope for Nonverbal Communication

Researchers have developed a brain-computer interface that can decode a person’s unspoken thoughts—capturing inner monologue with up to 74% accuracy. Unlike earlier systems, this BCI doesn’t require users to attempt speech, reducing the physical burden on people with conditions like ALS or stroke-related paralysis. The study, published in Cell, marks a milestone in translating internal brain activity into usable language.

Ethical concerns loom, including the risk of reading unintended thoughts. To address this, researchers added a mental passphrase trigger—“chitty chitty bang bang”—to activate decoding. It's a clever safeguard in a field where mind-reading is no longer just sci-fi. → Read the full paper here.

𝕏 TWEETS

Google DeepMind is Shipping!

🛰️ NEWS

What Else is Happening

👨‍⚕️ Doctors May Be Losing Sharpness to AI: A study found experienced gastroenterologists got 20% worse at spotting colon polyps unaided after just 3 months using AI tools, raising concerns about over-reliance.

💸 Eight Sleep Secures $100M Boost: Backed by F1 stars and top VCs, the AI mattress maker plans global expansion and medical integrations—touting 1B+ hours of sleep data and $500M in sales.

🚀 ChatGPT Go Debuts in India: OpenAI launches a budget-friendly tier at ₹399/month ($4.5) with major boosts to usage limits—India gets it first, with plans to expand globally based on feedback.

⚖️ Otter AI Sued Over Secret Recordings: A class-action suit claims the transcription app recorded private meetings without consent to train its AI, raising major privacy and wiretap concerns.

🧑‍💼 Newsrooms Race to Hire AI Experts: Media outlets are adding AI leads to navigate tools, ethics, and legal risks, despite tighter budgets than tech giants, the investment is becoming essential.

💼 JOB BOARD

Now Hiring: WorkSpan, Outreach, & More

📽️ VIDEO

GPT-5 Prompt Optimization Guide

GPT-5’s new prompt guide unlocks agentic control, faster reasoning, and coding tips—plus a prompt optimizer to refine workflows and boost results.

🧰 TOOLBOX

Scalable Databases, Smarter Influence, and Affordable Transcriptions

🗒️ FEEDBACK

Help Us Get Better

That’s a Wrap!

📢 Want to advertise with Forward Future? Reach 550K+ AI enthusiasts, tech leaders, and decision-makers. Let’s talk—just reply to this email.

Thanks for reading today’s newsletter—see you next time!

Matthew Berman & The Forward Future Team

🧑‍🚀 🧑‍🚀 🧑‍🚀 🧑‍🚀