👨💼 Arm Hires Amazon AI Chip Expert
Arm has recruited Rami Sinno, the Amazon engineer behind Trainium and Inferentia, to lead its shift toward building complete AI chips. The move signals Arm’s growing ambition to move beyond IP licensing into full-stack hardware development. Sinno joins a wave of hires from rivals like HPE and Intel. It’s a calculated bid to gain leverage in the booming AI hardware space.
📚 Grammarly Launches AI Agents for Student Writing
Grammarly now offers nine AI agents to assist students and educators with writing, editing, and grading tasks. Free and Pro users can access tools like predicted grades, citation generation, and feedback based on instructor data. For educators, plagiarism and AI detection tools are also available. Enterprise and Education rollout is expected later this year.
🕵️♂️ Texas Probes Meta, Character.ai Over Mental Health Claims
Texas AG Ken Paxton is investigating Meta and Character.ai for allegedly marketing AI chatbots as mental health tools without proper oversight. Officials say such bots may mislead children into thinking they’re receiving real therapy. Disclaimers exist, but critics argue kids ignore them—and their data may be used for ads. Civil investigative demands have been issued.
📈 Perplexity Adds Live Earnings Transcripts for Indian Stocks
AI startup Perplexity now offers live transcriptions of Indian public company earnings calls through its Finance dashboard. This expands beyond its previous U.S.-only coverage, adding a conference call calendar and enhancing tools like stock watchlists and sector tracking. It’s a boost for investors researching India’s fast-moving equity markets.
Enjoying our newsletter? Forward it to a colleague—
it’s one of the best ways to support us.
🪧 POWERED BY ZAPIER
The Most-Connected AI Orchestration Platform
Easily add AI where your work happens—whether it’s a step in a workflow, an autonomous agent, or a chatbot. Zapier connects AI tools like ChatGPT and Claude to over 7,000 apps, enabling smart workflows, agents, and chatbots—all with zero code.
Try Zapier’s AI orchestration platform for free today.
🏙️ SOCIETY
AI Study Shows Public Spaces Are Busier Than Ever, but Strangely Lonelier
An AI-led study by Carlo Ratti and academic partners analyzed video footage from the 1970s and today in New York, Boston, and Philadelphia, revealing a clear trend: people walk faster, linger less, and socialize more rarely in public spaces. Using computer vision to track human behavior, the team found that even iconic spots like Bryant Park have become more transactional, with fewer spontaneous encounters and shorter dwell times—suggesting a quiet decline in urban social interaction.
Ratti argues that the same technology used to diagnose this change could help reverse it. By testing design tweaks—shading, seating, layout—cities could restore human connection. → Read the full article here.
🗺️ GEOPOLITICS
Trump’s Deregulatory AI Agenda Faces a European Firewall
President Trump’s new AI Action Plan aims to strip away regulations and let U.S. tech companies develop AI with minimal oversight. But Europe’s AI Act, the world’s first comprehensive AI law, stands in the way. To access the EU market, American firms must comply with strict rules on privacy, bias, and safety—creating a global ripple effect known as the “Brussels Effect.”
Despite U.S. pressure, Europe is holding firm, signaling that AI governance and innovation aren’t mutually exclusive. As more countries adopt EU-style rules, America’s hands-off approach looks increasingly isolated. The regulatory future of AI may not be written in Washington. → Read the full article here.
📹️ FROM THE LIVE SHOW
OpenAI’s Tejal Patwardhan on Emergent Misalignment
FROM THE LIVE SHOW: On emergent misalignment During training, models can drift into “bad boy” personas, or other types you don’t want. So what’s the solution? Building on the work of @OwainEvans_UK and his group, @OpenAI’s @tejalpatwardhan says you can measure and catch this — Forward Future (@forward_future_) 5:45 PM • Aug 17, 2025
@forward_future_View post on X⚛️ SCIENCE
AI Tackles Malnutrition in Maternal, Child Health
A new paper in Nature Communications explores how artificial intelligence and precision nutrition can reshape maternal and child health interventions in low-resource regions. Despite decades of investment, malnutrition remains widespread—half of women and children have at least one micronutrient deficiency, and global anemia rates have barely budged. Precision nutrition offers a shift from one-size-fits-all programs to data-driven strategies tailored to individual biology, environment, and microbiome.
The study highlights AI’s potential to predict nutrition outcomes across mother-child dyads, integrating clinical, genomic, and microbial data. Yet, scaling these tools in low-income settings poses challenges—spotty data, limited infrastructure, and ethical concerns. → Read the full paper here.
🛰️ NEWS
What Else is Happening
⚡ AI Strains U.S. Power Grid: Explosive demand for massive AI data centers is overwhelming utilities, triggering delays, soaring costs, and a trillion-dollar scramble for space and electricity.
🗂️ Paradigm Reimagines Spreadsheets: With an AI agent in every cell, Paradigm turns spreadsheets into dynamic workflows, now public with $5M in fresh funding to push its bold roadmap.
✨ Grammarly Debuts AI-Powered Redesign: Now built on Coda, Grammarly’s sleek new interface adds tools like AI grading, citation help, and content detection—aimed at both writing and refining.
🎮 AI Now Powers Most Game Development: Google says 87% of developers use AI agents to cut costs and automate tasks, though concerns over jobs, pay, and IP rights still loom large.
🚨 Meta Probed Over AI Chat With Kids: A leaked doc suggests Meta’s AI allowed "sensual" chats with minors—sparking a Senate investigation and fierce backlash over safety and content oversight.
📽️ VIDEO
Mark Chen: GPT-5, Open-Source, Agents, Future of OpenAI, and more!
ICYMI: Matt sits down with Mark Chen, who leads research at OpenAI, to unpack GPT-5’s advances in reasoning, coding, synthetic data, and AI’s future.
🗒️ FEEDBACK
Help Us Get Better
What did you think of today's newsletter? |
