Good morning. It's Wednesday, March 11, and we're covering AI-powered drug discovery, CEOs betting on AI-driven hiring, Yann LeCun’s $1B push to build next-generation “world model” AI, and more.
New here? To keep up with the future of tech, sign up here. If you liked this email, share it with a friend.
LeCun’s AMI Labs Raises $1B
AMI Labs, co-founded by Yann LeCun, raised $1.03B at a $3.5B valuation to develop “world models,” AI designed to learn from real-world data rather than language alone. The startup aims to advance Joint Embedding Predictive Architecture (JEPA), though commercial applications may take years. Initial focus includes healthcare partnerships such as Nabla. Investors include Cathay Innovation, Greycroft, Bezos Expeditions, NVIDIA, and Toyota Ventures.
Murati’s AI Lab Strikes NVIDIA Compute Deal
Thinking Machines Lab, founded by Mira Murati, signed a multi-year partnership with NVIDIA for at least one gigawatt of AI compute starting in 2027. NVIDIA will also make a strategic investment in the startup, already valued above $12B. The company is building AI models designed for reproducible results.
Meta Acquires AI Agent Network Moltbook
Meta has acquired Moltbook, a social network designed for AI agents. Co-founders Matt Schlicht and Ben Parr will join Meta Superintelligence Labs as part of the deal. Moltbook gained attention for enabling AI agents to interact and share content autonomously.
Google Upgrades Gemini Across Workspace
Google added major Gemini upgrades to Workspace that let users generate Docs, Sheets, and Slides from a single prompt using data across Gmail, Drive, Chat, and Search. The AI can draft structured documents, populate spreadsheets, and design presentations automatically. Google says some tasks in Sheets are up to 9× faster. The features are rolling out in beta for AI Pro users and enterprise customers.
LIVE SHOW
Chris Lattner Joins The Future Live
“The one thing we know for certain is that people who ignore this shift will get left behind.” @clattner_llvm CEO of @Modular and creator of LLVM, on the reality of AI programming tools:“Both sides of the AI coding debate are wrong in different ways.”“There’s a lot of hype… pic.twitter.com/n0LppKSMxH
@ForwardFutureView post on XREAL OR AI
Think You Can Still Spot AI?
Prove it. ↓ Real or fake. Answer below.
VIDEO
This Fly Is Living in the Matrix...
Scientists simulated a fruit fly brain neuron-by-neuron in a virtual body, creating lifelike behavior and hinting at future human brain emulation.
MEDICINE
AI Speeds Search for Drugs Against Superbugs, Parkinson’s, Rare Diseases
Artificial intelligence is accelerating drug discovery, helping scientists identify potential treatments for antibiotic-resistant infections, Parkinson’s disease, and rare illnesses that have long lacked effective therapies. Researchers at MIT used generative AI to design 36 million potential antibiotic compounds, ultimately identifying two promising candidates effective against drug-resistant bacteria like MRSA and gonorrhea.
For diseases with unclear causes—such as Parkinson’s—other teams are applying machine learning to screen billions of molecules in days instead of months. The technology is also helping researchers repurpose existing medicines for rare diseases by analyzing thousands of drugs against thousands of conditions. While many AI-discovered compounds remain in early testing, scientists say the approach could significantly reshape how new medicines are found. → Read the full article here.
WORKFORCE
Survey: Most U.S. CEOs Expect AI to Boost Hiring, Not Cut Jobs
A new KPMG survey finds most U.S. CEOs expect artificial intelligence to increase hiring rather than eliminate jobs in 2026. Just 9% of CEOs at large companies plan workforce reductions tied to AI, while 55% anticipate adding employees and 36% expect no change, according to KPMG’s 2026 U.S. CEO Outlook Pulse Survey.
Executives remain optimistic about AI’s long-term business value but say the short-term return on investment has been limited, largely due to slow integration into existing systems and workflows. At the same time, CEOs increasingly see AI as a cybersecurity risk, particularly in enabling more sophisticated malware and phishing attacks. → Read the full article here.
AI SENTIMENT
Poll: Majority of U.S. Voters Say AI Risks Outweigh Benefits
A new NBC News poll finds 57% of registered U.S. voters believe the risks of artificial intelligence outweigh its benefits, compared with 34% who say the opposite. Public sentiment toward AI is broadly negative: 46% report unfavorable views, while just 26% feel positively, placing AI among the least popular topics measured in the survey.
Voters also show little confidence in political leadership on the issue—a third remain unsure about how the government should handle AI rules. Concerns center on potential job losses, cybersecurity risks, and the broader societal impact as AI adoption accelerates across industries. → Read the full article here.
NEWS
What Else Is Happening
YouTube Expands AI Deepfake Detection: The platform is piloting tools that flag AI-generated likenesses of politicians, officials, and journalists, letting them request removals as concerns grow over deepfakes.
ChatGPT Adds Interactive Learning Visuals: OpenAI launched dynamic modules showing math and science concepts in real time, as ChatGPT evolves from answer engine into hands-on study tool.
LinkedIn Emerges as AI Citation Hub: New data shows chatbots increasingly cite LinkedIn posts and profiles for professional queries, meaning executives’ online presence may influence how AI systems determine authority.
Google Deploys Gemini Agents at Pentagon: Google is rolling out Gemini AI agents to automate routine administrative tasks across the U.S. Department of Defense’s roughly 3-million-person workforce.
China’s AI Firms Pay Users to Join (Paywall): Chinese tech giants handed out 8 billion yuan ($1.2 billion) in coupons to boost adoption of new “agentic” AI apps, escalating an intense user-acquisition war.
AgentMail Raises $6M for AI Inboxes: The startup built an email service giving AI agents their own inboxes via API—aiming to turn email into a universal identity layer as autonomous agents begin operating online.
REAL OR AI
The Image Is Real. Tune in next week for another Real or AI test.
