🥇 Anthropic Releases Claude Sonnet 4.5
Touting improved code-editing accuracy, stronger long-context performance, and enhanced agentic capabilities, Anthropic says Claude Sonnet 4.5 strikes a better balance between creativity and control than earlier models while reducing issues like deception and hallucination. It’s now powering updates to Claude Code and is available through Amazon Bedrock and GitHub Copilot.
🛍️ OpenAI Launches In‑Chat Shopping
OpenAI has introduced an Instant Checkout feature in ChatGPT, letting users buy products from Etsy and select Shopify merchants without leaving the chat. Payments are handled via Stripe, with no added fees for users and a small fee charged to merchants.
🚀 DeepSeek Unveils “Intermediate” AI Model
China’s DeepSeek has released DeepSeek‑V3.2‑Exp, calling it an “intermediate step” toward its next-generation AI system, with gains in training efficiency and long-sequence processing. The model debuts a “Sparse Attention” mechanism that slashes compute costs and cuts API prices by more than 50%.
✒️ California Enacts First U.S. AI Safety Disclosure Law
Governor Gavin Newsom has signed SB 53, the first U.S. law requiring major AI companies to disclose how they’re mitigating risks like loss of control or bioweapon misuse. The law mandates public safety reports, incident disclosures, and whistleblower protections—backed by $1 million fines for noncompliance.
🪧 POWERED BY CLOCKWISE
Introducing the First MCP Server for Time
Clockwise MCP helps AI schedule time like a human would. Schedule meetings, optimize your work calendar, and intelligently block time for tasks. Available today in Claude, Cursor, Relay, and more.
⚜️ SOVEREIGNTY
South Korea Bets Big on Homegrown AI to Cut Foreign Dependence
South Korea is investing ₩530 billion (~$390M) in five domestic companies—LG AI Research, SK Telecom, Naver Cloud, NC AI, and Upstage—to build sovereign AI models tailored to Korean language and culture. The initiative is part of a strategic push to reduce reliance on U.S. tech giants like OpenAI and Google, while enhancing national security and control over AI-era data flows.
Each company brings distinct advantages: LG is leaning into efficient, industry-specific models; SK is scaling AI through telecom infrastructure; Naver boasts a rare AI full stack; and Upstage is punching above its weight with a lean, high-performing model. The government will winnow the field over time, funding only the top two to lead Korea’s AI charge. → Read the full article here.
Enjoying our newsletter? Forward it to a colleague—
it’s one of the best ways to support us.
👩🎓 CAREERS
Gen Z Isn’t Waiting for AI to Take Their Jobs, They’re Pivoting
While pundits debate AI’s impact on the future of work, Gen Z is already reshaping their career paths—often without fanfare. Faced with the reality of automation, many are turning toward “AI-proof” jobs in trades, healthcare, and education, with 53% considering skilled labor and 47% looking to people-focused fields.
Side hustles, once branded and broadcast by millennials, are now practical defaults: thrift reselling, furniture restoration, and more. This isn’t disengagement—it’s a low-key, high-agency strategy for career resilience. In a noisy discourse, Gen Z’s quiet adaptability may be the smartest signal of all. → Read the full article here.
🦾 ROBOTICS
MIT’s “Steerable Scene Generation” Uses Gen AI to Build Smarter Robot Training Worlds
MIT and Toyota Research Institute researchers have developed a generative AI system called Steerable Scene Generation that creates diverse, physically realistic 3D environments for training robots. Unlike static or handcrafted simulations, this system uses diffusion models and reinforcement learning to build virtual scenes that mimic real-world physics—think cluttered kitchens or complex restaurant tables.
By steering generation with Monte Carlo Tree Search, it can adapt scenes dynamically to training goals, vastly improving realism and variability. The result: scalable, flexible training data that could bridge the gap between simulation and deployment in real-world robotics. → Read the full paper here.
𝕏 TWEETS
Claude Sonnet 4.5 runs autonomously for 30+ hours of coding?! The record for GPT-5-Codex was just 7 hours. What’s Anthropic’s secret sauce? — Yuchen Jin (@Yuchenj_UW) 5:03 PM • Sep 29, 2025
@Yuchenj_UWView post on X🛰️ NEWS
What Else is Happening
💼 WFH Holding Back U.S. Tech? Ex-Google CEO Eric Schmidt says remote work weakens America’s edge against China’s hardcore office culture, warns tradeoffs are needed to stay competitive.
✂️ Lufthansa Cutting 4,000 Jobs Amid AI Shift: Germany’s biggest airline is slashing staff as it leans on AI to streamline operations and cut costs after pandemic-era overexpansion.
🧩 OpenAI Adds Parental Controls to ChatGPT: Parents can now manage their kids’ ChatGPT access with new tools — including content settings and chat history visibility tied to age verification.
💰 AI Billing Startup Paid Raises $21M: Manny Medina’s new venture helps companies monetize AI agents—ditching seat-based pricing to unlock scalable revenue models. Valuation already tops $100M.
📊 Ex-Microsoft Duo Tackles Excel Dependence: Maximor, their new AI startup, raised $9M to automate finance workflows, reducing manual work across global enterprises.
📟 PROMPT OF THE WEEK
Write Your User Manual
Have ChatGPT analyze your conversations to draft a guide on your communication style, preferences, and quirks. Shareable with teams so they know how to work best with you.
Analyze my conversations to draft a guide on my communication style, preferences, and quirks — something I can share with my team so they know how to work best with me.🧰 TOOLBOX
Trending AI Tools
🗒️ FEEDBACK
Help Us Get Better
What did you think of today's newsletter? |
That’s a Wrap!
📢 Want to advertise with Forward Future? Reach 450K+ 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
🧑🚀 🧑🚀 🧑🚀 🧑🚀
