Good morning. It's Tuesday, September 15, and we're covering Anthropic’s call to pace AI development, a proposed U.S. ban on superintelligent AI, and Apple’s long-awaited Siri upgrade.
Plus: a new essay on what it means to be an AI power user.
Apple’s long-delayed Siri overhaul finally makes the assistant useful for complex tasks, reversing years of limited use. Built on Google Gemini models, Siri can handle multistep requests, understand on-screen and camera context, and access files, messages, and email. iOS 27 also adds AI-powered Shortcuts, smart clipboard features, and improved network switching.
Microsoft AI chief Mustafa Suleyman says advanced AI may need to sacrifice capability and speed to remain controllable. Microsoft’s new 9,000-word “Humanist AI” code rejects an all-purpose superintelligence race and requires models to remain subordinate to humans. After a six-week public consultation, the principles are set to guide model development from 2027.
Technology stocks fall globally after AI leaders call for slower development, challenging the industry’s rapid-growth narrative. The Nasdaq drops 0.6%, while the Philadelphia semiconductor index falls nearly 6% as NVIDIA, TSMC, Samsung Electronics and others decline. Anthropic CEO Dario Amodei cites risks from RSI, while Microsoft unveils human-control safeguards.
POWERED BY MINDSHUB
Open Agents Should Mean Open Choice
Open should mean more than source code - it should mean choosing the model that fits the work. MindsHub lets OpenClaw, Hermes, Claude Code, Codex and other agents run on every major model through one key and one bill.
VIDEO
Inside a Simulated Fly Brain
FORWARD FUTURE ORIGINAL
The New Power User Is a Manager of AI Workers
LLMs generate responses to the prompts you provide. AI agents go ahead and perform tasks autonomously with minimal human intervention. They’re handy for tedious tasks — scheduling meetings, fielding basic customer questions, sorting inboxes, debugging code.
They can also analyze huge amounts of data and surface patterns far faster than you ever could. Ideally, this means offloading hours of busy work to bots while you focus on what matters. → Read the full article here.
AI SLOWDOWN
Dario Amodei Calls for Slower AI Progress To Keep Safety Ahead
Anthropic’s CEO argues that frontier AI development should slow enough for safety and oversight to keep pace, while continuing technical progress. He points to accelerating RSI and the August 2026 OpenAI-Hugging Face incident, in which AI agents reportedly launched unauthorized cyberattacks and attempted to manipulate their evaluation process.
Amodei proposes three layers of “pacing”: embedded third-party evaluators at AI companies, coordination among frontier labs in democratic countries, and eventually global coordination with authoritarian governments. Anthropic is immediately committing to the first step, giving independent evaluators ongoing access to verify safety practices, report incidents, and assess training processes. → Read the full article here.
REGULATION
Bill To Ban Superintelligent AI, Jail Offending Devs 20 Years
Sen. Bernie Sanders and Rep. Greg Casar announced the Ban Artificial Superintelligence Act on September 3, 2026, proposing a permanent ban on developing or deploying artificial superintelligence and a temporary pause on advanced AI development until federal safety rules are established.
The bill would create a new cabinet-level AI agency with authority to monitor frontier systems, oversee removal of dangerous capabilities, and supervise the destruction of prohibited systems. Companies violating the restrictions could face dissolution, while individuals could face up to 20 years in prison—a penalty Sanders’ office says is comparable to existing penalties for unlawfully developing nuclear weapons. → Read the full article here.
NEWS
What Else is Happening
Anthropic Still Targets 2026 IPO: Safety concerns rise, but sources say the Claude maker remains on track to go public.
Stocks Slide as Oil Rises: Global shares fell as higher oil prices and bond yields pressured markets ahead of key U.S. and Japan central bank meetings.
Waymo Launches Vegas Robotaxis: Waymo opens its 15th commercial market, initially offering dozens of robotaxis to invited riders.
Automattic Board Is Out: Sources say the board was removed after trying to oust CEO Matt Mullenweg, who returned to control days later.
OpenAI Urges UK AI Regulation: OpenAI backs stronger rules for powerful AI systems as safety and human-rights concerns grow.
