OpenAI launches a way for enterprises to build and manage AI agents

Key Takeaway: OpenAI launched OpenAI Frontier, an end-to-end platform that lets enterprises build, connect, govern, and manage AI agents—including agents built outside OpenAI.

More Insights:

• Frontier is “open,” so companies can manage third-party agents alongside OpenAI-built ones.

• Agents can connect to external data and apps to execute real business tasks beyond OpenAI’s platform.

• Includes controls to restrict agent access and permissions—enterprise-grade governance is a core feature.

• Designed like employee management: onboarding, defined access, and a feedback/review loop to improve performance over time.

• Early customers cited include HP, Oracle, State Farm, and Uber; access is limited for now, with wider rollout planned.

Why it matters: AI agents aren’t useful because they can “chat”—they’re useful because they can act inside real systems, and that’s exactly where risk lives; whoever becomes the default layer for agent identity, permissions, monitoring, and improvement could end up owning the “operating system” of enterprise AI.

Superpower Daily Exclusive: You Could Be an Investor in High-Potential Startups Like These

Superpower Daily and Alumni Ventures are teaming up this week only to give readers early access to high-growth startup opportunities, including some of today’s most exciting AI, Deep Tech, Quantum Computing, Cybersecurity, and Space companies co-invested alongside top VC firms like Andreessen Horowitz (a16z), Bessemer, & Y Combinator.

You get:

• Curated deal flow of high-potential AI First startups

• AV is already investing alongside elite lead venture firms in these deals

• No cost to see deals

• No obligation to invest

Don’t miss your chance before access closes.

The tech firms embracing a 72-hour working week

Key Takeaway: A growing number of tech—especially AI—companies are openly promoting 70+ hour, in-person workweeks (a “996” culture) to move faster than competitors, despite strong evidence of burnout, health risks, and diminishing returns.

More Insights:

• Startups like Rilla and Browser-Use market extreme hours as a badge of ambition and “fun,” screening for people who want that intensity.

• The 996 model (9am–9pm, six days/week) surged in China, got political and legal pushback by 2021, but its mindset is resurfacing globally through AI’s competitive race.

• Venture pressure matters: some investors reward founders who grind 70–80 hours, but critics argue forcing employees to do the same confuses “time spent” with “output.”

• Health consequences are real: major global analyses link 55+ hour weeks to higher risks of stroke and heart disease, contributing to hundreds of thousands of deaths.

• Productivity often peaks around ~40 hours; beyond that, fatigue erodes performance—some research suggests 70-hour weeks can produce little more than 50-hour weeks.

Why it matters: The AI boom is turning “speed” into a workplace ideology—and when long hours become a hiring filter instead of a temporary sprint, innovation starts to look less like progress and more like a system that trades human health and sustainable creativity for the illusion of winning.

Elon Musk is getting serious about orbital data centers

Key Takeaway: SpaceX is pushing a massive “orbital data center” satellite network idea forward at the FCC, while Musk publicly argues that space-based solar power could make AI computing cheaper and claims space will become the best place to run AI within a few years.

More Insights:

• SpaceX filed plans with the FCC for a data-center-like network involving up to a million satellites, and the FCC has opened a public comment process.

• FCC chair Brendan Carr unusually amplified the filing on X, signaling political tailwinds if Musk remains aligned with Trump’s circle.

• A formal SpaceX–xAI merger strengthens the case that Musk is building integrated space + AI infrastructure, not just floating a concept.

• Musk’s central pitch: solar panels generate far more power in space, making it easier to scale computing off-planet.

• Critics note the weak link: power is only one cost, and servicing/repairing failing GPUs in orbit could be a major practical hurdle.

Why it matters: If “AI infrastructure” starts shifting off Earth, it won’t just be a tech upgrade—it’s a power shift: the future of computing could become dependent on launch capacity, spectrum access, and political regulators as much as chips and algorithms, turning orbital real estate into the next high-stakes battleground.

Other stuff

• This guy installed OpenClaw on a $25 phone and gave it full access to the hardware

• AI’s latest 20-something billionaire got his start at L.A. garage sales

• From Svedka to Anthropic, brands make bold plays with AI in Super Bowl ads

• From churches to chatbots: How AI is fusing with religion

• Man who videotaped himself BASE jumping in Yosemite arrested, federal officials say. He says it was AI

• AI Doesn’t Reduce Work—It Intensifies It

• AI.com bought by Crypto DOT com founder for $70mn in biggest-ever website name deal

Hope you enjoyed today's newsletter

⚡️ Join over 300,000 people using the Superpower ChatGPT extension on Chrome and Firefox.