Friday, April 17, 2026
Code is free and agents are the new workforce.
April 17 · 7 videos
OpenAI banned code editors for nine months.
Ryan Lopopolo says code is now free.
ServiceNow predicts 2.2 billion agents entering the workforce.
Leopold Aschenbrenner made $2.5 billion on AI infrastructure.
Claude Opus 4.7 is out but Anthropic is holding back.
OpenClaw hit 30,000 commits in record time.
The bottleneck has shifted from implementation to attention.
“Implementation is no longer the scarce resource of what it means to do the job of software engineering. Code is free.”
Harness Engineering: How to Build Software When Humans Steer, Agents Execute
Ryan Lopopolo · AI Engineer · 46 min
Watch on YouTube →Ryan Lopopolo from OpenAI introduces Harness Engineering, a paradigm where implementation is no longer scarce. He argues that engineers must transition into managing fleets of thousands of agents.
- OpenAI engineers were banned from using code editors for nine months to force reliance on agents.
- The team currently consumes 1 billion output tokens daily to automate the development lifecycle.
- Software engineering is shifting from implementation to system design and meta-programming through instructions.
- A single engineer can now theoretically manage a fleet of up to 5,000 virtual agents.
- Repositories must be redesigned to be agent-legible using prompt-injected lints and modular architectures.
- The goal is a fuzzy compiler model where humans define requirements and machines execute parallel tasks.
Scaling Global Organizations in the Age of AI with ServiceNow CEO Bill McDermott
Bill McDermott · No Priors: AI, Machine Learning, Tech, & Startups · 57 min
Watch on YouTube →ServiceNow CEO Bill McDermott outlines the shift toward an agentic business model. He explains why enterprise platforms remain the essential control towers for AI integration.
- McDermott predicts 2.2 billion agents will enter the global workforce in the coming years.
- ServiceNow already handles 90 percent of its customer service cases through autonomous agents.
- Rebuilding established enterprise platforms with LLMs is 10 times more expensive than using existing systems.
- The business focus is moving from systems of record to systems of action where workflows actually close cases.
- ServiceNow currently manages 85 billion workflows in flight across its global platform.
- Leadership in the AI age requires high emotional intelligence to provide the ambition software cannot replicate.
Claude Opus 4.7 Is Out... But Anthropic Is Hiding Something
Ejaaz · Limitless Podcast · 30 min
Watch on YouTube →This episode covers the release of Claude Opus 4.7 and the massive financial gains in AI infrastructure. It highlights the strategic withholding of more powerful models due to security concerns.
- Claude Opus 4.7 delivers a 3x improvement in vision benchmarks over the previous 4.6 version.
- Anthropic is reportedly withholding a more powerful model called Mythos due to cybersecurity and compute constraints.
- Leopold Aschenbrenner realized 2.5 billion dollars in gains this week through his AI infrastructure fund.
- Allbirds stock surged 850 percent in a single day after announcing a pivot to GPU provisioning.
- Anthropic secondary shares are trading at a valuation of 850 billion dollars before an official IPO.
- Microsoft is becoming tool agnostic by integrating Claude into Word and Excel despite its OpenAI partnership.
State of the Claw — Peter Steinberger
Peter Steinberger · AI Engineer · 44 min
Watch on YouTube →Peter Steinberger discusses the explosive growth of OpenClaw and the management of open-source AI foundations. He warns of the security risks inherent in agentic systems.
- OpenClaw reached 30,000 commits and 2,000 contributors within its first five months of existence.
- The project receives an average of 16.6 security advisories per day, many of which are AI-generated slop.
- Agentic systems with access to private data and untrusted content create a legal trifecta of risk.
- The OpenClaw Foundation uses a Switzerland model of neutrality to allow collaboration between Nvidia and Microsoft.
- Taste is the ultimate moat in software development to identify and eliminate generic AI smell in code.
- Personal agents provide data sovereignty by allowing users to bypass the silos of large tech companies.
How Acquisition.com Makes Money
Alex Hormozi · Alex Hormozi · 16 min
Watch on YouTube →Alex Hormozi breaks down the monetization ecosystem of his family office. He explains how free high-value content builds a high-trust brand for private equity deals.
- Acquisition.com manages a portfolio of 10 companies generating over 250 million dollars in annual revenue.
- The strategy relies on making free content better than the paid products offered by competitors.
- Aggregate bargaining power allows the community to negotiate superior rates with vendors and agencies.
- The model funnels business owners from free books into advisory practices ranging from 5,000 to 135,000 dollars.
- Hormozi aims to bridge the credibility gap by having active practitioners teach business instead of theorists.
- The long-term vision is a B2B Disney where software, real estate, and education reinforce each other.
Lying or Incompetence? | The Starmer-Mandelson Scandal
Rory Stewart · The Rest Is Politics · 21 min
Watch on YouTube →Rory Stewart and Alastair Campbell analyze the political fallout of Peter Mandelson's failed security vetting. They critique Keir Starmer's management of high-level personnel.
- Peter Mandelson failed the Developed Vetting security process for his appointment as US Ambassador.
- Keir Starmer has lost two chiefs of staff and three communications directors in less than two years.
- The Foreign Office allegedly overruled security advice without informing the Prime Minister's office.
- High staff turnover is cited as evidence of a lack of grip on organizational culture and personnel.
- The podcast argues that leaders cannot outsource responsibility for top-tier diplomatic appointments.
- Throwing subordinates under the bus erodes institutional loyalty and prevents honest upward communication.
The Psychology of Addictive Discipline
Rob Dial · The Mindset Mentor Podcast · 15 min
Watch on YouTube →Rob Dial explains the transition from forced discipline to identity alignment. He describes the neurological mechanisms that make consistency difficult.
- Identity Lag is the period where behavior changes but the brain remains rooted in the old self-image.
- The brain prioritizes efficiency, energy conservation, and familiarity, which creates resistance to new habits.
- Discipline is a bridge to identity: once the identity shifts, the perceived effort of the task disappears.
- Building evidence through repetition is the only way to convince the brain to update its self-image.
- Success in long-term projects depends on shifting from push motivation to pull identity alignment.
- Discomfort is a signal of transition and growth rather than a signal to stop the activity.
References
PeopleRyan Lopopolo (https://x.com/_lopopolo) · Bill McDermott (https://x.com/BillRMcDermott) · Peter Steinberger (https://x.com/steipete) · Rob Dial (http://coachwithrob.com) · Leopold Aschenbrenner · Alex Hormozi · Keir Starmer · Peter Mandelson · Rory Stewart
ToolsServiceNow · OpenClaw · Claude Opus 4.7 · NVIDIA CUDA · Skool · Claude Mythos
PapersSituational Awareness