OP 320: When Speed Replaces Thinking

A bit of a different OP this week. I hope you enjoy it!

Author

Darren Herman
March 10, 2026

Happy Tuesday, OP Community.

Last week was awesome - capped off by spending the weekend in Providence, RI, at a dance competition for my daughter until late Sunday evening.  She did terrific, and I’m super proud of her.

Earlier in the week, we hosted our Inaugural Padel Invitational for Silicon Alley Sports. I had only played padel once before and had a blast.  I think 99% of us did - well, probably 100% did, but there was a twisted ankle on the court. Huge shoutout to the OP community who came out and played… You know who you are.

There’s a large OP contingent at the flagship pickleball and golf invitational on May 18 - make sure to grab your spot before it sells out… even if you do not play, you can still attend.

Going to change up this OP a bit today… because why not? I was able to read a few articles while at the dance competition and wanted to include them here.  I hope these articles and podcasts help add value to your week.

  • Do Not Outsource The Thinking (Michael Stricklen)

    According to Stricklen, the problem that coding agents using AI technology present is that, although the speed is excellent, the code that is produced may be of poor quality because the developer is not putting in the effort of rigorous thinking and verification. The problem Stricklen sees is an "asymmetry" that allows the developer to avoid the slow evaluation process in the name of speed. 

  • Staying Human In the Age of AI  (Rishad Tobaccawala)

    Tobaccowala suggests that AI is not an incremental but rather a structural change, in that it is creating a new brain rather than extending the human brain. It is not just extending knowledge, as previous technologies were, but rather making information free and non-monetizable, changing 20-30% of what we currently do in an industry that doubles in power every seven weeks. However, the way forward is not just about making it more efficient but rather focusing on what the machine can't do: intuition, insight, imagination, emotional intelligence.

  • How Universities Can Prepare Graduates For An AI-Driven World (Nido Qubein)

    This article examines how universities need to rethink curricula to prepare graduates for an AI-transformed workplace. It focuses on embedding AI literacy and practical life skills into higher education.

I hope you have a wonderful week.

Be well, do good.
Darren

Told yah I was changing things up this week…

Your AI Dashboard Is Lying to You (And What to Measure Instead) (Jeremy Utley's Blog)
AI adoption is not a training problem.. It's a deletion problem; leaders must convert dashboards into uncomfortable but productive conversations about real behavior change, not just metrics.

Episode 160: Matthew Egol on Why We Need an AI-Specific Industry Association (Marketecture Podcast)
Industry leaders discuss agentic advertising standards, AdCP governance, and how AI agents are reshaping marketing planning, creative, and measurement across the industry.

What Every Company Can Learn from Private Equity (Harvard Business Review)
PE-backed companies consistently deliver faster, more substantial gains than peers through six core practices: recurring full-potential due diligence, matched management teams, operational streamlining, eliminating unprofitable revenue, executing with accountability, and treating leadership time as scarce—fostering sharper strategic focus and stronger resource alignment.

Private Equity Revival Accelerates, but the Next Winners Will Be Operational Outperformers (InvestmentNews)
Firms must now deliver much stronger earnings expansion ("12 is the new 5") through immediate post-acquisition transformation initiatives, including operating partners, data analytics, digital transformation, and AI tools.

Evolving models and monetization strategies in the new AI SaaS era (McKinsey)
As AI systems perform work rather than support it, consumption-based pricing aligns customer value with the number of units of work completed, while founders must balance compute costs against pricing expectations.

Tunguz shares nine practical observations from a year of building AI agent systems, offering builders a framework for choosing among frontier models, fine-tuned specialists, and optimizing tool-calling reliability for production workflows.

Familiar productivity approaches (restructuring, downsizing) are hitting diminishing returns; leaders must redesign workflows and decision-making rather than pursuing structural cost-cutting.

Enterprises will realize they need their own forward-deployed engineers to work through data, process, architecture, and AI automation, as these engineers will know the business and industry better than borrowed resources from software vendors.

CMO and AI strategist breakdown critical shifts, including changes in agency relationships, generative search discovery, AI agents owning workflows, and human judgment becoming the ultimate competitive advantage.

Thanks again for subscribing to the OP!

What did you think of today's OP letter?

Login or Subscribe to participate in polls.