Tools for progress #57
How ChatGPT is making you dumb, A fungible worldview, The case for being the culture hire
We (humans) read hundreds of articles on company building, angel investing, and self-management and curate the best ones into a weekly summary—helping founders and operators stay on the top of their game.
Better thinking
How ChatGPT is making you dumb (10 minute read)
MIT found that ChatGPT can make you forgetful when used the wrong way. Most people who let AI do all the writing couldn’t remember what they just wrote minutes later. The fix is simple, you have to think through your ideas first, then use ChatGPT to improve them. Then use your brain to really tailor the output to your needs. Don’t use the initial output.
AI in the classroom (13 minute read)
AI could revolutionize education through personalized tutoring. Benjamin Bloom proved in 1984 that one-on-one instruction dramatically improves learning but was too expensive to scale. AI finally makes this accessible to everyone. Early studies show promise, though concerns about cheating and dependency remain. Success depends on using AI as a learning assistant rather than a replacement.
UK Gov “Modern Industrial Strategy” (very long read)
The Government’s new Modern Industrial Strategy landed this morning - a ten-year blueprint for backing growth sectors and attempting to tackle some of the UK’s deepest structural problems. It includes more funding to startups via BBB, measures to tackle high energy costs, and boosting regional hubs. It’s a long read.
Operational tactics
Retention: the situationship of SaaS (6 minute read)
SaaS retention has become like a situationship where you have to keep proving your worth every quarter as AI disrupts entire markets. This guide covers four retention levers including defining real “aha” moments beyond setup tasks, driving adoption of features that actually matter, preventing payment failures, and strategically winning back churned users. Churn isn’t just a metric problem but your most honest feedback about what users don’t value.
The $170bn A/B testing machine (26 minute read)
This is a long but intriguing read about Booking.com running 1000s of A/B tests daily on their website. They turned two $294M acquisitions into a $170B+ travel empire by letting data trump executive opinions to optimize every conversion. Small, systematic improvements compound into massive competitive advantages over time.
Fighting for context (2 minute read)
Legacy software companies are fighting to protect their data as AI reshapes workflows and threatens their dominance. Instead of building open platforms like before, companies like Slack are restricting API access while others acquire AI startups outright. For startups depending on these partnerships, platform risk has never been higher as incumbents choose between collaboration and self-preservation.
Refer and we’ll send you our favorite books as a “thank you” for spreading the word.
Venture investing
Asset quality drives secondary returns (5 minute read)
Chasing discounts in venture secondary markets is often the wrong strategy. Counterintuitively, funds bought at 15%+ discounts actually underperform those with smaller discounts. Since venture is a winner-take-most market where the top 100 exits drive nearly half of all returns, success comes from getting exposure to elite companies at reasonable prices rather than bargain hunting mediocre assets.
A fungible worldview (11 minute read)
This piece looks at how venture capital’s focus on doing more deals faster creates a “fungible worldview” where individual choices don’t seem to matter. Cluely just raised $15M from a16z using controversy and attention-grabbing stunts as their main strategy. The author argues this approach lets firms fund anything that makes money regardless of values. Every choice shapes who you become, so speed shouldn’t trump principles.
The opportunities in consumer AI (12 minute read)
Consumer AI is overlooked while everyone chases enterprise AI deals. This piece maps AI opportunities across major consumer spending categories like shopping, travel, and health, plus explores why chat interfaces aren’t always the right choice. Unlike past consumer apps that prioritized growth over revenue, AI’s high costs mean these startups must prove people will actually pay from day one, making it much easier to get a read on winners earlier.
Managing your career
Why I quit my software engineering job at Google (10 minute read)
A Google engineer quit after four years to chase entrepreneurship. The job felt like just a stepping stone, and he realized no amount of promotions or salary bumps would create real passion for working at someone else’s company. Sharing this for those who are on the fence, wondering if their big tech job will really make them happy.
The state of engineering leadership in 2025 (7 minute read)
This survey of 600+ engineering leaders reveals mixed feelings about AI’s impact on the industry in 2025. While hiring improved from 2024’s layoffs, 60% of leaders say AI hasn’t boosted team productivity and 51% view AI’s industry impact negatively. Engineering leaders are working longer hours with more responsibilities, struggling with team motivation amid uncertainty about AI’s future role in software development.
The case for being the culture hire (3 minute read)
As the tech world orgiastically competes for who can be the most disciplined, most execution focused, and most technical, Strachan pitches something else: just be fun to work with. Because as execution gets commoditized by tools like ChatGPT, what remains is how you make people feel. Don’t be boring. Become the culture hire.