Weekly Reading List: 08/11/2025

Things I Read This Week That Made Me Think

Lately, I’ve been thinking about how much of modern life runs on outsourced discernment.

Who we date, what we buy, what we build, what we believe, it’s all increasingly shaped by systems that claim to know us better than we know ourselves. The week’s reading took me across that entire spectrum: from AI-generated intimacy to the erosion of moral agency in public life. Together, they sketch a portrait of a culture that’s learning to live with the aftertaste of automation: the weird mixture of relief, dependency, and dread that comes from letting something else decide for you.

It’s not just about technology. The AI dating features and the collapsing startups are symptoms, but the pattern is older. Whenever our tools begin to think—when they start anticipating what we want instead of simply executing what we ask—we lose a small part of our friction with the world. And friction, I think, is where real choice lives. It’s what separates intent from instinct, momentum from meaning. Reading through these pieces this week, I kept noticing how often people are building systems to remove friction, and how often that absence comes back to haunt them.

Because when you remove friction, you also remove feedback. And without feedback, you stop learning.

1. The New York Times on AI Dating Apps

The article about AI companions and dating assistants felt less about romance and more about the future of calibration. Facebook Dating’s new AI tool can now match you with “a brunette who works in tech and lives in Brooklyn.” The more precise the filters, the more narrow the possibilities. These apps are reconditioning our brains to expect frictionless intimacy: pre-screened, pre-answered, pre-imagined. It’s strange to think that our emotional lives are being A/B tested, but that’s what’s happening. The piece left me wondering whether a generation trained on compliant AI partners will know how to handle the beautiful, frustrating resistance of real people.

2. TechCrunch on Cluely’s Pivot

Roy Lee’s Cluely started as a rage-bait experiment in cheating-“a tool to cheat on everything”- and has now pivoted into an AI meeting assistant. The story is a neat encapsulation of the startup delusion that virality equals validation. Lee’s admission that they “maybe launched too early” isn’t a confession; it’s a worldview. The piece reminded me that virality is entropy disguised as progress. Real compounding happens in rooms, not trending tabs. Every few years, tech gets high on its own distribution and forgets that hype is not a moat. This was one of those stories that makes you want to say: slow down.

3. 404 Media on the FBI and Archive.is

The FBI’s attempt to unmask the owner of Archive.is is a test case for the tension between preservation and power. Archive.is has been a refuge for those documenting inconvenient truths, and that’s exactly what makes it dangerous to institutions. What struck me most was how the act of archiving (once a bureaucratic duty) has become a radical one. The infrastructure of memory is now adversarial. We live in an era where even saving a page can be seen as defiance.

4. Inside Cursor (JoinColossus)

The profile of Cursor reads like a fever dream for anyone who has ever worked in early-stage tech. Their hiring process, run through a Slack channel where people drop names of talent they admire, sounds chaotic, cultish, and effective. It’s a rare portrait of a company scaling curiosity instead of compliance. The line that stayed with me: “The biggest existential risk to Cursor is that its early commercial success could distract from taking the biggest swings possible.” It’s true for individuals, too. Sometimes the hardest part of winning early is not mistaking traction for destiny.

5. Fidji Simo on AI as Multiplication, Not Subtraction

Fidji’s essay argues that most companies treat AI as a cost-cutting tool rather than a creativity multiplier. Her framing (AI as abundance, not austerity) felt clarifying. She describes how OpenAI uses its own systems to multiply human capacity instead of replacing it. What I took from it: the most interesting companies in the next decade won’t ask “what can AI do for us?” but “what does this let us attempt that was previously impossible?” Abundance is an attitude before it’s a capability.

6. Iter Intellectus on Digital Surveillance and Love

A study on 322 couples revealed how phone-checking erodes both trust and biology: cortisol spikes, oxytocin drops, sexual polarity collapses. Reading it was like watching modern attachment theory meet endocrinology. We talk about boundaries as emotional hygiene, but this reframed them as biochemical necessity. The takeaway was brutal in its simplicity: checking your partner’s phone is physiological self-sabotage.

7. A Guide to X for Deep Tech Founders

This guide for deep tech founders on how to engage on X (and not sound dead inside) was both tactical and cultural. What makes it interesting is how it reframes social media as an engineering problem: feedback loops, signal-to-noise ratios, community resonance. It reminded me that visibility is a skillset, not a personality trait. The smartest builders often underestimate how much storytelling multiplies credibility, and how “boring” businesses need narrative oxygen more than any other kind.

8. AI’s Hidden Crisis (Shae Omonijo)

This was one of the sharpest essays I’ve read on the political economy of data. The writer points out that the true bottleneck in AI is consent. The open web has been scraped dry, and what remains, archival, analog, private, is locked behind ethical and legal constraints. What resonated was her optimism: that this scarcity could create new, dignified work in digitization, metadata, and rights management. The future of AI might depend less on GPU clusters and more on librarians.

9. Mass Intelligence (One Useful Thing)

Ethan Mollick’s framing of “Mass Intelligence” is perfect: a billion people suddenly armed with tools that used to be the privilege of experts. Every institution was built for a world where intelligence was scarce; now it’s cheap, abundant, and destabilizing. The piece made me think about how we’re still missing the language for this shift. We don’t yet know how to measure wisdom in an age where everyone can simulate expertise. The institutions that survive this era will be the ones that learn how to turn ubiquity back into judgment.

10. The FT on Neom

A fascinating postmortem on a fantasy that physics couldn’t sustain. The Line was supposed to be Saudi Arabia’s architectural leap into the future, an urban corridor stretching through the desert. But the math broke before the money did. What lingered with me wasn’t the failure of ambition, but the failure of feedback. When power doesn’t permit pushback, gravity always wins.

11. Plum Data Labs on Cancer’s Economic Toll

This piece hit close to home. It quantified what families already know: that cancer bankrupts more than bodies, and that women, especially caregivers, bear the heaviest invisible costs. The numbers are damning, but the deeper insight was about systemic empathy: how our financial and healthcare infrastructures are still designed around idealized, male breadwinner households. Economic modeling without gendered reality is just elegant math that hurts real people.

12. Newslaundry on India’s Women’s Cricket Team

The women’s team won the World Cup, but there were no parades, no pay parity, no primetime celebrations. The report breaks down how “market forces” are used as moral cover for discrimination, that visibility itself is rationed. The story reminded me that merit doesn’t disrupt bias; money does. And money flows where narrative goes. Until we learn to tell women’s victories as national triumphs, they’ll keep being priced as exceptions.

13. The Point on Schattenfroh

A deep dive into a thousand-page postmodern novel that treats language like jazz. I don’t usually care for literary maximalism, but the reviewer’s framing of excess as redemption stuck with me. Lentz’s chaos isn’t just indulgence; it’s also rebellion against cultural obedience. The piece reminded me that art still has a moral role: to fracture inherited forms so something new can breathe. We forget that creation, at its best, is a kind of revenge.

14. Henrik Karlsson’s “Looking for Alice”

One of the most unexpectedly moving essays on love I’ve read in a long time. Karlsson’s argument, that you don’t love a category, you love a particular, is the most honest defense of non-cynical intimacy I’ve seen. His idea that “you can’t afford not talking to Alice” reads like a counterspell to the algorithmic dating story from earlier. If the AI era flattens desire into filters, this essay reintroduces randomness, specificity, and grace.

15. Vikram Shah’s Self-Interview

This strange, self-aware interview of a writer interviewing himself could have been a gimmick, but instead it lands as a meditation on performance, authenticity, and economic precarity. Shah treats writing as both devotion and delusion, something you do to justify your existence to yourself. I found it quietly devastating, especially his remark that he’s “constantly seeking opportunities to explain myself, even as I feel repelled by the act.” That’s the trap of all creative work in a performative age: to be seen, you have to translate yourself faster than you can understand yourself.

What ties these fifteen pieces together is a kind of crisis of agency. Whether it’s a dating algorithm, a data model, a government archive, or a literary monologue, everyone is wrestling with the same question: who gets to interpret reality?

The pattern that emerges is subtle but consistent. We’re automating judgment in the name of efficiency, compressing complexity into templates, and outsourcing taste to systems that reward predictability. But the cost of that convenience is comprehension. The more the world anticipates us, the less we have to confront it.

Maybe that’s why the pieces that lingered most (Karlsson on love, Omonijo on consent, Fidji Simo on abundance) were the ones that asked for something harder than optimization. They asked for care. Care is inefficient by design. It slows us down, forces context, makes space for reflection. It’s the only real antidote to automation because it requires participation.

So this week, I’m trying to resist the seduction of frictionless life. To remember that not everything that runs smoothly is running well, and that sometimes, the most human thing we can do is to stay with the lag.