Digesto

For decades, banks and large financial institutions have served as the backbone of our economic systems. Their services were essential, their expertise exclusive, and their infrastructure unrivaled. But today, a new force is emerging that challenges their dominance not through confrontation, but through quiet, scalable substitution:  AI agents . AI agents are software systems that can reason, fetch data, analyze financial instruments, and produce actionable insights—all without human intervention. They don’t just replicate the work of a junior analyst or a robo-advisor; they represent a fundamental shift in who controls financial knowledge, access, and execution. This shift has major implications. It threatens to erode the fee-based models that many financial institutions rely on. It undermines the very idea of centralized financial authority. Most significantly, it lays the groundwork for a splintered financial ecosystem: a future where finance is no longer controlled by a fe...

A Surprising Personal Discovery in Frame Proportions While designing a minimalist aluminum bicycle frame, I stumbled upon something remarkable that appears to have gone unnoticed in cycling literature: the relationship between front-center and chainstay length across most mountain bike frames approximates the golden ratio (φ ≈ 1.618). This mountain bike features an almost exact golden ratio. Image: pexels.com - Jahangeer Bm - Superimposed Golden Rule vector: Wikimedia commons The front-center is the horizontal distance from the front wheel axle to the bottom bracket, while the chainstay length runs from the bottom bracket to the rear wheel axle. When you divide the front-center by the chainstay length across various bicycle types, a consistent pattern emerges. Road bikes typically show front-centers of 680-720mm with chainstays of 410-420mm, yielding ratios between 1.62 and 1.71. Mountain bikes display front-centers of 700-750mm and chainstays of 430-450mm, giv...

In an era where we casually share vacation selfies, location data, and even our DNA and heartbeat with apps and platforms, one thing we continue to fiercely protect is… our suitcase. What's in a suitcase? - Images: Grok The average piece of checked or carry-on luggage contains nothing revolutionary: a few outfits, some toiletries, maybe a laptop or paperback novel. Yet, we insist on opaque bags, as if the contents were deeply private. It’s a habit we’ve inherited, not one we’ve ever really thought about. And it may be time to ask: why? Privacy Theater: A Game We Keep Playing This instinct to hide our belongings is less about protecting personal information and more about repeating a social ritual. We cover our bags not because we need to, but because that's what everyone does. It’s behavior on autopilot. But here’s the irony, we’ve already surrendered most of our truly private data . We have already surrendered everything which is important: Our phones track us, our...

Many users choose Linux for reasons of control, transparency, and privacy. So it's worth paying attention when software quietly undermines those principles. Recently, I discovered that Dell’s Linux Assistant for Ubuntu contains built-in telemetry code,  code intended to log and send metrics about user behavior. The kicker? It fails silently due to a minor Python bug. The result is what you might call an accidental privacy enhancement . Hidden Telemetry Dell’s Linux Assistant is supposed to make life easier for Ubuntu users running Dell hardware. It manages drivers, firmware, and support tools. Nowhere in the user interface or documentation does it mention telemetry. Yet, digging into its code  or, in my case, encountering a stack trace reveals that it tries to log an event when the user closes the app: QueueMessageClient(...).LogLogMetricEvent("FRONTEND", "Dell Linux Assistant closes") This function uses a scheduler to send telemetry data: usage logs,...

* Reclaiming Meaning in the Age of Instant Everything* I’m from early Generation X, raised with a particular dream: that AI and robotics would *finally* liberate humanity from drudgery. Not to make us lazier, but to give us the freedom to pursue *nobler* things: philosophy, art, reflection, even idleness with purpose. That dream still flickers. But lately, it’s flickering in the shadow of something darker: not a future of liberation, but one of numbing convenience, where meaning is bypassed in the rush for efficiency, and the effort that once made life feel rich is treated like a bug in the system. A recent article summed up this fear powerfully. It argued that we're systematically destroying the biological reward system that makes effort feel worthwhile. Not by accident, but *on purpose*. In our quest to eliminate friction from life, we’ve also eliminated the very struggle our brains are wired to reward. The satisfaction of cooking, creating, learning, and slowly improving ourselv...

As "artificial intelligence" grows in power and generality, the question of how to align these systems with human values, or at the very least, with human intentions, becomes existentially important.  Most of the contemporary alignment conversation focuses on optimization, feedback, and safety: how to guide a model’s behavior using techniques like reinforcement learning from human feedback (RLHF), red-teaming, or reward modeling. These are all valid concerns. But what if the deeper issue lies not in the process, but in the substrate? What if the problem isn’t just that our data is noisy, biased, or incomplete, but that deception itself is woven into the very structure of human thought? This isn't a novel claim in philosophy or literature. From Plato’s cave to Freud’s defense mechanisms, from Nietzsche’s “will to illusion” to postmodern critiques of ideology, thinkers have long suspected that much of what we call “knowledge” or “tru...