Artificial-Intelligence on Aneesh Sathe

The End of Identity: AI, Plasticity, and the Divergence Machine

Sun, 08 Mar 2026 06:28:21 +0000

Over the past year, the Contraptions club has been reading through history—from Giordano Bruno and Montaigne to Spinoza, Adam Smith, and Hume. We are now using using Venkatesh Rao’s Divergence Machine framework as a lens to make sense of the modern world.
For context, Venkat posits that human history operates through massive “world machines”. The “modernity machine” was constructed around 1200 and operated at a steady plateau of capability from 1600 to 2000. It is now in a state of rapid, partially scheduled disassembly. In its place, the “divergence machine” was constructed around 1600 and has been operating in fully deployed mode for about 25 years.
Looking at this transition through the philosophers we’ve studied, my feeling is that over the centuries, we’ve witnessed a gradual peeling back of the layers of imagination that were once heavily layered on top of nature. We can map this peeling back directly to Rao’s divergence concepts.

The Kernel and the Ark

Sun, 11 Jan 2026 08:28:42 +0000

I. The Wall and the Infinite #

It is possible that the history of the modern West hinges on a single, melancholic misreading of Voltaire. When Candide, exhausted by the Lisbon earthquake and the brutalities of the Seven Years’ War, finally withdraws to the banks of the Propontis to utter his famous dictum—“Il faut cultiver notre jardin”—he is not proposing a program of agricultural management. He is issuing a plea for containment. To cultivate a garden, in the shadow of such overwhelming chaos, is an act of stoic resignation. It is an admission that the world is too vast, too violent, and too unintelligible to be governed by reason. One builds a wall against the infinite, and within that limited circumference, one tends to the soil. The garden is a refuge from nature.

The Deep Dark Terroir of the Soul

Fri, 26 Dec 2025 02:17:37 +0000

This is the third and final part of the Thicket Series:
Part 1: Logic of the Thicket and the Unsearchable Web
Part 2: The Architecture of Resistance

The history of the working subject might be best understood not as a ledger of wages or a sequence of industrial breakthroughs, but as a study in the migration of the Master. In the eighteenth century, the Master was a concrete presence, a figure residing in the castle or the cathedral, distinct from the worker by a physical and social chasm. One knew where the authority lived because one could see the smoke from its chimneys. By the nineteenth century, this figure had moved into the factory office, closer to the rhythm of the machine but still identifiable by the suit and the watch. The twentieth century saw a further dissolution; the Master became atmospheric, blending into the very walls of the institutions that housed us—the schools, the hospitals, the barracks.

The Architecture of Resistance

Tue, 23 Dec 2025 05:38:32 +0000

The seventeenth-century Hague, the mid-twentieth-century Levant, and the digital terraforming of 2025 have a shared preoccupation with the “Average.” Whether it is the theologian’s way or predictive stats, control begins by smoothing out the landscape. The project of power is a project of cartography and illumination—an attempt to banish the dark corners where the unmapped might grow. Thus, the history of resistance, of being “against the world”, is less a history of rebellion than a history of seeking cover.

The Shelter as Epistemic Engine

Wed, 17 Dec 2025 06:35:42 +0000

This is a continuation of my ongoing exploration of places and spaces. Previously: We need homes in the delta quadrant, Thinking with places, Problems are places questions are spaces.

Introduction: The Terror of the Open Field #

We tend to think of “Space” as a vacuum—an emptiness waiting to be filled. But geographically and philosophically, Space is actually a condition of high-entropy potential. As Yi-Fu Tuan famously articulated, space is “freedom,” but it is also “possibility without orientation.” It is the open field where everything is possible, which means nothing is yet distinct.

The Tortured Artist Is So Yesterday

Mon, 08 Dec 2025 05:08:29 +0000

41 years ago, Samuel Lipman wrote that an artist’s life is a “constant—and constantly losing—battle” against one’s own limits. That image has lasted because print culture taught us to imagine the artist as a solitary figure whose worth is measured by the perfection of a single, final work. Print fixed texts in place, elevated the individual author, and made loneliness part of the creative job description.

That world is slipping away.
And with it, the tortured artist.

Four Early-Modern Tempers for a World That Can Summon Itself

Sat, 06 Dec 2025 10:12:37 +0000

This is a partial synthesis of the books read through 2025 in the Contraptions Book Club.

We live in a moment when the whole of human culture has become strangely available, no longer just an archive but something that behaves like a responding presence. A sentence typed into a search bar or messaging window returns citations and, more strikingly, continuations: pastiche, commentary, new variations of ideas that never existed until the instant we requested them. The canon now behaves more like a voice than a library. It is easy to treat this as convenience, yet summoning culture alters our relation to meaning in ways we are only beginning to see. The question is no longer whether we can find the relevant text, but what it means to think in a world that can generate its own echoes.

Why Every Biotech Research Group Needs a Data Lakehouse

Tue, 29 Jul 2025 06:04:54 +0000

start tiny and scale fast without vendor lock-in

All biotech labs have data, tons of it. The problem is the same across scales. Accessing data across experiments is hard. Often data simply gets lost on somebody’s laptop with a pretty plot on a poster as the only clue it ever existed. The problem is almost insurmountable if you try to track multiple data types. Trying to run any kind of data management activity used to have large overhead. New technology like DuckDB and their new data lakehouse infrastructure, DuckLake, try to make it very easy to adopt and scale with your data. All while avoiding vendor lock-in.

Work or Play? Ludic Feedback Loops

Mon, 28 Jul 2025 05:57:58 +0000

In his substack post today, Venkatesh Rao wrote about reading and writing in the age of LLMs as playing and making toys respectively. In one part he writes about how the dopamine feedback loop from writing drove his switch from engineering to writing. For him, writing has ludic, play-like, qualities.

I have made almost all my “career” decisions as a function of play. I originally started off with a deep love of plants, how to grow them and their impact on the world. I was convinced I was going to have a lot of fun. I did have some. My wonderful undergrad professor literally hand held me through my first experiments growing tobacco plants from seeds. But that was about it. My next experiment was with woody plants and growing the seeds alone took 6 months, and by the end I had 4 measly leaves to experiment with. I quickly switched to cell biology.

Briefing: The State of Explainable AI (XAI) and its Impact on Human-AI Decision-Making

Thu, 24 Jul 2025 06:10:10 +0000

This post is a sloptraption, my silk thread in the CloisterWeb. The post was made with the help of NotebookLM. You can chat with the essay and the sources here: XAI NotebookLM Chat

I. Executive Summary #

The field of Explainable AI (XAI) aims to make AI systems more transparent and understandable, fostering trust and enabling informed human-AI collaboration, particularly in high-stakes decision-making. Despite significant research efforts, XAI faces fundamental challenges, including a lack of standardized definitions and evaluation frameworks, and a tendency to prioritize technical “faithfulness” over practical utility for end-users. A new paradigm emphasizes designing explanations as a “means to an end,” grounded in statistical decision theory, to improve concrete decision tasks. This shift necessitates a human-centered approach, integrating human factors engineering to address user cognitive abilities, potential pitfalls, and the complexities of human-AI interaction. Practical challenges persist in implementation, including compatibility, integration, performance, and, crucially, inconsistencies (disagreements) among XAI methods, which significantly undermine user trust and adoption.

AI: Explainable Enough

Wed, 23 Jul 2025 05:50:31 +0000

They look really juicy, she said. I was sitting in a small room with a faint chemical smell, doing one my first customer interviews. There is a sweet spot between going too deep and asserting a position. Good AI has to be just explainable enough to satisfy the user without overwhelming them with information. Luckily, I wasn’t new to the problem.

Coming from a microscopy and bio background with a strong inclination towards image analysis I had picked up deep learning as a way to be lazy in lab. Why bother figuring out features of interest when you can have a computer do it for you, was my angle. The issue was that in 2015 no biologist would accept any kind of deep learning analysis and definitely not if you couldn’t explain the details.

The secret flag of content

Sun, 20 Jul 2025 06:36:28 +0000

I don’t have any fun when I use LLMs to write. It may have perceived utility: popping out a LinkedIn article or two everyday. But I bet no one is actually reading. It’s a strip mall for a thumb stroll.

LLMs suck at writing. The summaries that LLMs give with the “Deep Research” are so poor in quality that I start to skim it. Yes, I skim the thing that is already a summary.

Chatbots, Bats & Broken Oracles

Sat, 05 Jul 2025 05:54:05 +0000

I had the strangest conversation with my son today. There used to be a time when computers never made a mistake. It was always the user that was in error. The computer did exactly what you asked it to do. If something went wrong it was you, the user, that didn’t know what you wanted. After decades of that being etched in today I found myself telling him that computers make mistakes, you have to check if the computer has done the right thing and that is actually ok. A computer that hallucinates also provides a surface for exploration and seeking answers to questions.

We Need Homes in the Delta Quadrant

Tue, 29 Apr 2025 07:44:02 +0000

Place is security, space is freedom. — Yi-Fu Tuan

Starfleet Log, Delta Quadrant—Classified Briefing

At the edge of the known, maps fail and instincts take over. We don’t just explore new worlds—we build places to survive them. Because in deep space, meaning isn’t found. It’s made.

I. Interruption of Infinity #

The Delta Quadrant is a distant region of the galaxy in the Star Trek universe—vast, largely uncharted, and filled with anomalies, dangers, and promise. It is where the map ends and the unknown begins. No stations, no alliances, no history—just possibility.

Domain Ontologies: Indispensable for Knowledge Graph Construction

Wed, 15 Jan 2025 08:06:00 +0000

AI slop is all around and increasingly extraction of useful information will face difficulties as we start to feed more noise into the already noisy world of knowledge. We are in an era of unprecedented data abundance, yet this deluge of information often lacks the structure necessary to derive meaningful insights. Knowledge graphs (KGs), with their ability to represent entities and their relationships as interconnected nodes and edges, have emerged as a powerful tool for managing and leveraging complex data. However, the efficacy of a KG is critically dependent on the underlying structure provided by domain ontologies. These ontologies, which are formal, machine-readable conceptualizations of a specific field of knowledge, are not merely useful, but essential for the creation of robust and insightful KGs. Let’s explore the role that domain ontologies play in scaffolding KG construction, drawing on various fields such as AI, healthcare, and cultural heritage, to illuminate their importance.

Jan. 8, 2025: Count your DIGITS! Drunk Bayesian

Wed, 08 Jan 2025 08:06:00 +0000

NVIDIA Project DIGITS #

Around 2015 I was putting together funds in academia. Convincing IT, senior professors, and finance that yes, it was worth giving me a LOT of cash to build a workstation with multiple GPUs.

…
“No, it isn’t for gaming.”
…
“Yes, it will change the world.”
…
“No, there are no university rules that hardware bought multiple invoices across multiple departments can’t be used in the same box.”
…
“Yes, I’m aware that all my individual quotes are just below the bureaucracy summoning purchase limits.”
…
“Yes I tried random forest with the other stats and ML methods, this really is better. How do I know? Well…”

Jan 4, 2025

Sun, 05 Jan 2025 00:30:49 +0000

Bayesian Thinking Talk (youtube) #

Talk details from Frank Harrell’s blog - includes slides

This beautiful talk about Bayesian Thinking by Frank Harrell should be essential material for scientists who are trained in frequentist methods. The talk covers the shortcomings of frequentist approaches, but more importantly the paths out of those quagmires are also shown.

Frank discusses his journey to Bayesian stats in this blog post from 2017 which is also in the next section.

Reimagining AI in Healthcare: Beyond Basic RAG with FHIR, Knowledge Graphs, and AI Agents

Fri, 03 May 2024 21:15:05 +0000

Introduction #

While exploring the application of AI agents in healthcare we see that standard Retrieval-Augmented Generation (RAG) and fine-tuning methods often fall short in the interconnected realms of healthcare and research. These traditional methods struggle to leverage the structured knowledge available, such as knowledge graphs. Data approaches like Fast Healthcare Interoperability Resources (FHIR) used alongside advanced knowledge graphs can significantly enhance AI agents, providing more effective and context-aware solutions.