Articles for category: AI Tools

March 15, 2025

ikayaniaamirshahzad@gmail.com

A Copilot For The Mind (2023)

Large language models are perhaps the ultimate study buddy, that one thing that might help you actually absorb information while you’re reading. That’s Dan Shipper’s vision, as outlined in this essay from March 2023. Dan’s essay from last year stands out as a prescient reflection on what’s come since. And with OpenAI’s DevDay set for October 1 and Every taking a quarterly Think Week, we thought it was ripe to republish as part of our week of essays on the power of ChatGPT. ICYMI: We created eight custom wallpapers based on Every’s art for iPhone or Android. Download them for free.—Kate

March 15, 2025

ikayaniaamirshahzad@gmail.com

Merklemap runs a 16TB PostgreSQL

Merklemap runs a 16TB PostgreSQL. Interesting thread on Hacker News where Pierre Barre describes the database architecture behind Merklemap, a certificate transparency search engine. I run a 100 billion+ rows Postgres database [0], that is around 16TB, it’s pretty painless! There are a few tricks that make it run well (PostgreSQL compiled with a non-standard block size, ZFS, careful VACUUM planning). But nothing too out of the ordinary. ATM, I insert about 150,000 rows a second, run 40,000 transactions a second, and read 4 million rows a second. […] It’s self-hosted on bare metal, with standby replication, normal settings, nothing

March 15, 2025

ikayaniaamirshahzad@gmail.com

Perform fully customizable rolling aggregations

Deephaven’s version 0.33 release introduced several new features and improvements. Two methods that have the potential to save users a lot of time are the new Rolling Formula update_by operations: rolling_formula_tick and rolling_formula_time. These operations support user-defined formulas, which can be used to calculate rolling aggregations over a specified window of rows or time. Support for user-defined formulas opens up a whole new world of versatility for update_by. This is exciting for a few reasons. First of all, support for user-defined formulas opens up a whole new world of versatility for update_by. Rather than being limited to a single update_by

March 15, 2025

ikayaniaamirshahzad@gmail.com

Data Machina #249 – Data Machina

Generative AI Music. In the last year or so, Generative AI Music has improved massively. Although early days, today you can generate some pretty decent, short duration music of all kinds with AI. If you like creating music and AI, here is a list of interesting Generative AI music stuff. Facebook AIR MusicGen. Probably one of the pioneering models in AI quality music generation. MusicGen has sparked a whole universe of MusicGen derivative models of all kinds, and it’s the model behind many musicgen apps. The model is based on a single stage auto-regressive Transformer model, and unlike Google LM,

March 15, 2025

ikayaniaamirshahzad@gmail.com

Prompting Fundamentals and How to Apply them Effectively

Writing good prompts is the most straightforward way to get value out of large language models (LLMs). However, it’s important to understand the fundamentals even as we apply advanced techniques and prompt optimization tools. For example, there’s more to Chain-of-Thought (CoT) than simply adding “think step by step”. Here, we’ll discuss some prompting fundamentals to help you get the most out of LLMs. Aside: We should know by now that, before doing any major prompt engineering, we need reliable evals. Without evals, how would we measure improvements or regressions? Here’s my usual workflow: (i) manually label ~100 eval examples, (ii)

March 15, 2025

ikayaniaamirshahzad@gmail.com

What Developers Often Get Wrong When Interacting with AI

Hey there, fellow developers! I see you. You’re grinding through code, debugging, and sometimes, in a moment of desperation, turning to AI for answers. But let’s be real—sometimes, our interactions could go a lot smoother. In this blog, I’ll highlight the common mistakes developers make when interacting with AI and how to get better, more useful responses. If you’ve ever felt frustrated with AI, this might be why. 1. Vague Questions = Vague Answers Imagine a client sending you a message saying, “Hey, my website is broken. Fix it.” No details. No error messages. Just vibes. That’s exactly what happens

March 15, 2025

ikayaniaamirshahzad@gmail.com

Computational Humanities Research 2024

Predicting spatial representations in literature is a challenging task that requires advanced machine learning methods and manual annotations. In this paper, we present a study that leverages manual annotations and a BERT language model to automatically detect and recognise non-named spatial entities in a historical corpus of Swiss novels. The annotated data, consisting of Swiss narrative texts in German from the period of 1840 to 1950, was used to train the machine learning model and fine-tune a deep learning model specifically for literary German. The annotation process, facilitated by the use of Prodigy, enabled iterative improvement of the model’s predictions

March 15, 2025

ikayaniaamirshahzad@gmail.com

Fine-tune FLUX.1 with an API

FLUX.1 is all the rage these days, and for good reason. It’s a fast, powerful image generation model that’s easy to use and fine-tune, and it generates stunning images. Last week we brought you a guide to fine-tuning Flux with faces. That guide used an entirely web-based flow to create a fine-tuned Flux model, without writing a single line of code. We heard from some users that they would like to fine-tune Flux with an API, so we’re back this week with another tutorial that shows you how to do just that. In this guide, you’ll create and run your

March 15, 2025

ikayaniaamirshahzad@gmail.com

The Mechanics of Monocular Depth Estimation in Estimating Depth from 2D Images

Introduction Depth Perception is a process of understanding 3D objects and judging how far these objects are. It helps in tasks like navigation, object manipulation, and scene understanding. Despite its importance, estimating depth from a single image, known as monocular depth estimation, is a challenging problem for Artificial Intelligence (AI) to solve. However, advancements in machine learning, particularly deep learning or AI, have significantly improved the accuracy and reliability of monocular depth estimation. Convolutional Neural Networks (CNNs) and other deep learning architectures have shown great promise in learning to predict depth from 2D images by leveraging large datasets and powerful