RSSAmplifier

Blog

@sebinsua

@sebinsua

sebinsua.comRSS feed ↗12 posts

Latest posts

Esc

LBA : Load-bearing assumption. The basis of your understanding rests on an assumption. What are you assuming? How can you verify these assumptions? Every conclusion downstream from an unchecked assumption inherits its uncertainty. TEA : Reading tea leaves. Don’t assume that the information you currently hold is a useful diagnostic. What is salient to you right now might not be relevant unless you…

The “old boy”

Professional advice is expensive, contained by prestige hierarchies and cordoned off from the public. But the upper echelon is the “old boy” . Men who wrote the book on a topic and then retired into boutique consultancies they run for Love and for Justice in their spare hours. If you’re polite, direct and don’t waste their time, they’ll often provide counsel expecting nothing in return.

Is dismissing the algorithmic tech interview a memetic hazard?

Algorithmic bathwater you might not like it, but mindlessly grinding leetcode actually does make you a good programmer — kache (@yacineMTB) August 3, 2023 The belief that algorithmic tech interviews do not test for real-world programming skills is a memetic hazard. There might be truth to it but it serves as displacement from what is often the real underlying reason for resistance: the fear…

Use notation to unmask relationships when problem solving

HackerRank has a simple programming problem called Number Line Jumps that states: You are choreographing a circus show with various animals. For one act, you are given two kangaroos on a number line ready to jump in the positive direction (i.e, toward positive infinity). The first kangaroo starts at location x 1 ​ and moves at a rate of v 1 ​ meters per jump. The second kangaroo starts at location…

Scaling dot products is all you need

This is the second installment to my series on machine learning, which builds on my initial post on neural networks “Bridging the gap between neural networks and functions” . In this post, we’ll explore how scaling up dot product operations using matrix multiplication enables modern machine learning. As with the rest of this series, we’ll keep the explanations intuitive yet detailed, aiming to…

Inverse number spiral

Recently, while working on a problem from the CSES Problem Set known as #1071: “Number Spiral” , I accidentally misread the problem description and instead of finding N for a given [ y , x ] , solved the inverse by finding [ y , x ] for a given N . This problem turned out to have an interesting solution based on simple mathematical reasoning and as I couldn’t find any other solutions for this…

Type-safe tensors

Recently, Ben Newhouse released a TypeScript-based implementation of GPT called potatogpt . Although the performance may be slow, it contains a very interesting approach to type-checking tensor arithmetic . This approach eliminates the need to run your code to verify whether operations are allowed or to keep track of the sizes of tensors in your head. The implementation is quite complex, employing…

Bridging the gap between neural networks and functions

As AI increasingly affects our lives, businesses may look to their software engineers for explanations regarding AI model outputs and guidance implementing and integrating these systems. Understanding the foundations of AI may become essential, even for software engineers that do not aspire to becoming AI researchers. In this series of articles, I shall attempt to produce a layperson software…

Making technology choices

Back in 2010, the adage “the right tool for the job” was prevalent within the developer community , but by 2015 it was seen by some as motivated reasoning and in general harmful advice. Nowadays it’s common to hear that we should “Choose Boring Technology” and “optimize our engineering decisions for iteration speed” . My view is that the adage “the right tool for the job” suffered from being…

Avoiding false cause

Tech teams often cargo-cult the practices and patterns that are established earlier on by their creators. This can lead to useless boilerplate proliferating and sometimes to the creators eventually deciding that they have no choice but to disavow their earlier approaches. “Many don’t realize I wrote all that stuff when React was relatively new and people shared what they learned, not instilled…

Interfaces as loops

I had an experience the other day that transformed my thoughts on interfaces and I’d like to share it. I believe that not enough of us understand what an interface really is and that this is hampering our ability to provide good user experience. One of the problems we have when answering the question “What is an interface?” is that we don’t have an accurate representation of what an interface…

Enhanced autodidactism for the chronically lazy and hyperactive

If you find it difficult to concentrate: you might think it is difficult to finish projects or learn new things. This requires protracted effort, not something you are able or willing to give. Maybe you’ll give up: why bother forcing yourself to continue when you have no natural inclinations. But don’t. Don’t try to push harder when you start to lose concentration. Do what you are most comfortable…