Spicy sleight

Someone once shared a trick with me: if you’re not a great cook but want to sell food, your best bet is to make it spicy. Spiciness is often used as a trick to cover up culinary shortcomings. Let’s call this spicy sleight. Before we move on, it’s worth pointing out that this phenomenon is…

LLM doesn’t understand

I have a useful test to evaluate claims on LLM’s ability to “understand”. This example shows that at this current moment, even one of the smartest LLMs, GPT-4o, is truly “merely” a heap of statistical operations and is not as “intelligent” as humans. Here I asked [1] GPT-4o to merge the different definitions I have…

Picking a post title with the help of LLM

While writing that previous post, I needed a title for it. Asked chat GPT: I want to create a title for a blog post the post will talk about the value of engaging with a material, how cognitive effort is the key to effective learning I want to contrast it against the idea of merely…

Why rubber ducking works. The value of externalization.

https://www.youtube.com/watch?v=wI2g8V2uw3w&ab_channel=ElyseMyers Saw this video earlier. Immediately thought, “this is such a good example to use when I want to talk about externalisation”. Yes, externalisation is such a boring technical term. Conveniently, computer people calls it: “rubber ducking”. // leaving a meta-trace here to talk about an underrated magic that human has access to: creating a…

[Part 2] The YouTube script: youtube2llm.py

Hey, as promised, I’ll start introducing the different scripts in the Nootroscripts project. youtube2llm.py [TOC] Overview: analyse, embed, and ask anything about the content of a YouTube video. Feel free to skip to Example usages for a quick start and refer to the documentation-like sections later if you’re not an RTFM kind of person. Input…

Two ideas for using LLMs in software development

Two ideas for using LLMs in software development Other than coding assistants and code completions, Software developers spend substantial amount of time making technical decisions related to naming and organizing codes. We do this so 1) the programs can run (are programmatically correct) and 2) we can maintain it with the least amount of cognitive…

Nootroscripts: A suite of LLM-powered scripts for curious curators to help synthesize content in different formats and sources, more efficiently

Hey friends, today I am releasing my LLM-related Python scripts for working with youtube, podcasts, and articles. These CLI-based scripts have boosted my consumption and production process so much. Here’s the link to the github repository: Nootroscripts. Check out the comments in the code. It should help figure out the usage. The code is solid…