Lately, I discovered glances - a really nice-looking substitute for the top tool.
Just make it work.
Something that works has infinitely greater value than something beautifully crafted but not working.
early returns early returns early returns early returns early returns early returns early returns early returns early returns
SMART = science, math and art
Long COVID update: Chest pain is gone. I thought it was the main reason why I couldn’t make myself do things. And I feel normal, but still can’t make myself work. Strange, but at least there’s no pain.
developers, developers, developers, developers, developers, developers, developers, developers, developers, developers, developers 🙂
🌸 Mon, Oct 9, 2023 🌸
Congrads Unity community. CEO John Riccitiello is gone.
🌸 Wed, Oct 4, 2023 🌸
I like Bill Gates. Is it somehow related to the alert?
Is it ethical to lie to the bot, or to the LLM?
Lying to a bot, such as ChatGPT or any other LLM (language learning model), presents a different set of ethical considerations than lying to humans. Here are some points to consider:
No Emotional Harm: Bots don’t have feelings, emotions, or consciousness. Lying to them doesn’t cause them emotional distress, as it might with a human.
Purpose and Utility: The purpose of bots like ChatGPT is to provide information, answer questions, and assist users. If you provide false information, you’re potentially reducing the utility of the system for your specific session, but you’re not permanently “harming” the system.
Training Data: Although individual interactions with a user do not typically retrain the model, in systems where user interactions do influence the model’s future outputs, repeatedly providing false data could, in theory, degrade the model’s usefulness. This is more a consideration about the collective accuracy and integrity of a system, rather than an ethical concern about “harming” the model itself.
Your Own Ethics: Ethics is often personal. If you believe in being truthful and honest in all interactions, whether with humans or machines, then lying to a bot might feel wrong to you.
Intent and Outcomes: If you lie to a bot to maliciously manipulate someone else or cause some form of harm (e.g., using the bot to craft a lie for human recipients), then the ethical concern is more about the intent and the potential harm to humans than about the lie to the bot itself.
Privacy and Misleading Data: Some people might deliberately provide false information to bots as a privacy measure, ensuring their real data or intentions aren’t recorded or analyzed.
In general, the primary ethical concerns surrounding interactions with AI models center around the humans using them, the data they’re trained on, and the outcomes of their use, rather than the act of lying to the machine itself. If you’re considering the ethical implications of an action, it’s always good to consider the broader context and potential consequences.
I have a protocol, then I put my laptop in the backpack. I put the charging cable first, then the mouse, and lastly the laptop. The reason I am doing this is so I don’t forget the charging cable and mouse. And it happened. This weekend, I put the charging cable in my backpack and I forgot to put the laptop.
Pick random person for code review:
clang-repl $'#include <array>\n#include <iostream>\nsrand(time(nullptr));std::array names = { "Alice", "Bob", "Charly", }; std::cout << names[rand() % names.size()] << std::endl;'
Converting values to HEX from bash:
clang-repl $'#include <iostream>\nusing namespace std; cout << hex << (unsigned)-543 << endl;'
TIL: Do not instruct GPT to give short answers. Let it wander so it can generate better quality answers at the end of the day.
The Visual Studio editor is cursed. Why does pressing Enter not do what it should, and why does it indent the code and indent it incorrectly? Why does backspace not delete one space, but instead delete everything to the beginning of the line? Why is the editor blinking so much? How do you split the editor vertically, and why is splitting it horizontally a good idea? Can it be just a normal text editor, like Notepad for example?
🌸 Fri, Sep 8, 2023 🌸
Mind of a 3D graphics engineer: there is no spoon, everything is just a triangle.
🌸 Wed, Sep 6, 2023 🌸
What will be the point in 2023 to release a song with the piano tuned to something other than concert A?
🌸 Tue, Sep 5, 2023 🌸
I discovered an interesting bias in my brain. I have some issues with the ML model I am working on; some tuning issues, it is too repetitive and tends to finish monologues earlier than desired. So I was trying to solve it by myself and it never crossed my mind to ask ChatGPT for help. Like it is odd to ask an ML model to fix issues in another ML model.
🌸 Tue, Sep 5, 2023 🌸
I heard that in some prisons there is a way to torture prisoners by not giving them furniture to sit on; they have beds but not chairs. I recently got a glimpse of this sort of torture. The chair in the office was very uncomfortable. I bought a new one and it feels so good to sit on the comfortable chair.
🌸 Fri, Sep 1, 2023 🌸
“insurance is legalized gambling” - Robert A. Heinlein
There is no magic bullet in software engineering (khm Adjile). You just have to sit on your ass and code. If you spend time in meetings, you are not making progress. Progress is only made when you sit on your ass and code.
How to convert an ePub file into plaintext: ebook-convert input.epub output.txt
I have 13 billion parameters and 70 billion parameters Llama 2 sitting on my hard disk. And I realized, I have some chances to survive a zombie apocalypse. I have knowledge on my hard disk to bootstrap the civilization.
Should I try to ride the AI wave and create something around it? I have made multiple attempts in the past, but the technology was not ready at the time and it was quite expensive. Now, we have Llama which we can run offline.
We also have some successful examples like Seinfeld, so I can take away some ideas from it.
I am thinking of making a virtual streamer which will give some nonsensical live news, have a junk cooking show, and a ridiculous museum.
What do you think?
I think we should start calling people who have reached the age of 25 “quarter-centurians”. Sounds cool!
If you have a TikTok full of nerdy science videos and you want to watch pretty dancing girls, ask your friend mohjii to watch your TikTok for 15 minutes. Satisfaction guaranteed.
Previous - 1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 - 10 - 11 - 12 - 13 - 14 - 15 - 16 - 17 - 18 - 19 - 20 - 21 - 22 - 23 - 24 - 25 - 26 - 27 - 28 - 29 - 30 - 31 - 32 - 33 - 34 - 35 - 36 - 37 - 38 - 39 - 40 - 41 - 42 - 43 - 44 - 45 - 46 - 47 - 48 - 49 - 50 - 51 - 52 - Next