Hacker News Reader

agileAlligatoron Why were Victorian elites so effective?

The proximate cause of nationalization of EIC was the Sepoy Mutiny of 1857.

agileAlligatoron Why were Victorian elites so effective?

For instance the colonial office (the bit in London that administered the colonies centrally) had 14 staff in 1816, and grew to about 25 by 1850

Important context:

1. The Colonial Service never administered India, the biggest colony.

2. Lot of these colonies had representative institutions of their own. "We have 43 colonies, and not 50, 60, or 70, as has been variously stated; 27 of these have representative institutions, some in a higher degree of perfection than others, no doubt, but the whole have representative institutions, including Canada, New Brunswick, Jamaica, Antigua, Barbadoes, and all the minor West Indian colonies." [0]

[0]: https://hansard.parliament.uk/commons/1849-04-16/debates/807...

agileAlligatoron Dots: Always-on agents

Yeah this is why you have your top guy (Fable/astra) split everything into discrete tasks and go one layer at once. Get 10 subagents working on parallelized tasks that have no dependencies on each other through the night.

agileAlligatoron Dots: Always-on agents

I just YOLO it. Just don't give it access to anything that can produce permanent consequences. The cost of a mistake is maybe one hour of fixing it. If what it produced is 80% right and 20% wrong, by letting it run overnight, you've gotten 80% of the work done that otherwise wouldn't have happened.

agileAlligatoron Chat-based Large Language Models replicate the mechanisms of a psychic's con

https://chatgpt.com/share/6a5fdc7a-d6f8-83e8-bbea-8deb42cfed...

Terrence Tao's conversation with ChatGPT is very illuminating.

HN Discussion: https://news.ycombinator.com/item?id=49010345

agileAlligatoron A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

Yeah that is what I said, cost has gone down.

agileAlligatoron A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

Not really. Actually, for the purposes of cybersecurity, local models are far superior. Both offense and defense.

agileAlligatoron A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

That is the direct effect of reduced cost. Jevon's paradox type effect: cost goes down demand goes up. You can AI-check so many more things that would be very time consuming earlier.

agileAlligatoron A heap overflow and SSO misconfiguration to compromise OpenAI internal repos

The only thing AI has changed is that it has dropped both: the cost of attack and the cost of defense. Nothing in the game has materially changed; the game has just sped up.

agileAlligatoron TiVo to charge money for skipping commercials in your own recordings

Yep. IP is not true "property" either, since it is not "scarce"

agileAlligatoron Judge rules Trump administration’s blacklisting of Anthropic was illegal

What is "quality"? What is the objective test of it? I have made my definition clear: if people are choosing a product consistently, it must be of some value to them. What is your definition?

agileAlligatoron The turbulent AI era is here

https://daviddfriedman.substack.com/p/shrinking-hierarchy

This is almost a direct counter to your essay, albeit written before you

agileAlligatoron Judge rules Trump administration’s blacklisting of Anthropic was illegal

on what basis do you claim the local product is "good" if people are not choosing it?

agileAlligatoron Show HN: Mcploitable – The "Metasploitable" of the Model Context Protocol

mcploitable is a set of MCP servers that do not know they are part of a training range. Each is designed with intentionally vulnerable tools and/or overly permissive prompts. They are designed to replicate real-life agentic attacks and demonstrate the OWASP Agentic Top 10. There are also 3 guided simulations for those ASI categories that don't lend themselves well to being a vulnerable box.

agileAlligatoron Show HN: Mcploitable – Vulnerable MCP Servers for the OWASP Agentic Top

I did run the comment through AI, maybe that's what triggered the flagging.

agileAlligatoron 3D framework for the web, built on Svelte and Three.js

That's way cool man. I love the aesthetic.

agileAlligatoron Google made me ruin a perfectly good website (2023)

I honestly don't think it's possible to have a QA team large enough to handle the gajillions of websites that come up and disappear every day. They just have to come up with better and better metrics until they find one that approximates the human experience the best.

agileAlligatoron Anthropic: Prompt Library

> you'd simply disable WASD controls until the word is typed successfully.

Which it doesn't do, but would very simply be rectified by a follow up prompt asking it to do that

agileAlligatoron Show HN: MonsterWriter – Write a thesis, post, or organize notes

This project comes three months too late for me sadly... I'll still check it out though

agileAlligatoron I should have loved biology (2020)

Exactly how I learned to touch type; only it was Discord not IRC. The more things change, the more they remain the same.

agileAlligatoron Facebook bans ads for Votes for Women boardgame, claims "sensitive social issue"

He means that welfare beneficiaries and women vote in way that is against his image of what a capitalist democracy should be.

agileAlligatoron Facebook bans ads for Votes for Women boardgame, claims "sensitive social issue"

Yes, that is what he said. Welfare beneficiaries and women are two groups libertarians find are tough constituencies to appeal to, and both of these groups have increased in size since 1920.

agileAlligatoron Book list for streetfighting computer scientists (2022)

> eventually only read vol 2 because I was interested in (pseudo) random number at that time

Your comment comes at the perfect time, I am looking at learning about random number functions and tests of randomness myself. Do you have any other resources about these topics?

agileAlligatoron Conway's Game of Life is omniperiodic

I envy today's generation and their free access to high quality maths resources haha

agileAlligatoron Saying Goodbye to OpenSubtitles.org API

I completely forgot I had that active haha

agileAlligatoron Saying Goodbye to OpenSubtitles.org API

If I read the post correctly, they are making changes to the API, not the website itself, right?

You will still be able to visit the website and download the subtitles you want. You just won't be able to make use of the API without paying for it.

Edit: Nvm they are going to insert ads everywhere

agileAlligatoron Nature retracts controversial superconductivity paper by embattled physicist

I'm having a hard time believing those papers weren't written by humans after a cursory glance

agileAlligatoron Copying Angry Birds with nothing but AI

Steps 1 and 2 are more reliant on human-to-human communication if you are developing an application to be used by some end user and not something that just ingests data and processes it. It takes skill to translate non-technical speak into a clear goal and specifications required to achieve that goal, to probe the third party and ask specific questions to tease out their expectations of the application. That, LLMs can't do yet. I have no doubt that they'll get better at that in the near future though.

agileAlligatoron Copying Angry Birds with nothing but AI

The workflow has shifted a step up:

1. Define problem

2. Generate specifications <-

3. Write the code

Knowing how to do step 3 will get less and less important everyday.

agileAlligatoron Nota is a language for writing documents, like academic papers and blog posts

Reminiscent of Gwern's website[0] (click on any one of the articles to see what I am talking about). Was that an inspiration in creating this?

[0]: https://www.gwern.net/index