Hacker News: lilerjee

New comment by lilerjee in "We all depend on open source. We will defend it together"

lilerjee — Fri, 26 Jun 2026 17:30:07 +0000

Do you pay for the opensource projects which the "AI" companies used to train your LLMs?

Used them first to train LLMs to earn money, then exploit them again?

Current AI is like the film company producing TV series or movies

lilerjee — Tue, 23 Jun 2026 06:29:55 +0000

Article URL: https://devcyc.life/current-ai-is-like-the-film-company-producing-tv-series-or-movies/

Comments URL: https://news.ycombinator.com/item?id=48641141

Points: 2

# Comments: 0

New comment by lilerjee in "Has AI already killed self-help nonfiction books?"

lilerjee — Wed, 17 Jun 2026 10:48:14 +0000

If you work for only 4 hours a week, what practical things could you complete and do you really like the things what you are doing?

If you really like to do the things, you will spend many hours on it.

New comment by lilerjee in "Artificial intelligence is not conscious – Ted Chiang"

lilerjee — Thu, 04 Jun 2026 09:28:22 +0000

"Artificial intelligence is not conscious". This is a common sense.

LLMs are not even artificial intelligence, and they are just advanced automation systems.

AI is just marketing language.

Too many tools or tool overload

lilerjee — Sat, 18 Apr 2026 17:33:00 +0000

Article URL: https://devcyc.life/too-many-tools-or-tool-overload/

Comments URL: https://news.ycombinator.com/item?id=47817760

Points: 3

# Comments: 0

New comment by lilerjee in "The Zettelkasten method in Obsidian"

lilerjee — Sat, 11 Apr 2026 19:13:20 +0000

So I said "Most time". I know the situation you said. The situation exists but is not regular (I think it's occasional except you want to create a complex system by notes).

By the way, the most important thing is thinking, not the notes, but notes can help you thinking and creating better systems.

More and earlier thinking, more notes, better systems.

I hope it can help you and other people.

New comment by lilerjee in "The Zettelkasten method in Obsidian"

lilerjee — Sat, 11 Apr 2026 04:19:34 +0000

It is too complicated. We just get, save or write something, maybe with some categories, keywords, or tags.

After saving, maybe you need some organization later, but most time they are just there. Most time you search content by categories, keywords, or tags.

I think we need right tools for different requirements.

What Are Intelligence, AI, Automation System? Their Essences? How to Distinguish

lilerjee — Thu, 12 Mar 2026 17:40:51 +0000

What are Intelligence, AI, Automation System (AS), current AI? Do you really know the essential meanings of them? How to differentiate them?

Comments URL: https://news.ycombinator.com/item?id=47354494

Points: 2

# Comments: 0

New comment by lilerjee in "Are You becoming a Slave to AI? (Maybe you are losing chances to grow) (2025)"

lilerjee — Wed, 18 Feb 2026 18:44:06 +0000

The main views on AI are similar with mine

Are You becoming a Slave to AI? (Maybe you are losing chances to grow) (2025)

lilerjee — Wed, 18 Feb 2026 18:44:06 +0000

Article URL: https://medium.com/@eatulrajput/are-you-using-ai-as-a-crutch-or-becoming-a-slave-to-it-65edd3299dda

Comments URL: https://news.ycombinator.com/item?id=47064588

Points: 2

# Comments: 1

New comment by lilerjee in "Creators of Tailwind laid off 75% of their engineering team"

lilerjee — Fri, 09 Jan 2026 16:23:29 +0000

It's time to think about seriously "Could we create a license to make AI companies pay for your content?" or "Create a technology to ban AI bots effectively",

please refer to https://news.ycombinator.com/item?id=45729809

New comment by lilerjee in "Obsidian/Joplin does NOT support manual/arbitrary sorting of folders or notes"

lilerjee — Fri, 28 Nov 2025 15:52:46 +0000

Sorting or queuing is a common occurrence in life and work. For example, sorting the exam scores of students in the same grade at a school. Sorting is useful for lists with a few items or a lot of items.

Arbitrary sorting is particularly useful for prioritizing tasks.

Tags are mainly used for identification or grouping, and are quite different from sorting; the two are distinct.

Given the universality of sorting or queuing in real life, I believe it should be a fundamental feature of note-taking software.

Obsidian/Joplin does NOT support manual/arbitrary sorting of folders or notes

lilerjee — Fri, 28 Nov 2025 14:30:51 +0000

When I finally decided to try Obsidian, I discovered that it couldn't sort folders (except for the default alphabetical sorting). Further searching revealed that most note-taking software (including Joplin) didn't allow manual or arbitrary sorting of folders or notes.

I believe manual or arbitrary sorting is a basic function that any bookmark manager or note-taking software should provide.

Why can't most software do this? Is it that difficult?

Comments URL: https://news.ycombinator.com/item?id=46078970

Points: 4

# Comments: 3

New comment by lilerjee in "AI's Dial-Up Era"

lilerjee — Tue, 04 Nov 2025 11:28:17 +0000

What are the disadvantages of AI?

The author didn't mention them.

AI companies robbed so much data from the Internet free and without permission.

Sacrificing the interests of owners of websites.

It's not sustainable.

It's impossible for AI to go far.

New comment by lilerjee in "Free software scares normal people"

lilerjee — Fri, 31 Oct 2025 08:20:29 +0000

One of my product design principles:

Concise

Focus only on the parts that are really needed, necessary, or the most important; hide the currently unnecessary or secondary, or simply remove the truly unnecessary and secondary; remove all unnecessary

Maybe it can help you

Ask HN: Could we create a license to make AI companies pay for your content?

lilerjee — Tue, 28 Oct 2025 06:48:19 +0000

The majority of traffic for many websites comes from AI or data companies. These companies use your data for free without your permission, process it, and then sell it to you or others. This is robbery and theft.

Technical measures cannot fundamentally solve the problem; they can only mitigate it, and consume significant resources, and waste time and energy.

Legal, commercial, and ethical means can truly resolve the problem, just as open source software licenses address theft and piracy.

I think the better solution is as follows:

Create a license that prohibits or makes AI companies pay to use your content.

1. Include your license link in your `robots.txt` file, or create another file called `license.txt` in the server's root directory.

2. Include your data use policy in this file. You can prohibit data use or specify the conditions for use, such as payment, price, and duration.

3. If you require AI companies to pay for data use, publish your payment link on a page of your website.

4. If an AI company fails to pay or violates your `license.txt`, you can take legal action against them.

5. If you don't want to, don't have the time, or if litigation is ineffective, you can publicly denounce the company and damage its reputation. Don't underestimate the power of moral condemnation and business reputational damage.

6. It's easy to check if an AI company is using your content. If you confirm that, just take actions against them.

What do you think of this solution, or do you have a better one?

Comments URL: https://news.ycombinator.com/item?id=45729809

Points: 1

# Comments: 3

New comment by lilerjee in "AI Startup Founders Tout a Winning Formula–No Booze, No Sleep, No Fun"

lilerjee — Fri, 12 Sep 2025 14:33:14 +0000

Talking nonsense seriously.

"Winning Formula", attract your attention

No Sleep, No Fun -> No Efficiency, No Creativity -> No good product, No good company -> No health, No life

New comment by lilerjee in "Reshaped is now open source"

lilerjee — Thu, 11 Sep 2025 12:52:45 +0000

Too much JavaScript, I think using CSS to implement the same UI is better and simpler.

New comment by lilerjee in "PKM apps need to get better at resurfacing information"

lilerjee — Wed, 10 Sep 2025 12:42:04 +0000

Good article. I am solving the similar problems in my product[1], which purpose is more practical and uses hierarchy (like file system) and keywords to organize and link information.

If you want to resurface them, you can show them in hierarchy(show children, or descendants in any level), search or filter by keywords, sort them by created time, URL, order, or other fields.

Resurfacing from current context can be very complicated, you should create some fields to track your activities, create many routines to handle these information automatically, and create relative UI. These are advanced features based on the current basic features.

[1] https://metaessen.com/

New comment by lilerjee in "Cognitive load is what matters"

lilerjee — Sat, 06 Sep 2025 16:15:51 +0000

Solving problems is the purpose and focus.

How to solve problems?

The simple, direct, and effective method is good one method sometimes, not all time. For example, use very complicated method to obfuscate code in order to increase the difficulty of cracking, and there are many environments to use complicated code or methods.

So one is the purpose, one is the method.

Do not think "Lower Cognitive Load" is the purpose.

"Lower Cognitive load" is just the by-product sometimes, not the purpose.