aavci

I don't see any answers here with a link to best practices or a grading system or something promoting clean websites. What can be done in this space?

You get things rating accessibility, methodologies for building apps but nothing about standards for website quality?

Where is the ‘sign in with google’ pop up?

Any better word than ‘meat-books’?

I scrolled through the comments as I am curious if anyone shared some ideas: What is the future of authoring by humans like? Instead of books being one of either ‘meat’ or ‘machine’, as this author is claiming, what would it look like for human authors to leverage AI for simply getting assistance for the stories they author? Could there be a healthy, non sloppy use for AI?

Can someone that uses these models provide some tips or docs on how best to get to use these at home to test around? Is the best thing a virtual environment?

Please help me understand. Is it more effective to have a general model like this instead of a state machine that leverages other models and a deterministic set of code?

Other than privacy and security and simplified development, what are other valid reasons you'd prefer to run this instead of a backend LLM?

I don’t see a problem here. Do you?

Also, Linear themselves are also working on this.

I liked looking at this.

Every cafe is different and cafes in each country vary heavily by city and area and I am not sure if I have recommendations on how to make crowd sourced data more reliable but I like the work done.

Why are some companies listed multiple times?

How do you populate this index?

I love the idea of such frontpages. There should be one for small scale b2c apps as well to help discovery

Thank you for clarifying. Also, great work! I am looking forward to trying it out for my personal learning note taking this week.

Does anybody mind explaining how the web of articles in the first image helps the writer?

Nice blog btw

FOMO is making me feel like I should mess around with openclaw but I can’t see any use cases that I can’t accomplish with other tools. What should I do based on this article?

How else can you be in the right place and right time to discover a problem to solve that can’t be seen from afar?

Thanks for sharing. This seems to me like a more useful approach than an agent that has full control over a computer

Interesting to see it took them so long to implement this. Claude was super limiting without the ability to have a scheduler or a connection to events

Thank you for sharing. This looks nice.

Do you have any further plans for this project?

"Nobody caught it in review" and "they're in someone's head" sound like issues to work on.

Some todos that come to my mind:

- Review more thoroughly.

- 'Vibe code' some unit tests

- Document and communicate the things that are in team members heads that should be openly shared

How does this compare the building your own bot that has access to these tools: - web plugin - api access to messaging - access to a job scheduler

How does this compare to fine tuning?

Interesting to see. I thought they were promoting fine tuning

Cheaper. Every month or so I visit the models used and check whether they can be replaced by the cheapest and smallest model possible for the same task. Some people do fine tuning to achieve this too.

I found myself asking Claude to teach me a topic recently instead of going to the perfectly written documentation. I didn’t like that. Went back to the documentation and used Claude to complement the knowledge.

I wonder if promoting open-source tooling and best practices could make it easier for new apps to adopt security features like E2E encryption. For example, someone building a chat app might not add E2E encryption unless they have access to user-friendly tools and are encouraged to do so.

Startups that initially choose the more private implementation version often face a disadvantage. They may not see immediate benefits and instead experience drawbacks, such as caring a bit more than their competitors. For example, an AI plugin using local large language models for privacy might not be rewarded as much as a competitor who fully embraces cloud-based solutions.

This looks really cool. Is the use case of getting an LLM to respond with custom TUIs something you have thoughts about?

This is pretty cool.

I am wondering why you chose to go with 4 files only instead of organizing the code into separate folders?

I’m not sure if they implied not having any dependencies anywhere