4
points
annjose
Joined 23 karma
annjose.at.hn
- annjose parentGenuine question to understand - have you tried this approach to build or break any habit for yourself? What were the learnings from it - what worked and what didn't? And how did you tweak the approach for the next habit?
- 1 point
- Congrats — well deserved! I love the game and play it every day; it's actually the first thing I do in the morning. A big fan of Hard mode! My best friend has also started playing it and we share the results with each other.
Just one feedback - on desktop browsers, I can see the list of answered clues below the textbox, but on the phone (Brave or Firefox on Android), I don't see that list. I am not sure if this is a feature or a bug, but it’s a feature I miss when playing on my phone. Seeing those answers gives that little “aha!” moment of satisfaction.
I also made a custom GPT - Bracket GPT [0] that helps in solving the clues when I am stuck. It doesn’t directly give the answers, but offers hints to help nudge you to the solution. It’s a fun companion when you're totally blanking.
[0] https://chatgpt.com/g/g-67e0f124cd408191943faadb3d70c6df-bra...
- This is the script that assembles the structured comments and generates the summary - https://github.com/levelup-apps/hn-enhancer/blob/main/script...
You can run it as: node summarize-comments.js <post_id> Example: node summarize-comments.js 43597782
And the summary will be put in the "output" folder.
You need to set the environment variable (in this case OPENROUTER_API_KEY because LLama4 is currently available at OpenRouter).
- This! I love the pure joy of picking both the destination and the path. No pressure, no goal — just the joy of building for its own sake.
These two lines really hit home:
> You don’t have to listen to any other voices here, except that quiet one inside of you that’s gently urging you to do the thing you know you need to do.
> You don’t need to know where it’s going to lead. For that matter, it doesn’t have to lead anywhere. Nothing ever has to come of it.
That freedom is everything. Just creating because it feels right (to me).
- 4 points
- 3 points
- I experimented with vibe coding [0] yesterday to build a Pomodoro timer app [1] and had a mixed experience.
The process - instead of typing code, I mostly just talked (voice commands) to an AI coding assistant - in this case, Claude Sonnet 3.7 with GitHub Copilot in Visual Studio Code and the macOS built-in Dictation app. After each change, I’d check if it was implemented correctly and if it looked good in the app. I’d review the code to see if there are any mistakes. If I want any changes, I will ask AI to fix it and again review the code. The code is open source and available in GitHub [2].
On one hand, it was amazing to see how quickly the ideas in my head were turning into real code. Yes reviewing the code take time, but it is far less than if I were to write all that code myself. On the other hand, it was eye-opening to realize that I need to be diligent about reviewing the code written by AI and ensuring that my code is secure, performant and architecturally stable. There were a few occasions when AI wouldn't realize there is a mistake (at one time, a compile error) and I had to tell it to fix it.
No doubt that AI assisted programming is changing how we build software. It gives you a pretty good starting point, it will take you almost 70-80% there. But a production grade application at scale requires a lot more work on architecture, system design, database, observability and end to end integration.
So I believe we developers need to adapt and understand these concepts deeply. We’ll need to be good at:
Without this knowledge, apps built purely through AI prompting will likely be sub-optimal, slow, and hard to maintain. This is an opportunity for us to sharpen the skills and a call to action to adapt to the new reality.- Reading code - Understanding, verifying and correcting the code written by AI - Systems thinking - understand the big picture and how different components interact with each other - Guiding the AI system - giving clear instructions about what you want it to do - Architecture and optimization - Ensuring the underlying structure is solid and performance is good - Understand the programming language - without this, we wouldn't know when AI makes a mistake - Designing good experiences - As coding gets easier, it becomes more important and easier to build user-friendly experiences[0] https://en.wikipedia.org/wiki/Vibe_coding
- > Spends hours crafting the perfect anti-doom-scrolling strategy only to immediately doom-scroll through HN comments about doom scrolling.
Spot on!
> Has an M2 Max with 64GB RAM but probably still complains when Chrome opens more than 5 tabs.
Not true, I have 40 tabs open!
> Created a tool to generate portfolios in 5 minutes but spent 5 hours explaining how to optimize YouTube settings. Priorities!
Ouch! Brutal and funny at the same time.
Thank you for making this!
- I can totally relate to this feeling of frustration with doom scrolling - I was in the same boat a few months ago, especially with YouTube. Fortunately, I managed to break free from the cycle by tweaking a few settings in YT itself (no special browser extensions needed). These are the changes that helped me:
Main settings that gave me a starting point:
1. Uninstalled YouTube app and now using only the browser version (on mobile and desktop)
2. Turned off Watch History in https://www.youtube.com/feed/history - "Pause watch history" (you can only pause the watch history and YT will periodically remind you to turn it back on. OH yeah, nice try Google!).
3. Turn off AutoPlay (toggle switch on the video player toolbar)
4. Tweak all the settings in https://www.youtube.com/account_playback - disable info cards and video previews (the setting that makes the videos to play when you hover over thumbnails)
After making these changes, your YT homepage and History page will be empty spaces - no videos at all. It is so refreshing! As a bonus, now YouTube shorts show only short-forms content from the channels you subscribed to. So it is more meaningful than some random junk.
Additional habits that helped me:
1. Subscriptions - I subscribed to specific channels that I want to follow - eg: Dave2D, MKBHD, fav cooking channels, NPR etc. and watch their videos via Subscriptions link
2. Topic-specific playlists - save interesting videos that I want to save for later - e.g: 'Health', 'Good recipes' etc.
3. Related videos - When a video is playing, YT shows a bunch of 'related videos' on the right. Most of these videos were not really related to the video, instead they are just trigger content. So I do two things here:
With these changes, I watch videos in one of three ways only - by searching for specific topics, or selecting from my playlists, or browsing through Subscriptions.- select the 'Do not recommend channel' from the vertical "..." menu in each video. - if the related video is genuinely interesting to me, choose the menu option 'Add to watch later'This was a big shift from a "push" to a "pull" model and has effectively stopped my doom scrolling habit in just a couple of weeks. I feel like I am watching YT on my own terms now.
- 2 points
- Cool. I am running this on M2 Max 64GB. Here is how it looks on my terminal [1]. Btw, the very first run after downloading the model is slightly slow, but the subsequent runs are ok.
[1] https://asciinema.org/a/fFbOEfeTxRShBGbqslwQMfJS4 Note: This recording is in real-time speed, not sped-up.
- Yes, it runs a quantized [1] version of the model locally. This version uses low-precision data types to represent reduced weights and activations (8-bit integer instead of 32-bit). The specific model published by Ollama uses 4-bit quantization [2] and that's why it is able to run on MacBook pro.
If you want to try it out, this blog post[3] shows how to do it step by step - pretty straightforward.
[1] https://huggingface.co/docs/optimum/concept_guides/quantizat...
- Ollama has released the quantized version.
https://ollama.ai/library/codellama:70b https://x.com/ollama/status/1752034686615048367?s=20
Just need to run `ollama run codellama:70b` - pretty fast on macbook.
- And the site is down. >>> Due to the large wave of new users joining over the past several days, we have encountered technical issues which have left many experiencing service interruptions. We thank you for your patience and encouragement as we work to make Vero available to everyone.