So, using my 9PM-12AM for so many days, I built HackerComments.ai – a reimagined way to explore HackerNews discussions. Instead of scrolling endlessly through nested threads, HackerComments extracts, summarizes, and organizes comments into these sections:
TL;DR – A quick summary of key takeaways
Comments Mind Map – A 3D force graph to visualize discussions
Comments Highlights – Top topics that people are debating
User Perspectives – Insights from key contributors
This was my first solo full-stack project, and I couldn’t have done it without LLMs and GenAI. I can build models and algorithms, but software and infra weren’t my expertise. AI made it all more accessible, from coding to simplifying complex discussions.
HackerComments.ai reimagines how you explore HackerNews discussions by extracting and organizing comments into digestible sections like TL;DR summaries, a 3D comments mind map, debate highlights, and key user insights. Say goodbye to endless scrolling through nested threads and get a holistic view of the conversation in just a few clicks.
For battery, I think it can be connected directly to power instead of always rely on battery. I may need pure-battery when moving around (hopefully nothing more than 2hr) but have it plugged in when I am stable.
Maybe having LLM providing insights on more math proving techniques (which is the soul of optimization) is more meaningful than having it directly solve the optimization. The ROI of having a effective proof vs. brutal forcely search for a solution is simply day and night.
- The general idea of providing guiding prompt + scoring for better objective values is interesting. Though I doubt how this scale since it requires a lot of guiding/customization towards different/bigger problem, but I’d love to think further on it.
- Maybe let LLM help explain it’s thinking process/logic to help improve existing algorithms (rather than using it as a standalone-optimizer). I once did that for an allocation problem - and it was able to show a basic algorithm for a feasible solution.
- A essential topic in optimization is about proving optimality, maybe having AI providing insights on proving could also be cool.
- Author compared their algorithm with heuristics on randomly generated TSP problems (why not TSPLIB), the claim is that LLM can do better than heuristics on small problems. They showed an interesting metric on # of success suggesting we might need to sample multiple LLM runs for a good results.
- One big question I did not find an answer is how they replicate the runs given the stochastic nature of LLMs. Even with a zero temperature, LLM is only relatively less random. This extends to many LLM-application papers and hence must be papers talking about it.
The reality is no matter how genuinly your delivery driver ask "hi how can I help", he/she does not care about it. There is simply no time for him to care about it. Otherwise, you will see your delivery fee increase since.
I think taking an average 1.5-2 years here is just bad statistics to under-estimate the real-world scenario. It is a heavly skewed where India and China have way extended wait time vs other countries (practically no wait time). The greencard process is provision by country and hence average = bad statistics
I know this is about how to convert text-to-music. But, is it just me to find most of the sample "music" to be melody-less. The beats and flow make it sounds like "music" but it is truly not. I'd call it text-to-noise for now lol
EA is supposed to be pretty happy with its SimCity BuildIt as it is a new way of generating lots of revenue through micro-transactions and makes the balance sheets look better. Meanwhile, players are pretty happy with Paradox's real passion about how to build a better city simulation game. For what it is worth, it is a win-win-win for EA, players, and Paradox Interactive.
It is what you believe it to be (most of the time politically). People use it prove the theory. Other people use it to disprove. I can only see the fun with technicality on using the real data to simulate this.
This was my first solo full-stack project, and I couldn’t have done it without LLMs and GenAI. I can build models and algorithms, but software and infra weren’t my expertise. AI made it all more accessible, from coding to simplifying complex discussions.