Surely any sane person vibe coding a note taking app just has it save all the notes as markdown files to disk? At that point making a backup is trivial and they're unlikely to get corrupted.
I vibe coded a tiny MUD-style world sim where LLMs control each character. It's basically a little toy sandbox where LLMs can play around. There's no real goal to this, I just thought that it would be fun, like a more advanced tamagochi.
One of the issues I encountered initially was that the LLMs were repeating a small set of actions and never trying some of the more experimental actions. With a bit of prompt tweaking I was able to get them to branch out a bit, but it still feels like there's a lot of room for improvement on that front. I still haven't figured out how to instill a creative spark for exploration through my prompting skills.
It has been quite exciting to see how quickly a few simple rules can lead to emergent storytelling. One of the actions I added was the ability for the agents to pray to the creator of their world (i.e. me) along with the ability for me to respond in a separate cycle. The first prayer I received was from an agent that decided to wade into a river and kneel, just to offer a moment in stillness. Imagining it is still making me smile.
Unfortunately, I don't have access to enough compute to run a bigger experiment, but I think it would be really interesting to create lots of seed worlds / codebases which exist in a loop. With the twist being that after each cycle the agents can all suggest changes to their world. This would've previously been quite difficult, but I think it could be viable with current agentic programming capabilities. I wonder what a world with different LLM distributions would look like after a few iterations. What kind of worlds would Gemini, Claude, Grok, or ChatGPT create? And what if they're all put in the same world, which ones become the dominant force?
I’ve been messing around with a similar project (but in a grimdark/cosmic horror setting). I was running into the same issue, agents getting stuck in a loop. What worked for me was adding dwarf fortress/rimworld like systems. The random events and systems influencing systems worked wonders for me.
Hypothesis: limiting usage / tokens could have a positive effect on project quality, since it forces the developer to think more carefully about the problems they're working on. When you're forced to stop and slow down, you try to be more deliberate with token usage. But if you have unlimited tokens you can just keep generating infinite lines of code without thinking as hard about the problem.
I've seen people on social media bragging about how they're able to produce a mountain of code as if this was praiseworthy.
Since we're on the subject of waste removal, most mammals also pee for roughly the same amount of time [0], around 20 seconds. If you look up a video of an elephant urinating, it's quite the spectacle and the flow is voluminous.
Amazing. Exactly the insightful comments that I hope to see on HN...perhaps not on the current subject but nonetheless
> On-Topic: Anything that good hackers would find interesting. That includes more than hacking and startups. If you had to reduce it to a sentence, the answer might be: anything that gratifies one's intellectual curiosity.
That's really cute, it reminds me that Will Wright (creator of The Sims) has referenced this book "The Ants", by Bert Holldobler in multiple occasions as a key inspiration for his games (and in particular SimAnt) and systems thinking. Did you come across that during your research phase or had you not heard about it? I haven't read it yet, but maybe someday I'll get around to it.
Hey, I'm the one who built this particular challenge!
I had no clue, but thanks for the book lead! It didn't come up directly, but SimCity 2000 and especially SimCity 4 had a huge impact on me growing up / I still spin up SimCity 4 from time to time, so I imagine there's a massive indirect influence haha.
This is actually all built in nextjs/react, though the initial sketches of the sim had no visualization and were just running in my terminal using bunjs
Will Wright gave at least one great lecture at SIGGRAPH. Wonderful thinker and communicator. #ACM plz unlock this culture.
Ants have fascinated me from a young age, I’d construct a viewing platform so I could watch the hive at work.
My favorite were the carpenter ants, so smart. It feels like they have a theory of mind. I didn’t hurt them, but if you disturbe their nest under a board, the efficiency and expediency they would exhibit while collecting their young was fascinating.
*edit, I meant to ask if you had any other book recommendations?
Well, I'm currently reading through "Designing Virtual Worlds" by Richard Bartle. He's known for being one of the creators of MUD (multi-user dungeon). I'm not far along enough to make a judgement on the quality of the contents, but I keep seeing the book title pop up everywhere so it seems important.
Speaking of SimAnt, I soon discovered the bug that by moving into the top corner, the black ants would conquer all squares with no intervention from that one sideways and downwards and too easily win the game.
Still, it was fun just messing around with the ants, watching the trails, and chasing spiders by calling forth all ants.
I think the desire is that in the long-term AI should be able to use any human-made application to accomplish equivalent tasks. This email demo is proof that this capability is a high priority.
This guy's videos were immediately going viral after the conflict began. I enjoyed and found them educational, but I'm taking all of his claims with a grain of salt because I also don't know much about the region or its history. He talks very authoritatively which makes for compelling storytelling but conflicts of this magnitude require much more context to really understand.
For context, there's been a massive project to produce as close to perfect of a color-corrected version of Dragon Ball and Dragon Ball Z. It can be found online in traditional anime torrenting sites. It's really an outstanding labor of passion and a true testament to the global community's love for this series.
I'd imagine that there's some discussion about how to make the most out of the tool as well as discussion of experiments and capabilities. I'm not even sure what exactly "Microsoft Copilot" entails anymore because of the multiple rebrands, but having a place where you can discuss exploring plugins and other adjacent features seems useful.
Not quite the same, but recently I was recently looking around for communities centered around Claude Code for discussion about people's workflows as well as discussion about what plugins people are using and if they notice it making a significant difference.
Since the technology is still evolving, having an active community can help you discover new patterns and explore the space more effectively.
> [...] I'm not even sure what exactly "Microsoft Copilot" entails anymore [...]
Watching from the sidelines (not a Microsoft user), I've completely lost track. Between this, the Azure 365 cloud whatever stuff, I have no idea what many of the products even exactly are any more.
Simply put Microsoft is the worst company at naming stuff. Even when they come up with a good name for something, they'll name 3 other totally different products the same thing to maximize confusion.
I gotta say though, I'm actually not sure which VMware (well Broadcom I suppose) products I use anymore. I'm pretty sure they took the Aria name off something else they called Aria for a little while. So Aria is no longer Aria but they still have Aria but it's what used to be called XYZ
Xbox Series with X > S (so if you want the high end of the current generation you want the Xbox Series X; if you want mid-range things are more complicated because you can now get an Xbox One X, but not the Xbox One, used for much less than you'd get an Xbox Series S for and which one is "better" is a dice roll depending on the games you want to play and if 4K matters to you…)
Series is a real weird word to use there. But it also doesn't help that the versions are extra complicated because with "PC-like compatibility" in everything after the Xbox One playing just about the entire same library you need a bit of a matrix to figure out which is best for you if you don't care about the "latest and greatest".
Oh wow yes, completely forgot about that one. To me, it's a complete blur made from single words and letters, one series x s one box 360? Maybe they should create a 365, with MS office pre-installed. Or something.
Seriously? Does anybody know what Copilot is? I don't think I have ever seem a "Copilot user", so I don't know what it looks like. Is it the little macro key on new laptop keyboards? The chatbot you get in Bing? A technical philosophy? Or is it in essence just copilot.com, the mediocre chat interface which you used to get free GPT-4 three years ago?
I wish. I got a Dell laptop for work and they've replaced the right Ctrl key with a Copilot key, and (because it's a locked-down work sysyem) the only thing I can remap that to is the Windows menu. And I keep hitting it out of muscle memory, interrupting everything. But at least now it doesn't launch Copilot.
Which I could add is "the only AI approved for use by IT" because they hate us.
> Which I could add is "the only AI approved for use by IT" because they hate us.
It's the same at our place. It's basically the lowest effort way as we already have data agreements with Microsoft 365 it eliminates a lot of the paperwork. And they do promise that they won't train on data even in the free (well, included with basic M365) version for corporate users. A lot of others don't unless you pay.
It's too bad because it seems to be the worst AI around. Even compared to ChatGPT itself which uses the same model as copilot in MS Office. I don't really understand why there's such a difference. If you do pay the $30 it's a bit better especially the researcher.
Double check if the (hopefully not locked) BIOS gives an option to customize the CTRL key. I had a previous work laptop which also got cute with the CTRL button, but thankfully did let you remap it.
Colin and Samir have some really in-depth videos with MrBeast which help document and explain his rise in fame, influence, and prominence. He's one of the most dedicated optimizers of our time, and he's been responsible for shaping the YouTube meta for years at this point. At this point a significant portion of his brain is probably fully allocated towards optimizing for the most engaging YouTube videos possible. The "24 hours with MrBeast" video helped contextualize his fame to me, it's really rock-star level. It's a shame that not many people engage with him on a more technical level, I think he would have a lot of interesting insights over which to nerd out about.
I agree, even more generally, in that if you think about any medium from a creator perspective there is so much more depth to it.
Pop music is not my thing, but I read a book about it, and the max martins and production function of these giant hits, and its endlessly fascinating optimization problems.
reply