Hacker Newsnew | past | comments | ask | show | jobs | submit | neomantra's commentslogin

I've been evangelizing vibe coding, because we are wielding something much more powerful now than even ~3 months prior (Nov was the turning point).

Now that Prometheus (the myth, not the o11y tool) has dropped these LLMs on us, I've been using this thought experiment to consider the multi-layered implications:

In a world where everyone can cook, why would anybody buy prepared food?


That is hilarious.... and to prove the point of this whole comment thread, I created reddit-kv for us. It seems to work against a mock, I did not test it against Reddit itself as I think it violates ToS. My prompts are in the repo.

https://github.com/ConAcademy/reddit-kv/blob/main/README.md


Typo-Driven Development!

Thanks for sharing this — I appreciate your motivation in the README.

One suggestion, which I have been trying to do myself, is to include a PROMPTS.md file. Since your purpose is sharing and educating, it helps others see what approaches an experienced developer is using, even if you are just figuring it out.

One can use a Claude hook to maintain this deterministically. I instruct in AGENTS.md that they can read but not write it. It’s also been helpful for jumping between LLMs, to give them some background on what you’ve been doing.


In this case, instead of a prompt I wrote a specification, but later I had to steer the models for hours. So basically the prompt is the sum of all such interactions: incredibly hard to reconstruct to something meaningful.

This steering is the main "source code" of the program that you wrote, isn't it? Why throw it away. It's like deleting the .c once you have obtained the .exe

It's more noise than signal because it's disorganized, and hard to glean value from it (speaking from experience).

I wasn’t exactly suggesting this. The source code (including SVG or DOCX or HTMl+JS for document work) is the primary ground truth which the LLM modifies. Humans might modify it too. This ground truth is then rendered (compiled, visualized) to the end product.

The PROMPTS.md is communication metadata. Indeed, if you fed the same series of prompts freshly, the resultant ground truths might not make sense because of the stochastic nature of LLMs.

Maybe “ground truth” isn’t exactly the right word, but it is the consistent, determined basis which formed from past work and will evolve with future work.


> because of the stochastic nature of LLMs.

But is this "stochastic nature" inherent to the LLM? Can't you make the outputs deterministic by specifying a version of the weights and a seed for the random number generator?

Your vibe coding log (i.e. your source code) may start like this:

    fix weights as of 18-1-2026
    set rng seed to 42

    write a program that prints hello world
Notice that the first two lines may be added automatically by the system and you don't need to write or even see them.

> But is this "stochastic nature" inherent to the LLM?

At any kind of reasonable scale, yes. CUDA accelerators, like most distributed systems, are nondeterministic, even at zero temperature (which you don't want) with fixed seed.


I see what you are saying, and perhaps we are zeroing in on the importance of ground truths (even if it is not code but rather PLANs or other docs).

For what you're saying to work, then the LLM must adhere consistently to that initial prompt. Different LLMs and the same LLM on different runs might have different adherence and how does it evolve from there? Meaning at playback of prompt #33, will the ground truth gonna be the same and the next result the same as in the first attempt?

If this is local LLM and we control all the context, then we can control that LLM's seeds and thus get consistent output. So I think your idea would work well there.

I've not started keeping thinking traces, as I'm mostly interested in how humans are using this tech. But, that could get involved in this as well, helping other LLMs understand what happened with a project up to a state.


I've only just started using it but the ralph wiggum / ralph loop plugin seems like it could be useful here.

If the spec and/or tests are sufficiently detailed maybe you can step back and let it churn until it satisfies the spec.


Isn't the "steering" in the form of prompts? You note "Even if the code was generated using AI, my help in steering towards the right design, implementation choices, and correctness has been vital during the development." You are a master of this, let others see how you cook, not just taste the sauce!

I only say this as it seems one of your motivations is education. I'm also noting it for others to consider. Much appreciation either way, thanks for sharing what you did.


Doesn’t Claude Code allow to just dump entire conversations, with everything that happened in them?

All sessions are located in the `~/.claude/projects/foldername` subdirectory.

Doesn't it lose prompts prior to the latest compaction?

I’ve sent Claude back to look at the transcript file from before compaction. It was pretty bad at it but did eventually recover the prompt and solution from the jsonl file.

It’s loses them in the current context (say 200k tokens), not in its SQLite history db (limited by your local storage).

I did not know it was SQLite, thx for noting. That gives the idea to make an MCP server or Skill or classical script which can slurp those and make a PROMPTS.md or answer other questions via SQL. Will try that this week.

It doesn't lose the prompt but slowly drains out of context. Use the PreCompact hook to write a summary.

aider keeps a log of this, which is incredibly useful.

Here's the link to the Brooking's report from the NPR article, to read it in full: https://www.brookings.edu/articles/a-new-direction-for-stude...

I've only skimmed it, but I note that all this research is before Nov 2025 and is quite broad. It does get some into coding, mentioning GitHub CoPilot and also refers to a paper about vibe-coding, where the conclusion is that not understanding the artifacts is a problem.

So all this reporting is before Gemini 3 and Opus 4.5 came out. Everything is really different with the advent of that.

While substitute teaching just before Xmas 2025, I installed Antigravity on the student account of the class computer and vibe-coded two apps on the smart board while the kids worked on Google Classroom. This was impromptu, to liven up things, but I knew it would work because I had such amazing experiences with the tool the week before.

* [1] Quadratic Formula Explorer for Algebra 2

* [2] Proving Parallelograms for Honors Geometry

Before the class ended, I then gave a quick talk the gist was: "I just made these tools to understand the coursework by conversing with an LLM. Are you going to use this to cheat on your homework or to enhance your understanding?"

I showed it to a teacher and then she pointed me to existent tools like them on educational web sites. But that was missing the point that we can just manifest the very hyper-specific tools we need... for example how should the Quadratic Formula Explorer work for someone with dyslexia?

I'm not sure what the next steps with all this is, but certainly education needs to adapt. The paper notes "AI can enrich learning when well-designed and anchored in sound pedagogy" and what I did there is neither, so imagine how sweet it is gonna be when we weave this into educational systems by skilled curriculum designers.

[1] https://conacademy.github.io/quadratic_explorer/ [2] https://conacademy.github.io/proving_parallelograms/


I have a two-fold approach to this:

* With specific positive or negative feedback, I will issue friendly complements and critiques the LLM to reinforce things I like and reduce things I don't.

* Rather than thinking sycophantic/antagonistic, I am more clear about its role. e.g "You are the Not Invented Here technologist the CEO and CTO of FirmX will bring to our meeting tomorrow. Review my presentation and create a list of shortfalls or synergies and as well as possible questions".

So don't say "please suck at your job", give them a different job.


Very cool, thank you for creating and sharing. Literally yesterday YouTube searched (for the first time in a long time) for "D blues backing for Harmonica". I am interested now in exploring LLM paths for that.

You did a great job on the visualizations!!

I couldn't get sound to work on OSX yet, but will keep trying. We uke here, so will make an issue for that ;)

I work on NTCharts and also just yesterday had an LLM fix a visual bug using world understanding, one that I had thought about multiple times prior. In the end, the solution was obvious once revealed and I had overthought the problem.

Video in OP Miditui project is amazing, I missed it on first skim.


Thanks, everything was vibe coded, so visualizations thanks to vibe coding. I'm prompting like key user/product manager/architect, so expressing user needs, and ensuring the overall architecture seems reasonable.

For uke, would you expect same functionality but with 4 strings?


November 2025 was when AntiGravity and Gemini 3 came out and everything changed for me. Six months earlier, I had tried to vibe-code the 21+ verification page to AgentDank (an OSS cannabis MCP server connecting LLMs to open data via DuckDB SQL). After hours, I couldn't get a page fully working.

I tried it with AG+G3, I prompted both the age 21+ screen AND the chat interface. It one-shotted a working version of both in less than a minute.

I was immediately free to start exploring my idea! Adding multiple personality Budtenders, a Stash box for frequent item; it would create mocks and tests. So liberating.

Then I had this other idea in my head for a while, that since DuckDB is broadly portable and can target WASM, I could play with the datasets in the browser and much of the app doesn't need an MCP-connected LLM or any backend services.

  next up, there is another mode where we will browse and visualize the cannabinoid contents.  the dataset will be the data here https://github.com/AgentDank/dank-data    we will use apache echarts for visualizations.  we can probably embed duckdb in the browser and do it all the queries there.   we can have some simple UI for exploring, as well as raw SQL query

And it one-shotted an entire DBA application interface with a custom UI and visualization to explore the data. Then I asked for some 3D WebGL charts with echarts and we got that working too.

So WASM is gonna be as important as ever because we have tons of software which can be compiled to WASM, Web is the UX meeting point, and LLMs can help bring it all together.


Thanks for sharing and taking the time to document your repo. I’m also sometimes unsure of “self-promotion” — especially when you don’t have anything to sell, including yourself.

I sometimes don’t share links, due to this and then sometimes overshare or miss the mark on relevance.

But sometimes when I do share people are excited about it, so I’ve leaned more to sharing. Worst is you get some downvotes or negative comments, so why not if there is some lurker who might get benefit.

When you don’t blog or influence, how else but in related HN comment threads are like-minded people gonna know about some random GitHub repo?

My second level hope is that it gets picked up by AI crawlers and get aligned somewhere in the latent space to help prompters find it.

ETA: “The [Prompt Engineer] skill was optimized using itself.” That is a whole other self-promotional writeup possibility right there.


hah thanks for the compliment.

yeah last time I shared it, I got a whole lot of hate for vibe coder self promotional BS so I decided to tread a bit more carefully this time.

I encourage you to try to prompt engineer skill! It’s one of the easiest to use, and you can literally use it on anything, and you’ll also immediately see how the “dynamic prompt workflow” works.


It's techno-freemasonry. One must break through the symbolism. The author wielding it and transmitting it cannot just plainly say the knowledge. We don't have the vocabulary or grammar for these new things, so storytelling and story universes convey it. The zoomorphism and cinematic references ground us in what all these bots are doing mimetically.

I'm excited the author shared and so exuberantly; that said I did quick-scroll a bunch of it. It is its own kind of mind-altering substance, but we have access to mind-bending things.

If you look at my AgentDank repo [1], one could see a tool for finding weed, or you could see connecting world intelligence with SQL fluency and pairing it with curated structured data to merge the probabilistic with the deterministic computing forms. Which I quickly applied to the OSX Screentime database [2].

Vibe coding turned a corner in November and I'm creating software in ways I would have never imagined. Along with the multimodal capabilities, things are getting weirder than ever.

Mr Yegge now needs to add a whole slew of characters to Gas Town to maintain multi-modal inputs and outputs and artifacts.

Just two days I go, I had LLMs positioning virtual cameras to render 3D models it created using the Swift language after looking at a picture of what to make, and then "looking" at the results to see the next code changes. Crazy. [3]

ETA: It was only 14 months earlier that I was amazed that a multi-modal model could identify a trend in a chart [4].

[1] https://github.com/AgentDank/dank-mcp

[2] https://github.com/AgentDank/screentime-mcp

[3] https://github.com/ConAcademy/WeaselToonCadova/

[4] https://github.com/NimbleMarkets/ollamatea/blob/main/cmd/ot-...


The true social media. Walk up and stick a quarter on the cabinet. With the ever present sounds of bowling balls hitting pins at the Sports Center, you know exactly which one is yours out of the seven up there. Everybody hovering around, watching and kibitzing. Emotions bounce from stoic concentration to exuberant trash talk. Respect is briefly granted to the kid running the joystick for a half-hour until the hollers and applause when a frame perfect dragon punch knocks him out mid-kick, dethroning the current champ. Quarter laid up again, back in the line for the next dopamine hit shared with strangers.

We are more connected than ever, yet still so far apart.


We had an Asian store across from the middle school where I hung out and we played Street Fighter for hours after school. The second generation Hmong that came out of the Vietnam war would would hang out and play. We all loved it! I'd often play Ken and they'd play Ryu, haha, we love our avatars. Sometimes I gave them a run for their money, sometimes they taught me new techniques, like a new sequence of moves.

Some of the other kids on my street went to private schools and I think they missed out on some of the lessons/bonding I got from interacting with a variety of people in public school. It's good to get out into social setting and mix it up with folks.


> Some of the other kids on my street went to private schools and I think they missed out on some of the lessons/bonding I got from interacting with a variety of people in public school. It's good to get out into social setting and mix it up with folks.

I went to private school, and would "miss the bus" after school on purpose so I would have to take the city bus home. There was an arcade in downtown Minneapolis a few blocks from the school where I'd hang out and play Mortal Kombat for an hour or two before heading home. Maybe stop by Shinders on the way to the bus stop to grab the latest copy of Wired or whatnot.

Definitely let me get out of the private school bubble a bit, and gave me some lifelong problem solving skills - both socially and practically speaking.


Wired was such a great magazine to read as a teen in the 90s. I remember just itching for the next issue.


Yet there was always that one kid that knew how to soft-lock Street Fighter II arcade cabinets with Guile. Samurai Shodown, The King of Fighters, and Mortal Kombat were also fun. =3


I had a pre-teen death rivalry with another kid over Samurai Showdown. I was the blue tuberculosis guy and he was the long haired samurai, and we'd meet Tuesday-Friday afternoons to burn quarters killing each other. At one point the owners used toy finger-handcuffs to tie us to the machine until one of us won. There was pizza!

I've always wondered what happened to him.


>I've always wondered what happened to him.

Probably, just ran out of Quarters for awhile... or discovered Final Fantasy VII on PlayStation. =3


Note that this wasn't a global phenomenon. We had SF2 in Spanish arcades, and I a lot of people wanting to play, but basically nobody played versus in the arcades I visited: The game was too expensive to make someone's investment last a single 3 round fight. So instead you'd see a line of people waiting to play single player, and helping each other out.

The multiplayer games that did well were all PvE, like Gauntlet or Knights of the Round. A very different culture.


> The game was too expensive to make someone's investment last a single 3 round fight. So instead you'd see a line of people waiting to play single player, and helping each other out.

Custom among my friends was to put a quarter in, and wait to press start until the computer was about to win its second round. Then the challenged essentially got to play you for free.

Doesn't work as well when it was two coins to start, one to continue though; in that case, the next challenger would rather pay one coin to challenge right away rather than letting the winner play the computer until almost dead at the price of two coins.


I recall during the 90s spending a bunch of time on SF2 and Mortal Kombat in arcades: shopping malls, bowling alleys, even some restaurant/bars that had a small arcade. One of the fun experiences was one arcade that Saturday mornings they had a "Freeplay" time for a few hours where everyone paid like $5 and every game was in Freeplay mode. It always amazes me how we all learned the special moves and fatalities word of mouth and eventually they'd get published in gaming magazines. The whole winner stays, loser pays - folks setting their quarter on the arcade to reserve their next spot. Many years later a coworker and I bought a very well used (cigarette smell and burns) MK2 machine for the office break area that took me back. Comically we found at least $10 worth of quarters inside the enclosure. Good times.


It was somewhat anesthetized compared to running out into an empty lot to play football but yeah better than staring at a screen and calling it networking.


Certainly! I left out the part where I would ride my BMX bike with my friends, without a helmet and without my parents tracking me, 5 miles in an urban environment to get to the arcade, then we'd hike the Hollywood Hills fire trails afterward.


Very Terminator 2!


Also giving off some Pet Sematary Part Two vibes.


I'm not so sure... Where I grew up there was no arcade.

It's easy to say that we are more connected but far apart, but only if you ignore the democratization that has come with that connectivity.


Might be a stretch, but have you tried climbing or similar social sports?

Climbing is a bit more for young adults, but since the wall is a shared resource, you have a lot of these social interactions, and it's mostly strangers as well, you just walk up, pay for a couple of hours and start climbing.

I'm sure there's social sports more appropriate for adults and elderly as well.


Even just keeping the focus on fighting video games, the community still exists. Search for something like "<your city> FGC" (ie "fighting game community") and you'll probably come up with several hits. Join the Discords, figure out the schedule for local events, then just hang out and be a positive person. I was playing SF2 with folks in person just this past Saturday.


You can still go out there and do things like join a running club. People are still going to arcades in Japan.

Comments like these are kind of ironic.


Nowadays' Japanese arcades are not like the ones GP is describing, most players don't interact with each other directly anymore.

Notable exceptions are places like Mikado centers that organize tournaments and keep the old flame alive.


I don't think the culture is the same due to cabinets having network capabilities now, but I do think it's possible.

At the taito station in Akihabara, I've met tourists a few times when I was in town for a large tournament (EVO Japan) and made friends from it. I've also had people watching me play, but unfortunately I don't speak Japanese.

I know there's a few arcades that still have some street fighter III: third strike cabinets with regulars. I can't speak for other games but at least for street fighter, people are almost always open and friendly.


I was there 2 years ago. I went inside one of the multi storey gaming places in Akihabara. The old school (90's and older) era games are a small section in one floor when there is 6 storeys of gaming.


That sounds like the Taito Station on the right side of the street. On the other side there is a Gigo with a whole floor for retro games, and Hey! that is focused almost only on retro games.


> Comments like these are kind of ironic.

Why, because there is one country in the world where this doesn't apply?

It's a commentary on modern Western culture, not a request for hobby suggestions.


>not a request for hobby suggestions

Of course it's not. Why look at anything positive or actually do something when you can instead engage in the tired tropes like looking at the past with rose tinted glasses as a way of comforting yourself.


You can be as positive as you want to be, and should absolutely take action and do things to better socialize.

But to pretend it’s remotely the same as it was 40 years ago is utterly ridiculous. Now when you do such things like a running club you are joining a group of very self-selecting people who for the most part have a certain personality type.

You simply do not get the diversity in group experiences as there used to be. It was either go through social discomfort or sit alone bored with zero social interaction. Now the friction to get that social dopamine hit is extremely low bar, and going beyond it the bar has been raised considerably.

Not to mention doing stuff like running club or rock climbing just feeds into the hyper-scheduled world the west has become. Spontaneous social interaction is important too, and those third places are increasingly scarce and involve far more friction. Which again self-selects for certain personality types and lifestyles.

For some people these changes are positive - much easier to find niche activities to do with others. For other people they are extremely negative.


>Now when you do such things like a running club you are joining a group of very self-selecting people

I'd disagree with this pretty strongly. I do workshops at a makerspace in Berlin, which is in itself a pretty nerdy place but we've got everything from pensioners to middle aged moms to obviously a lot of people from the university or tech work.

In much smaller cities not just here you'll find chess clubs, poetry slam groups, church choirs what have you. None of it hyper-scheduled or commercialized. I can't speak to what this was like 40 years ago I wasn't alive then but there's no shortage. I think the biggest difference is, people don't move. In the Western world mobility is at an all time low. If you were young and lived in a place where these opportunities didn't exist people literally just packed their bags and relocated. In the words of Morgan Freeman: https://youtu.be/oZcSivXEGys


I definitely agree that a big part of it is how immobile people are these days.

What I meant by hyperscheduled is that typically these activities revolve around setting a schedule in advance for everyone to commit to and plan around. This sort of thing simply does not work for me. At all. Maybe once every couple months or so.

For example my local makerspace requires at least days (if not weeks) advance booking for most tools. When I’m in a project mood or want to meet up with friends to hack on something it will be more of a “hey let’s go figure this out, meet you there in an hour!” situation.

What I personally miss are the social clubs/spaces - heck even neighborhood pubs - that used to exist as simple meeting points. Whoever was there happened to be there and you’d tend to slowly make more social connections over time. You show up when you felt like showing up, and probably find a handful of casual friends there no matter when you’d go.

There is an extreme dearth of such impromptu meeting points/gathering hubs at least where I live. If you want to meet with friends you typically are going to schedule it a few days out - even if it’s meeting up for drinks after work. With work from home that’s even far less of a thing since even coworkers are geographically dispersed vs. cutting out of work 30 minutes early to go grab drinks at the bar around the corner.

By the time I get through my exhausting work calendar each week all I want is some control over my day back - and let the day go by feels vs a calendar. This is the largest difference other than social media I’ve felt over the past few decades.


That's "just life" unfortunately. By the time most people reach their mid-30s, they accumulate enough commitments that they have to plan things in advance or they just don't happen.

If a friend turned up randomly and unannounced at my door and asked if I want to go to the pub, they have the following barriers to overcome:

I'm out for 10 hours a day during the work week and I'm asleep another ~8h. In the remaining 6h I want to go to the gym, I have to eat, I have to run errands, I have to spend time with my wife and in the remaining time I might just want some alone time. Odds are I'd have to turn my friend away, which would make me feel bad, even though I'd have gladly joined him in the pub if I could plan for it. And I have no kids! If I had kids, the odds are even worse.

We all wish we could be back in our 20s when we had all the freedom and none of the commitments, but the truth is that for 95% of people this isn't possible. When you are in uni (or fresh out), you have all the time and energy, but (generally) no money. So you can spend a lot of time with friends who are in a similar position. By mid-life (30-55), you have money and energy, but no time. And in your winter years, you have money and time, but no energy left.

In each of these phases, you can try to go against the flow and experience friction all the time, or you can try to make the most of it and adapt.

If you absolutely desire the freeform approach you describe, perhaps you need to step up and establish the clubs and spaces you'd like and select for members who have a similar desire.

Most spaces in cities have to cater to the lowest common denominator, and simply wouldn't be able to function without strict scheduling rules. It'd be patently unfair if the 20 year old uni student hogged the equipment when you turn up for your 2 hours of free time that you planned a week in advance, wouldn't it?


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: