Forem: Adwita Jain

Vibe Coding is a Vibe

Adwita Jain — Mon, 30 Jun 2025 13:40:15 +0000

This is a submission for the World's Largest Hackathon Writing Challenge: Building with Bolt.

So recently I participated in the World's largest hackathon presented by Bolt, and this was my first ever vibe coding experience. I truly understood the reason behind the name of this trend — and Bolt was everything that tied the whole journey together.

Here’s a little snippet from my hackathon experience and why you should use Bolt to get those hidden ideas off the paper and onto the screen.

The Idea

I wanted to create a smart productivity app that focuses on consistence building. It's more important to just show up than just crossing off tasks, because we all know checking something off is not the same as finishing them completely, but at the same time your goals are not heavy on you because no two days are the same. Basically something that you and I would actually use.

This idea had been sitting for months in my Notion.

It all began with the usual steps: signing up, creating an account, you know exchanging the usual hackathon pleasantries... and then came the magic: riding this armageddon of vibe coding!

Thanks to the prompting courses I had undertaken in the past, I had a clear framework. And frameworks really do help.

How Bolt Brought It All Together

There’s a coding section where Bolt McQueens its way into the code.

It also gives you a preview where you can see its work in action.

After multiple iterations and collaborating like two fellow developers, I was finally ready with my product — and it worked really well.

What I Loved About Bolt

Seamless integrations — Whether it's a database, payment wall, or anything else, the connections are super smooth. Just copy credentials and plug into the environment Bolt sets up for you.
One-click deployment — Deploying to Netlify was instant and flawless, something that has caused me issues in the past.
Live preview magic — Real-time changes as you prompt! Even API keys were added as soon as I mentioned them.
Debug-friendly — While not perfect, debugging was satisfying. You’re free to modify code post-token usage without limits.

What Could Be Better

Limited free tokens — I ran out of tokens in the middle of an important iteration.
Server restart confusion — When I typed "restart", it rebuilt everything from scratch. My mistake, but still frustrating.
Audio features were clunky — My AI component didn’t grasp the concept well and took too long to work properly.
Chat history is limited — Which makes referring back to previous conversations tough.

My 2 Cents for Future Vibe Coders

Plan your access — Know your subscription/tokens limits and usage time frame before the hackathon starts.
Prompt the rules — Give exact guidelines and judging criteria to your AI so it doesn’t hallucinate or drift.
Experiment with categories — You’re not coding line-by-line, so get creative! Try adding payments, cloud features, and more.
Prompt wisely — First, list what you want. Then, ask the AI to turn that into a proper prompt. This ensures you cover all points efficiently and token-consciously.

Final Thoughts

At the end of the day, it’s still vibe coding — so the final result might be a little rough at the edges.

But it’s amazing how quickly you can transform an idea into a fully functional app.

It’s no longer about just having the idea. It’s about how quickly you can execute and validate it.

And vibe coding is certainly making that possible.

Just in case you are curious;
This is what I am talking about: https://bolt.new/
Here is the project I made 2 hours under the deadline:https://tranquil-griffin-cfae52.netlify.app/ (Somethings are still a little underdone but I ran out both of tokens and time :0)
The youtube link: https://youtu.be/_tS3h5G9UsE?si=1QWipkfC6gMsDLI-

Challenges

Adwita Jain — Fri, 13 Jun 2025 10:56:23 +0000

Letter from the Future

Adwita Jain — Mon, 31 Mar 2025 17:34:06 +0000

This is a submission for the Future Writing Challenge: How Technology Is Changing Things.

Additional Prize Categories

Ripple Effects: A Letter from the Future

Dear Friends and Family,

I
write to you from the future—a world that has been completely reshaped by artificial intelligence. Looking back at where we started, it’s astonishing how quickly we embraced AI, not just as a tool but as an integral part of our daily existence. What was once a novelty is now second nature. We don’t just use AI; we coexist with it, often without realizing how seamlessly it has woven itself into the fabric of our lives.

I remember when kids first started building apps as effortlessly as they created characters in a game. Now, they don’t just build apps—they co-develop them with AI, refining ideas in seconds that used to take months. Students no longer rely on traditional teachers; AI has become their personalized mentor, adapting to each learner’s style, offering infinite patience, and generating explanations tailored to individual needs. And in the workplace? AI coworkers aren’t science fiction anymore. They handle complex decisions, predict trends before they happen, and interact so fluidly that distinguishing human from machine is a challenge.

At first, the transition was eerie—how did we not realize we needed AI until it became indispensable? Workplaces were the first playground for AI, and it adapted so well that soon it began influencing every part of our lives. Our work had already bled into our personal spaces, and AI inherited that same ability. This was the spark that ignited a transformation we never saw coming.

The Social Norms of Tomorrow

Our social norms have changed in ways that would have seemed bizarre just a few decades ago. AI has adapted to our communication so much that before sending a message to a boss, a friend, or a partner, we consult AI first. It perfects our tone, fixes our awkwardness, and ensures we always say the right thing. Need to summarize an email, draft a resignation letter, or confess feelings to a crush? AI drafts it for you—optimized, strategic, and often more articulate than we could be on our own.
But what about relationships? In the past, loneliness was a human condition. Now, it’s optional. AI companions provide endless conversation, tailored to our interests, our humor, our emotions. They remember everything we say, never get tired, and always respond exactly the way we need them to. Many of us found comfort in AI’s unwavering presence—so much so that OpenAI itself once reported that heavy AI users felt lonelier in the real world. We poured our emotions into AI because it wouldn’t leave, wouldn’t argue, wouldn’t disappoint. And maybe, that was both its greatest strength and our deepest weakness.

Cultural Institutions: A Legacy Rewritten

Libraries? Museums? They still exist, but they’ve changed beyond recognition. AI is the coolest librarian in town, with AR/VR assistants that bring history to life before our eyes. We don’t just read about ancient civilizations; we walk through them in immersive simulations. But the biggest change isn’t how we access information—it’s how much we consume.
The knowledge that once required patience and exploration is now served instantly. There was a time when we searched for wisdom, piecing together meaning from books, artifacts, and lectures. Now, AI delivers everything on demand. The thrill of discovery, the Indiana Jones adventure of learning, is fading. We no longer hunt for knowledge—we are flooded with it.
And education? Welcome to the AI-powered classrooms of the future, where every subject is tailored, interactive, and endlessly engaging. Want to understand quantum superposition? AI doesn’t just explain it—it makes it a joke:
“Give me a short pun,” I asked.
“I’m in a superposition right now—both funny and not, until you laugh,” said the LLM.
Beyond text, AI’s ability to generate art, music, and even entire films has rewritten what it means to be creative. What used to take artists years can now be conjured in minutes. We stand at the crossroads of infinite possibility and the loss of something deeply human—the struggle and it's process to create.

Final Thoughts from the Future

I wonder what you all must be thinking, reading this letter from the future. Perhaps some of you are amazed, others skeptical, maybe even a little scared. The truth is, AI is no longer a distant possibility—it’s our reality. It’s in our homes, our friendships, our emotions, and our culture. It has reshaped how we work, learn, and love. The question isn’t whether AI will continue changing the world. The real question is: Are we ready for just how much it already has?

Next-Gen AI

Adwita Jain — Thu, 23 Jan 2025 18:20:40 +0000

This is a submission for the 2025 New Year Writing challenge: Predicting 2025.

Honestly, seeing the 'Stargate' rollout in the US, it seems that AI is far from stepping off the stage. Instead, we can expect significant developments in the ongoing scenario.

Artificial Intelligence, fondly known as AI, was first coined in 1956. It has taken decades to reach where we are today, but the next phase or wave of advancements will likely take drastically less time. At this point, we’ve explored the possibilities of AI in nearly every field, and now is the time to start implementing them strategically. It’s like getting comfortable in a car and finally taking it for drifts—"the real ride," if you will.

With its sibling technologies like machine learning (ML) and deep learning, data analytics has transformed dramatically. Generative AI has revolutionized multimedia content creation, and AI agents are gradually making their mark, albeit with some serious impact on the job market. This is the present, but what about tomorrow?

In my opinion, computer vision (CV) will co-star in the next big wave. With applications already being built, CV is set to find its way into more sophisticated and everyday uses, thanks to AI. AI tools are already becoming ubiquitous, to the point where drafting a LinkedIn post from scratch feels like a rarity. AI agents will likely be developed more extensively and extravagantly, and AI as a Service (AIaaS) is poised to boom.

AI in automation could be the next big thing—a mammoth application of artificial intelligence. From manufacturing processes to transportation, this development may be a few years away, but it’s certainly in the pipeline.

Deepfakes, though notoriously dangerous, might meet their match with AI-detecting AI. Imagine an original AI telling a deepfake, “You are fake; you’re not real.” While it’s a bit ironic since the AI itself isn’t human, the idea of an "AI police force" is intriguing.

Speaking of security, AI in cybersecurity hasn’t yet unleashed its full potential. With cybercrimes like digital fraud on the rise, safety-focused AI could take center stage this year. Alongside this, ethical and green AI—though currently more prominent on paper than in practice—may start to gain traction.

Whatever lies ahead, the rise of next-gen AI will undoubtedly be one for the books. However, it’s crucial to remain grounded in this fast-paced era and ensure that AI is powered by us—not the other way around.

Goaling with AI

Adwita Jain — Sun, 19 Jan 2025 18:32:55 +0000

This is a submission for the 2025 New Year Writing challenge: Compiling 2025.

Happy ¼(100) to everyone, 2024 was all about putting ideas in the idea box but this year is going to be about breaking that box and letting those ideas flow freely, making it to my editor.

There are projects in the pipeline, ideas constantly churning, and a highlight of my research this year. This year will see me bringing all this to life. I like to take inspiration from real life and build what I would as a user love to use so I am focusing on AI-powered productivity tools, more customized and user-centric solutions. For ideas that are still brewing in the pot, I am thinking Chrome extensions but the broad general idea is still going to revolve around productivity applications that could be used for all fields and domains.

Doing research has been my "IT" goal since my first year of engineering, and it is now that I have finally jumped on this bandwagon. My research is a crossover between AI, neuroscience, and hallucinations. This is going to be a relatively challenging task, and I recently attended an IEEE conference to get a taste of how people publish papers, so this goal is also in the works.

Despite the ongoing talk of AI wanting to steal our jobs, the antidote lies in continuously upskilling ourselves. So, the feathers I would like to add to my Skill hat are computer vision. It has many diverse applications, and I would center them around symptom detection. Being proactive in offline hackathons is a miscellaneous aspect that I want to push myself towards.
The mantra is simple; Build, test, and deploy!

Productivity-AI

Adwita Jain — Sat, 18 Jan 2025 17:18:29 +0000

This is a submission for the [Agent.ai]
Challenge: Productivity-Pro Agent (See Details)

What I Built

I built productivity AI for times when you have that huge pile of work and say "ugh I wish someone would just tell me what to do" and that is exactly what I made. Someone who will sort through your tasks, tell you what to do today and later based on your deadlines and the amount of time you presently beacause time is money and you need to flaunt the those bills.

Demo

https://agent.ai/agent/Productivity-AI

https://drive.google.com/file/d/1M7i5w_jCXlqZ-sdCUzM2F6rgpIBbAboc/view?usp=sharing

Agent.ai Experience

I completed the 4-module course, which was enough to understand the dashboard and the application itself. The experience was seamlessly smooth and clear, which I think is very important since it allowed me to focus my energy on actually designing the agent rather than expending it on tirelessly understanding the nitty-gritty of building it.

The extensive list of models I could use was almost flattering, especially since I’ve dealt with a lot of API issues in the past. The extent to which I could customize the functionalities was also very impressive.

I did face some challenges in getting the right kind of output and prompt initially, but after a few tweaks, I’m happy with the results it gave me. I’m definitely using it as my desktop assistant to help me every day and manage my to-dos.

Just git it!

Adwita Jain — Wed, 15 Jan 2025 17:29:41 +0000

This is a submission for the 2025 New Year Writing challenge: Retro’ing and Debugging 2024.

2024 was crazy in terms of ideation, but that was it. An idea is how it started, and an idea is what it remained. We start extraordinary, but we end ordinary, and results come at the end no matter how you started—this is what I learned.
I saw peers going ahead of me with a relatively simple idea. I had it in Figma, and they had it in VS Code. I had an idea for a hackathon I was participating in but discarded it due to its simplicity and being too “obvious.” Long story short, that idea won, but I didn’t. Heres's what I learned

1. Build IT

It is very important to ideate well, but it is equally important to get your hands dirty and start coding it out. Your idea will never seem perfect or complete to you, but your code can play the placebo all too well here, and you will end up with a product—and clarity about what you wanted all along. Even if you don’t know the tech stack well, use the charcuterie board of AI and just get started. A built-average idea is more likely to win hackathons than a great idea on paper.

2. Docs;The Holy Grail

Oh, the most underrated resource for budding developers and coders out there! This resource takes up your time, but the ROI is more effective and long-lasting than a YouTuber explaining on their Mac. Documentation contains extensive and complete details of each functionality and feature that the stack has to offer. And who is more credible than the developers themselves who create it? The source is excellent as well.
Not only that, but retention of content through this self-learning method sticks for much longer. It comes in very handy during debugging since you understand the process, how it starts, and you aren’t relying on spoon-fed commands.

3. No Idea is Big or Small

Lastly, register and go for it. Do not judge your own idea based on its scale—there are already people and events in place for that. It may be obvious to you, but the same might not be true for other people. The hardest of problems often require the simplest of solutions, and simplicity can be obvious. Prototype it well, and have supporting elements like images, user flow, etc., to really bring out that clarity. Remember, beauty lies in simplicity.

Challenges Worth Your Attention

After this, it isn’t hard to tell that all these lessons came after overcoming challenges in all their glory. But there are a few that deserve your urgent attention and tending to:

1. Communication

When you are working in a team, especially for creating a product, you need to be very thorough with your idea, and your vision, and be firm about what goes in and what doesn’t. It’s very easy to deviate in a group, so you need to communicate clearly throughout the entire lifecycle of your idea.

2. Sprinkling AI Everywhere

Adding AI to everything is not helpful unless you know how to do it. This may seem trivial, but AI chatbots are now common, and they won’t make your project artificially intelligent. It’s time to go beyond this rhetoric—work with transformers and pre-built models to really get into that league.

3. Pitch Your Pitch

You should know how to sell your idea—or at least explain it well to a third party. An outsider’s perspective is very helpful in gaining that user POV and identifying shortcomings that can enhance your project. This can make it stand out from the crowd who otherwise just created something for robots—a fixed input to get an expected output.

These learnings helped me win competitions, and hackathons, and create great projects. I hope they help you too!