Alex Gedevani
The age of 'idea guys' taking back control is upon us, and thanks to AI, the gap between concept and creation is vanishing faster than ever. Anyone with internet access and anywhere on the IQ curve, can now become a creator, whether human, agent, or AI.
We'll soon look back nostalgically on the days when turning an idea into reality started with just a pen and paper. "Prompt-to-create" is now rapidly transforming workflows across nearly every function.
As an idea guy myself, I recently tested out some of the latest prompt-to-app platforms to see the latest progress. My experiments led me to "building" two apps this weekend: Apple Avalanche, a casual arcade game and Thirst Trap, a habit tracking app for drinking water.
Here are some of my reflections on the process and its implications.
Apple Avalanche is a simple game. A user controls dogwifhat with arrow keys to catch falling apples for points while dodging rotten ones. Rosebud is a creator first platform that makes game development and UGC accessible by using gen ai to reduce friction between concepts and the creation of games. Some may call it the Midjourney moment of games.
What I liked here is the ability to easily remix. I remixed the original 2d project/background and added in my own preferences. If I understand correctly, somebody can now take this same exact game and remix it in their own way if they please (Like swapping out the main character with own IP and rewarding players based on reputation accrued for playing the game).
Within the platform, I integrated dogwifhat as an asset to play with and toggled the different control speeds I wanted all through natural language prompting with code changing on the backend. I also implemented a points system for collecting good / bad apples. I generated a custom song about this game on Sona AI and embedded it as background music. Creating assets directly with the game was also simple, just prompt it. I leaned on chatgpt to formulate a specific prompt and then plugged that in to create the asset I wanted for the rotten apple.
Rosebud is a Sfermion portco, get in touch if you'd like to find a way to work with them!
Thirst Trap is a lightweight web app I spun up to gamify water intake. Users log their hydration by tapping a glass of water, which simultaneously nurtures a virtual plant and reveals a fun fact about water. Separately, I experimented with a multiplayer version where you can play with friends and have your water logs stack up vs others in a bar chart. The no-code editing could improve but remains powerful. Similar to how back when AI image generators took off, descriptiveness was important and many leveraged another AI to aid with crafting prompts, I can see AI-assisted no-code prompting being another way to ramp up.
I used Ohara, a no-code platform making it easy to turn ideas into apps and agents. What stood out with Ohara.AI was its ability to quickly connect APIs, offering building blocks for app expansion. In this instance with Thirst Trap, I connected perplexity API for fun facts. Text to image generation could also be used to create a more gamified version where the more you drink, the more an unrevealed egg hatches and reveals visually a water monster collectible.
There will be no shortage of no-code development platforms across web2 and web3 (Alchemist) coming to market as well. Over time, I expect most of these to directly provide the opportunity to incorporate crypto and augment the agent/app stack.
Reflections on the "click to create" economy
Revisiting my earlier screenshot post above, expecting the creative economy to only accelerate from here. As Sasha from Creator Ventures pointed out, we are in the age of slop, with many of even the top monthly grossing non-game apps benefiting from this trend.
Today is the hardest it will ever be to create. The quality of tools at our disposal will only get better.
The influx of apps and games will create more "15 minutes of fame" moments, like the beer app that once earned $10k-$20k daily. As accessible products have dominated crypto, we may now enter a mini-clip games era, but this time with lightweight AI-generated games becoming the next growth wave, created in just days.
In a crypto world heavily focused on infrastructure, people have been asking, 'Where are the crypto apps?'. Well, it looks like we have found the accelerant to this. AI-assisted creation will simply 100x+ the amount of experiments we are seeing.
As seen above, you can take it from the app kingmaker himself. The shift to an AI-first approach is well underway. App and agent creation is becoming largely commoditized. Things may shift from app development itself to curating data, refining models, etc.
For crypto ecosystems, this acceleration in developer velocity has significant implications:
Top-of-Funnel Growth: Lower barriers to entry for app creation encourages more creators, which in turn leads to more apps drawing new users into existing and new crypto ecosystems. Which crypto infra layer will be the first to embrace an AI-centric crypto app GTM approach for onboarding new devs and users?
Community Engagement: Fast, compelling experiences deployed to accommodate metas can deepen engagement between communities and NFT/IP-based projects. People simply want fun and productive things to do with their assets.
Open-Source Remixability: By default, many of these tools are open-source, which could enable network effects as creators build and remix atop each other’s work.
Additionally, integrating AI agents capable of autonomously creating games or apps opens up a new design space as well.
Look, I'm just an idea guy who hastily put together these experiments to test what's possible. I'm excited to spend more time "doing things" vs "thinking things" and learning on the fly. A year from now I hope to look back at this post and laugh at how much has changed since then.
What would we do without our beloved "memecoin devs" in the creative economy supercycle? Next will be an incoming wave of "app developers" and the pump fun for apps.
The only limit now is our imagination. Lock in.