Okay things are officially getting crazy. My 8 yr old son brainstormed an app idea w/ ChatGPT (just chatting back and forth like with a friend) called BoredBot to help kids figure out fun/enriching things to do when bored. Then we built the app in an hour from scratch to deployment with @Replit's new agent feature. Full chatbot, with voice to text and text to voice. The Replit agent set up the environment, wrote the code, debugged the bugs, and just told us what it needed to keep going (i.e. OpenAI API key). When we had more ideas to make it better, we just asked. Didn't write a single line of code. A few guesses on where we are going at the limit of all this (similar feeling messing with 01): 1) this agent stuff is no joke - legit felt like we had an expert programmer who just did what we asked and had good ideas along the way. This is weird to say but feels like execution could trend towards commoditization. 2) Creative insight and spark feel more valuable than ever, as will knowing how to collaborate with an AI to 10-100x iteration towards greatness, and having the taste to know when you've hit that mark. 3) It's a good time to be someone with range, comfortable spreading into new areas. A novice with AI can become "dangerous enough", incredibly quickly. Learning will accelerate too (i.e. the way @Replit was able to write a contextual explanation about a piece of code was fascinating). 4) If the above are true, I think we'll see a renaissance of solopreneurs / tiny startups innovating in all sorts of areas as people figure out how to use these tools to fill in the gaps of execution they lack and 10-100x the areas they excel. Was going to write this post any way but suffice to say I maxed out my @Replit Agent usage and this post can double serve as a pretty please request for more :)