new Claude feature: Let's build an animation!

It's Sunday and Claude has released some new features, so you know, let's play with them. I want to first start off by saying that I am not technical. In college, I studied like communications, media production and entrepreneurship, so zero CS background. So this is exactly how I go and try to learn new things. One of my tips is I love having dictation mode enabled on my laptop. So I use my MacBook, but go into your settings so that you can make a hotkey. I use my FN key and I tap it twice and it turns on dictation. So you'll see me use that so I can talk directly to my laptop instead of typing. Hey, so I think you released some new features, uh, especially visualization when it comes to learning. And I have an idea for making a video about innovation cycles and how mass populations adapt to new innovations that come to the market. Could you build that visualization for that and then also teach me how to use this tool better in the future. Okay, let's see. Okay, I'll come back when it's done. Okay, I think this took about two minutes for it to make and built this visualization. Here's the core framework. Now let me show the second layer, what's actually happening psychologically and socially at each stage, which is the more interesting part for a video. What drives each group, the real mechanics. What actually crosses the chasm. Where most innovations stall. Modern accelerators. Now, how to use this tool well going forward. What the Visualizer actually is: it renders inline SVG and HTML widgets directly in chat. Two modes: pure SVG for static diagrams, HTML for interactive elements (sliders, buttons, live state, charts via Chart.js). It's not image generation. It's code that renders in the browser. How to get the best output from it: Be specific about what you want visualized. "Make a diagram of X" is fine. "Make an interactive diagram where I can drag a slider to see Y change" is better. The tool supports true interactivity - sliders, toggles, click events, animations. You're not limited to static pictures. Tell me the purpose: "For a video" vs "for my own understanding" vs "to show investors" changes the density, color choices, and level of annotation I'll use. Request multiple diagrams when the topic has layers. I just built two - one for the structure, one for the mechanics. A third showing real historical examples (radio -> TV -> internet -> mobile -> AI) with timelines would complete the video asset. Ask for that version if you want it. Push on interactivity explicitly. I defaulted to static SVG here. I could have built this as a clickable stepper where each segment lights up with examples and quotes when you tap it - that would be far more useful as a video script tool. Ask for that version if you want it. The honest limitation: Complex layouts break if I try to cram too much into one diagram. When I see you asking for something to dense, I'll split it rather than produce a cluttered mess. Want me to build a third diagram - an interactive, clickable version of the S-curve where I can show us how to use this tool for, and content is to generate a interactive, 20 second video of the innovation curve that I'm going to use for a different video. So Claude asked me questions saying, hey, like, which arc should we animate? Like, what should the animation style feel like that? And it started building it. Okay, this really sucks. So I'm gonna take a screenshot of this and tell Claude to like, fix itself. Some tips that Claude had told me to do, like, get a better output is describe exactly what the user is going to be seeing or how you want to exactly interact and like click around with that thing that you're built. Like making a diagram. I said, I want you to make a 30 second animation where you display first, like, the timeline after that timeline is done, you will then start with, like, the AI timeline where we are in terms of adapting to that innovation. So I'll show the animation and then the second. I am pretty happy with how this turned out. I feel like I could probably use this in a video with just a few more minor tweaks. That would just take me an extra five minutes. But I did start this session at around 2:20 and it's 2:45. So being able to build this to 30 minutes and also make two different videos is pretty awesome. So now if I hit replay, it'll go through the current events that Claude researched that felt like events that led to different points in the mobile revolution. So it's going to go all the way and then end in about 2015. Once again, it's a little wonky that I can go fix later. But now it's gonna change into the AI one. I could definitely use this and talk about this in a video, um, and show you where we're at currently within the AI adoption curve. So here we go. Here's more stats. Boom, you're gonna end here. Once again, I had an idea for a video and there wasn't anything out there that suited, like, talk about, because everything is so outdated. Like, I could just build my own tool with, like, information that came out yesterday, right? This changes so much how we think about learning and what it means to build things. And it makes me to embrace even more human creativity, because the more different domains you have access to to learn about, the more breadth you have to combine different things and then make new things. Let me know what you think I should be building next. You in the next video.