Parabolica #41 - Future of no-code; Runway Gen-2 update
Future of no-code
Are you using ChatGPT to help you code? It’s increasingly getting easier to make things with the help of LLMs. You have your open source options too (CodeLlama).
With this workflow, you’d still have to copy-paste code from one window into the next. For instance, the ChatGPT tab into VSCode.
Well, if you’ve been wanted to create a website, no more “I don’t know how to code” excuses.
Soon, as long as you can type in English (or any language), you can make a website.
v0.dev, by Vercel, is in private alpha. I don’t have access yet but I’m eager to test it out. It’s been on my radar for a few months.
You’ll still need ChatGPT’s help with the backend if you’re not a software engineer (like me). But I’m eager to try it out and potentially stop using Bubble for my web app.
Check out some cool examples:
https://twitter.com/leeerob/status/1720993160245973020
https://twitter.com/jhooks/status/1720876155740532873
Runway’s new Gen-2 update
Runway is at the forefront of generative AI video. They released an update to their Gen-2 model last week. I tested it out and want show you what I’m seeing.
Here were the text2video prompts:
“Will Smith eating spaghetti”
“a tidal wave wormhole, zooming out”
“a woman looking at her cellphone, zooming out”
Compare that with Pika Labs and how Gen-2 results looked back in July.
The fingers and facial features have gotten much better.
Protip: Quickly test visual hooks with Gen-2 for your top video ads.
Improvements to Sovran.ai this week:
Control colors to keep your videos on brand
AI Voiceovers on final videos
You will soon be able to upload your own assets and brand guidelines.
Create smart ads in minutes.