Claude Artifacts Go Public

Share and edit Artifacts with anyone

Generated in Midjourney

Total reading time around 4 minutes.

Welcome to Visually AI!

🔮AI News This Week

Publish & Remix Claude Artifacts

Anthropic announced Claude Artifacts can be published publicly, allowing others to view and remix your creation, and can be remixed by others to build upon and modify the original content.

Publishing an Artifact makes it publicly accessible, while remixing allows you to start a new conversation and modify the Artifact.

Here are the key points to know:

• Publishing an Artifact makes it publicly accessible, allowing others to view and remix it

• Remixing allows you to build upon and modify the original content

• Published Artifacts are viewable on a separate, public-facing website

• Only the Artifact itself is published, not the surrounding conversation or context

• You can unpublish a specific version of an Artifact at any time

• Deleting a conversation containing a published Artifact will also remove the public version

---

My Artifacts Use Case: Building a Custom App

I built a cinematic video prompt generator using Claude Artifacts in about an hour.

Workflow:

I created a Project in Claude focused on AI video prompting and used this prompt: "Create a cinematic video prompt generator incorporating the factors and recommendations for effective video prompts in this Project knowledge base."

Claude generated a great text template, which I refined through follow-up requests to build an interactive app. I specified the maximum character count and categories to include.

I'm still adding more options and categories, so the app isn’t published yet. I tested the prompts in Runway’s Gen-3, and the results are impressive.

Here's an example followed by a demo of the prompt generator in Claude’s Artifact preview window:

---

Poe Previews

Poe Previews is a new feature that allows users to create and interact with custom web applications directly in chat, without requiring programming knowledge.

With Previews, users can build a wide range of interactive experiences, such as games, animations, and data visualizations, and share them with others via a dedicated link.

The feature is available for all users on the web, supports HTML output, and will soon be expanded to include additional formats.

I asked the Claude-3.5-Sonnet bot on Poe to create a character name generator with customizable options for gender and genre. The bot generated the HTML code to my specifications, and I was able to make additional requests to refine the functionality.

However, as a free user, I had limited requests available. In hindsight, I would plan my project better before starting a chat to ensure I get the most out of the bot's capabilities.

Try Poe Previews for free. You can see the process here:

You could have your AI service, tool, or event seen by Visually AI’s community of over 15,000 subscribers:

🚀 This Week’s AI Tools

Mirrorize: Launched maestro/v1 video fx generator, a simple high resolution text to video generator. (link)

Leonardo Describe with AI: Create prompts from images with Leonardo describing what it sees. (link)

Krea Scene Transfer: Create new scenes in seconds with perfect light and color consistency, and it’s free! (link)

LivePortrait: A new research paper from Kuaishou lets you animate images from a video reference. I used the HuggingFace demo with a Midjourney image and then I animated my own selfie, guided by an example video. (link) (project page)

---

💻Visually AI on YouTube

My co-host Miguel and I posted Episode 2 of our new collaboration series, Under the Hood.

This week, we talked about using ComfyUI workflows on the Glif platform to generate images:

Under the Hood | Episode 2: Intro to ComfyUI Workflows with Glifs

---

🖼️ Image Prompts

Prompt: An image of a dark, brooding figure sitting alone on a rooftop at night, overlooking a cityscape with tall buildings and a river.

Generated in Midjourney

Prompt: A cinematic still of a sleek black car speeding down a deserted highway at dawn, long shot, warm light with soft shadows, dramatic and thrilling, shot on an Arri Alexa, 8K.

.Generated in Midjourney

---

🎞️ Video Prompt

This was shared by my friend, Kallenia D on 𝕏 for image models, and I added “Hyperlapse, documentary style, VFX,” for dynamic motion to use as a video prompt for Gen-3.

I used ElevenLabs Video to Sound app to match sound effects.

Prompt: Hyperlapse: A multidimensional trippy colorful kite flying above a bleak black and white scene of a drab boring black and white neighborhood, documentary style, VFX

Thanks for reading, and have a creative week!