Anthropic engineer Thariq Shihipar says Claude users should manage projects and collaborate with AI agents through Artifacts rather than relying entirely on chat. Shihipar, a Member of the Technical Staff at Anthropic, made the case in a recent podcast interview on Latent Space. Artifacts, a feature Anthropic introduced in 2024, lets users clarify requirements, communicate preferences and collaborate with agents, he argues.

The adoption picture is stark. A Claude AI Facebook group with nearly three million users rarely hosts discussions about Artifacts, suggesting the feature is not well understood. Shihipar acknowledges users have not embraced Artifacts as a project management tool.

What does Shihipar recommend?

He wants users to build dashboard Artifacts for long-term projects. In his words:

"And so one thing that people are not doing yet that I'm trying to encourage is this idea of a dashboard artifact. So you have Claude working on a project long term. Maybe it's like a Kanban or something. It can store that Kanban data in a database. Multiple Claudes can access that data via the Artifact MCP. And that Artifact can talk to those Claudes as well."

He concedes the approach is demanding. Building these dashboards requires far more detail than a simple prompt.

"But there's a lot of work to do because it's so much more complicated than a multiple choice question. There's a lot more detail in terms of diagrams and code snippets and schemas or whatever it is for that problem. But Artifacts is the more AGI-pilled way of doing the, 'ask you the question,' basically."

Why did the podcast co-host push back?

Latent Space co-host SWYX challenged the recommendation, saying he prefers working directly with Claude without Artifacts. His concern is where feedback belongs:

"I think one thing that's unclear to me about the Artifact stuff is what feedback should go in through the Artifact and what feedback should go through a Claude, the chat. Because the more AGI-pilled one is to just feed everything to the Claude."

Shihipar disagreed. He sees Artifacts becoming the primary interface to AI agents:

"I think the more AGI-pilled one is to go through the artifact. And I think that we sort of imagine in the limit. I think that Artifacts will be your interface into the harness. You can comment on this live document of your plan of the work. You can see maybe multiple agents and different agents are doing this. And that Artifact is built for the current work that you're doing."

He added that Anthropic has not finished building the infrastructure: "I think we're still getting there from an infrastructure perspective. But yeah, I think on-the-fly interface for your harness is probably where things are headed."

How would local and cloud agents split the work?

Co-host Vihu Sapra asked why users should not simply do everything through Claude Code, which already returns rich HTML mock-ups, diagrams and other output. Shihipar answered by splitting the agent experience into three layers:

• The display layer. A surface UI hosted as an Artifact, with its own database, showing all agent work.
• The inference layer. The intelligence runs in the cloud, so users do not have to keep their computers on.
• The execution layer. "Hands" that can work locally on the user's machine, in a remote sandbox or wherever the work needs to happen.

Shihipar described the direction of travel: users will message a cloud-based Claude that can spin off many sub-agents, which communicate with each other. Those "local hands" will let an agent access the user's computer when it is online. "That's unpackaging the Claude Code experience right now. Right now, it all happens in one place," he said.

What is the takeaway for marketers using AI agents?

Shihipar believes Artifacts help users understand what agents are doing and guide their work more effectively than chat alone. He also recognizes Anthropic has more work to do before his vision of agent collaboration through Artifacts fully materializes. For teams already running multi-agent workflows in Claude Code, the message is to start treating Artifacts as the project dashboard and interface layer now, ahead of the infrastructure Anthropic plans to deliver.