Toggle sidebar Toggle sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Anthropic's Claude Can Now Create Interactive Visuals Directly in Conversations

MacRumors

MacRumors

macrumors bot
Original poster


Anthropic's Claude chatbot has been updated with support for inline visual content that will help it provide clearer answers.


Claude can now create custom visuals like charts, graphs, and diagrams. Visual content will be used when it better conveys an answer than plain text, and visual aids can also include real-world data like weather and recipes as long as web search is enabled. The visuals that Claude creates are distinct from Artifacts, and use HTML and SVG rather than image generation.

Claude is able to display current weather conditions and forecasts when users ask about the weather in specific locations, and it can provide formatted recipe cards that are easier to follow than a block of text. Weather and recipe data are only available on the desktop for now, because those visuals do not render in the iOS app.

Anthropic says that Claude is also able to ask structured questions using interactive multiple choice inputs instead of requiring users to type a response. Claude will use visuals when an answer calls for it, but users can also ask Claude to create a visual aid.

Visual responses and interactive content are available to all Claude users.

Article Link: Anthropic's Claude Can Now Create Interactive Visuals Directly in Conversations
 
Article Link
https://www.macrumors.com/2026/03/12/claude-visuals-conversations/
  • Haha
  • Like
Reactions: Konigi and Z-4195
Apple should've acquired Anthropic in their early days.

I use Claude everyday. Organising cumbersome file system tasks, creating to-dos in Things, quickly whipping up design documentations and of course vibe coding ideas and working prototypes. Amazing tool.
 
  • Like
  • Disagree
Reactions: haddy, Konigi, TheLunaPark and 11 others
Oblivious.Robot said:
The last time I talked with Claude, it kept arguing with me on what year this was 2025 or 2026.
I went back to ChatGPT.
Click to expand...
That's a pretty poor way to determine the strength of AI models. The data models are trained on has a fixed cutoff date... they aren't continuously updated. It's like complaining that a book can't tell you what the current date is.
Every time you use it it's like you are asking questions to a person who fell asleep in 2025 and just woke up. But it can use the internet just like you can if it needs to look up more up to date information. But it forgets that information the next time you chat.
 
  • Like
Reactions: MLVC, twistedpixel8, wilderkun and 1 other person
Jensend said:
That's a pretty poor way to determine the strength of AI models. The data models are trained on has a fixed cutoff date... they aren't continuously updated. It's like complaining that a book can't tell you what the current date is.
Every time you use it it's like you are asking questions to a person who fell asleep in 2025 and just woke up. But it can use the internet just like you can if it needs to look up more up to date information. But it forgets that information the next time you chat.
Click to expand...
I'd argue that its a perfectly reasonable way to determine if any given AI is having trouble with a simple task from a simple man with a simple question, Claude struggled and was gaslighting me into an argument of what year I quoted vs what the current year actually is. I'm sure it's a great app and lots of people love it, I had a poor experience at the very first onset and another app was just better without having to waste more of my energy I simply went to the one which worked. It's reasonable I think. 🙂
 
  • Like
Reactions: locovaca, dricci and Orange Bat
Whenever AI companies produce these marketing videos they should also show us a compilation of fails and errors otherwise it’s dishonest marketing and the result of that is people (like a few above) who think the sky is falling and humans are doomed.

If people experienced honest marketing and used these models they would know it’s just like any software - lots of bugs and disappointment.

Also, be aware that armies of bots exist on forums and socials whose task it is to make you feel you are worthless compared to AI. This is how they demoralize and enslave you by making you feel rejected third class citizens.
 
  • Haha
  • Like
Reactions: Orange Bat and Taustin Powers
Jensend said:
That's a pretty poor way to determine the strength of AI models. The data models are trained on has a fixed cutoff date... they aren't continuously updated.
Click to expand...

Yes and no. They can search online. Sometimes they will still argue against your lived experience until you sway them into admitting they are wrong.
 
  • Like
Reactions: Oblivious.Robot
coolfactor said:
Progress is mind-blowing. 🤯

The one question remains... when will AI become truly *curious*? Only then will it be truly sentient. Then, watch out...
Click to expand...

These are input-output systems. When there is no input from you they are no thinking, they are not living, they are just nothing.
 
  • Like
Reactions: Orange Bat
coolfactor said:
Progress is mind-blowing. 🤯

The one question remains... when will AI become truly *curious*? Only then will it be truly sentient. Then, watch out...
Click to expand...
If curiosity only appears after some threshold of neural complexity, then AI may never be sentient, because we most certainly are misunderstanding what curiosity actually is. And if life itself begins at a complexity threshold, that raises uncomfortable questions about the ideals we live by.
 
muzocan said:
Apple should've acquired Anthropic in their early days.

I use Claude everyday. Organising cumbersome file system tasks, creating to-dos in Things, quickly whipping up design documentations and of course vibe coding ideas and working prototypes. Amazing tool.
Click to expand...

The very fact that Apple didn’t acquire Anthropic tells us everything about how late the company is to the AI game. And I’m not convinced Apple would have led Anthropic to its current market‑leading position anyway, given how badly they’ve handled Siri since acquiring it back in 2010.

So in the end, consumers can probably be glad things turned out the way they did.
 
  • Like
  • Disagree
Reactions: Wx_Man and TwoBytes
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back