正在加载视频...

视频加载失败

"Hey Claude, create an interactive simulator that explains the Monty Hall problem." "Make it better." (It soon hit the maximum size for an artifact) Nice job, I thought the ability to play the game manually or autoplay with different strategies was clever

40,307 次观看 • 1 年前 •via X (Twitter)

10 条评论

Ethan Mollick 的头像
Ethan Mollick1 年前

It did make one error I needed to correct (co-intelligence!) with the cumulative wins graph, but otherwise this was all Claude.

Stephen Hill 的头像
Stephen Hill1 年前

I've been doing a few "Teaching with Generative AI" presentations lately and Claude Artifacts are always a crowd favorite. It's amazing how easy they are to build. You're definitely right about the need for "cointelligence" though.

Bailey Berro 的头像
Bailey Berro1 年前

now we need one for going for 2 when down 14 in NFL games

Simon P 的头像
Simon P1 年前

Love seeing Claude make learning fun! The Monty Hall problem is a classic for a reason – it's counterintuitive. The artifact's ability to let you play manually or autoplay is a great way to see the odds in action. #Claude #AI #MontyHall #Learning

Juan Stoppa 的头像
Juan Stoppa1 年前

what prompt did you use? I can see that it generated statistics graphs

Norbert Landsteiner 的头像
Norbert Landsteiner1 年前

And here's how a human does it: (This was meant for an article generalizing on the problem, emphasizing the temporal component and how this is a more common exploit. But I was disrupted and the text never emerged.)

Colonel Tasty🤌 的头像
Colonel Tasty🤌1 年前

This is incredible!

Ash 的头像
Ash1 年前

Already want to give Claude more resources for better artifacts.

ultimate0 的头像
ultimate01 年前

Been trying the same thing like 3 times over the last almost 2 years or so. Didnt turn out that nice. one prompt? ( besides the correction?)

Medical Axioms 的头像
Medical Axioms1 年前

This is easier! Switch wins 66%

相关视频