game – Open Thinkering

RetroEmoji Challenge: a simple game created using ChatGPT

Posted on March 8, 2023March 9, 2023 by Doug Belshaw

This morning, I came across Sumplete, a game invented and coded entirely by AI (ChatGPT). This sounded awesome so between meetings today, I had a go at something which, while not entirely original, is still something I wouldn’t have been able to create myself.

I present to you RetroEmoji Challenge!

It’s not perfect, or even that fun, but given that I literally just typed instructions into a box and then copy/pasted the output into CodePen, I’m very happy. It was fun.

How I made it

The Sumplete about page does a good job of explaining how to talk to ChatGPT. The only things I’d add are:

You’re limited on the free version of ChatGPT to a certain number of lines of output. You can get round this by asking it to give you the HTML, CSS, and JS separately.
ChatGPT can be a bit flakey and sometimes forgets what it’s talking about, crashes, or loses the chat history. If this happens, just tell it you want to share the code for a game and for it to tell you when it’s ready to receive it. Then paste in the code.
Sometimes the code ChatGPT gives you doesn’t work, or doesn’t work as intended. You can tell it this, and sometimes it fixes it!

I asked for a ‘retro 1980s vibe’ for RetroEmoji Challenge. And yes, ChatGPT came up with the name for it.

There’s a lot of trial and error, but there’s a lot of that with any kind of project, and I literally couldn’t have made this by myself. Copying-and-pasting the code into CodePen helps see the changes you make in realtime. It’s kind of addictive.

Informal learning, gaming, and #openbadges design

Posted on July 19, 2012 by Doug Belshaw

Burnout Paradise

One of my favourite games for the PlayStation 3 is Burnout Paradise. Apart from the racing and being able to take down cars in spectacular ways, one of the reasons I love it is because it’s a non-linear game.

What do I mean by that?

I mean that after a (very) lightweight introduction, the whole map is open to the player. You’re guided through the mechanics of the game as you play it, and you can choose what you want to do next. If you just want to drive around, that’s fine. In fact, there’s ‘challenges’ to complete (smashing through billboards, etc.) if that’s all you want to do. By driving around you actually discover some of the ‘formal’ challenges like races as well as the auto repair shops.

Burnout Paradise - map

Every now and again, either through winning races or completing stunt challenges you’ll unlock a new car. But you still have to go and find it and take it down. And there’s also the ‘stealth’ achievements you unlock unexpectedly. It’s a compelling, very rewarding game in its own right, never mind being able to play live online against other human opponents!

Recently, within the Mozilla Learning team we’ve been discussing the non-linearity of badge systems and how interest-based learning can be scaffolded and assessed. Obviously the assessment is ultimately going to lead to Open Badges, but a few of us feel that we can’t merely replicate the existing structures found in formal education. There’s not much point in using badges if the learning design still talks about a ‘101’ class or uses a Beginner/Intermediate/Advanced approach.

The question has come up, as it always does, about pre-requisites. There’s no getting away that some learning is built upon prior knowledge, the argument goes. That’s certainly true, but there’s ways of motivating the learner to want to undertake that prior learning. That way is by appealing to their interests.

As with anything new, the easiest way to get at what we can do is through metaphor. In this case, I think that a video game serve as a very useful model for what we want to do. Start with the player (in this case the learner) and scaffold experiences around them.

Does that make sense?