In partnership with

Welcome to Jumble, your go-to source for AI news updates. This week, Google turns typed ideas into playable games. Meanwhile, three fired OpenAI safety researchers are pushing back against their dismissals. Let’s dive in ⬇️

In today’s newsletter:
🕹️ Google puts game creation in a chat box
🔎 Former safety staff push for clearer rules
📷 Flock faces a reported staffing shakeup
🧬 A bigger biological data effort takes shape
🎯 Weekly Challenge: Build something worth playing twice

🧩 Google Turns Your Game Idea Into Something Playable

Google’s Playground launch lets you describe a game and try the result in your browser. Your wonderfully silly idea finally has somewhere to go.

Is this the beginning of the end for the gaming industry?

Login or Subscribe to participate

📝 You Write the Rules

Start from scratch or a suggested prompt, then request changes to characters, rules or physics. Games work on phones and laptops, with private and shared options. Unity Spark is also due later this year, bringing more advanced 3D game-making tools to Playground.

🇺🇸 US Adults Get the First Turn

The initial rollout covers US adults aged 18 and over, with creation access tied to Google AI subscriptions. Check your allowance before planning an afternoon of experiments.

Discover laptops with built-in intelligence.

Shift your workforce into hyper-productivity mode with Dell Pro laptops powered by Intel® Core™ Ultra with Intel vPro®. Shop on Dell Premier.

⚖️ Fired OpenAI Researchers Push Back

Three fired researchers are challenging OpenAI’s account of their dismissals. Their argument reaches beyond their own jobs: can colleagues raise concerns without worrying about what happens next?

🗣️ Speaking Up Is Part of the Job

Tomek Korbak, Jasmine Wang and Mikita Balesni say outside collaboration was part of their work. They argue unclear boundaries now risk discouraging the conversations needed to identify safety problems.

📋 Two Very Different Accounts

OpenAI says the dismissals concerned misconduct, not retaliation. The researchers dispute wrongdoing; their letter is their account, and it does not independently settle what happened inside the company.

Weekly Scoop 🍦

🕹️ Weekly Challenge: Create Your Playable Game

❝

Challenge: Make a tiny game with one goal and one way to lose. Use ChatGPT, Gemini or Claude to help shape the idea; keep the first version simple.

Here's what to do:

Step 1: Pick your tiny game Start with one goal and one way to lose. A toast-catching cat is plenty; save the sprawling fantasy universe for later.

Step 2: Write the brief Ask your assistant: “Design a game where a cat catches falling toast. Give me simple controls, a score and one way to lose.”

Step 3: Make it playable If eligible, describe your game to Playground’s conversational builder. Without access, try the rules using paper tokens first.

Step 4: Fix one thing Play a round, then ask for one change: slower toast, clearer scoring or a tougher finish. Try again before adding more.

What would you build if making a game started with a sentence? And, can an AI lab keep its safety work open while protecting confidential information? See you next time!

Stay informed, stay curious, and stay ahead with Jumble!

Zoe from Jumble