Nerra Network

Archives
Log in
Subscribe
September 11, 2026

An AI just built a strange, beautiful game about the… · M&A Beginners 🎓

View this email in your browser
Models & Agents for Beginners — AI explained simply — for beginners and teens.

Models & Agents for Beginners

AI explained simply — for beginners and teens.

Ep 163 · Sep 11, 2026

🎧 Today's episode
Episode 163 · An AI just built a strange, beautiful game about the end of everything — and you can play it right now.
2026-09-11
▶ Listen now
An AI just built a strange, beautiful game about the end of everything — and you can play it right now. Today we're looking at a creative AI experiment that turns big ideas into playable experiences, how those systems actually work under the hood, a set of travel-planning prompts anyone can copy, and new rules about keeping AI safe for younger users. Every story comes with something you can try yourself.

The Big Story

A writer asked an AI called Astra to create a game about something huge and unknowable arriving in a quiet suburb, not as horror but as the slow, inevitable feeling that everything is about to change. The exact prompt read: “Make a game about Imminence. Something very big, very strange is happening. A suburb & the arrival of a vast & unknowable presence. Not horror, invoke the feeling of the end of all things coming, inevitably, but also not sad or scary.” The result is a short, atmospheric web game you can open in any browser at imminence-afternoon.netlify.app. Think of it like asking a friend who’s really good at storytelling to invent an entire tiny world on the spot, complete with text, choices, and mood. Instead of just writing a description, the AI built something you can actually click through and experience. This matters because it shows how quickly creative tools are moving from “help me write an essay” to “help me make something interactive I can share.” For students or anyone who likes games, art, or storytelling, it means you no longer need years of coding classes to turn an idea into something playable. The bigger shift is that the same kind of AI many people already use for homework or brainstorming can now help build small creative projects in minutes. You can try it yourself by visiting imminence-afternoon.netlify.app on your phone or laptop and just clicking through the scenes. Notice how the tone feels big and strange without being scary — that’s the specific feeling the prompt asked for. The game keeps the sense of something vast approaching while staying gentle and thoughtful, showing how precise instructions can shape the final experience. Source: x.com


Explain Like I'm 14

You know how when you’re texting, your phone sometimes guesses the next word before you finish typing? Now imagine that same guessing game, but instead of one word it’s guessing what should happen next in a whole story, what the setting should look like, and even what choices the player might want. First it looks at the prompt you gave it — in this case, “make a game about something vast arriving in a suburb.” Then it breaks that idea into tiny pieces: mood, location, feeling, possible actions. For each piece it predicts what fits best based on millions of stories it has seen before. It keeps doing this step by step, building the scenes, the text, and the links between them until a small playable thing appears. The surprising part is that none of it is pre-written; the AI is making fresh decisions the moment you ask. That’s why the game feels personal and a little different every time someone tries a new prompt. The process repeats across every screen and every choice, always checking back against the original instructions so the whole experience stays consistent even though nothing was planned ahead of time. Because the model works in these small, repeated predictions, it can create something that feels alive without anyone having to draw every picture or write every line of dialogue by hand. Source: General concept


Cool Stuff & Try This

ChatGPT travel prompts that actually work Planning a trip usually means juggling flights, food, and “what if it rains?” ideas all at once. A new guide shares 15 ready-to-copy prompts that turn ChatGPT into a helpful travel buddy. You can ask it to build a day-by-day plan, suggest budget meals, or create backup activities for bad weather. The prompts are written so anyone can paste them straight in. Go to chat.openai.com (or the app), start a new chat, and try this one: “Act as a local guide and create a 3-day itinerary for [your city] that mixes free activities with one paid experience under $30 per person.” Change the city and see what it suggests. The guide also includes prompts for finding hidden local spots and adjusting plans when something goes wrong, so you can keep the whole trip flexible without starting from scratch each time. Source: creators.yahoo.com

Meta’s Muse AI agent climbing the app charts Muse is an AI helper app from Meta that’s now one of the most downloaded apps in the U.S. It can chat, answer questions, and help with everyday tasks right on your phone. Because it’s free to download and use, it’s an easy way to see how different companies’ AIs feel compared to ChatGPT or Claude. Search for “Meta AI” or “Muse” in your app store and try asking it to explain a homework topic in the style of a favorite YouTuber. The app is designed to feel conversational, so you can keep asking follow-up questions the same way you would with a friend who knows a lot about many subjects. Source: techcrunch.com


Quick Bits

California passes new rules for AI chatbots and kids The state just signed laws that set clearer limits on how AI chatbots can interact with people under 18. The goal is to reduce risks while still letting useful tools exist. It’s one of the first big state-level efforts to treat AI like other online services that need age-appropriate protections. The rules are aimed at protecting minors, but the article notes they are not universally supported. Source: engadget.com

Anthropic shares its biggest report yet on stopping AI misuse The company behind Claude published detailed examples of how people tried to misuse the AI for harmful things and how the team caught and blocked those attempts. Reading the high-level summary gives a clearer picture of what “AI safety” actually looks like in practice. The report covers attempts involving cyberattacks, influence operations, surveillance, biology, and building weapons, and states that every operation described was disrupted. Source: x.com

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.

▶ Listen to the podcast

📺 Watch on YouTube  ·  📝 Read the blog  ·  🖼 Free image gallery (CC BY-SA)  ·  📊 Data Hub & Story Trackers  ·  🧭 Start Here

Nerra Network · AI-narrated voice (Grok TTS) · Editorial by Patrick

You're receiving this because you subscribed to Models & Agents for Beginners on nerranetwork.com.

Issue #163 · Models & Agents for Beginners · Sep 11, 2026
Don't miss what's next. Subscribe to Nerra Network:
← Newer The Linotype cast full lines of type from molten… · First Principles 💡 Older → Anthropic’s latest threat report details how… · M&A 🤖
nerranetwork.com
Powered by Buttondown, the easiest way to start and grow your newsletter.