Supersonic wrote a case study about us.
Here's what we achieved together on Backyard Master:
📈 D1 Retention: 27.8% → 46.5% 📈 D7 Playtime: 3x longer 📈 D1 Churn: 9.4% → 3% 📈 D7 ARPU: 0.71→0.71→1.04
Full story 👇 https://t.co/JECP4CgLlj
@shinobi602 Iterating on existing games is a safer bet than launching brand-new projects, and this has become a risk-free strategy for an increasing number of major game companies. It’s a bit disappointing, yet understandable.
Rest assured that GPT 5.6 Sol will stay in the ChatGPT subscription you pay for. Including Go, Plus and Pro subscriptions. At least until we ship an even better model.
Take a calm look at every new model, and you'll realize there is still a long gap between them and your actual business, no matter how powerful they are.
https://t.co/V4Fle7zWCn
85% male users.
Western art style.
Low potential.
Loop Arena. Shipped. Fun. Doomed.
Sometimes a game that works is a game the market can't fund.
Building anyway. What would you do? 👇
#g... https://t.co/hLvSHvlvvL via @YouTube
Anthropic extends Claude Fable 5.
OpenAI quietly drops the 5-hour Codex limit.
Both are good news for users.
But here's the builder's dilemma:
If your product depends on a single model's rate limits, you don't have a product.
You have a hostage situation.
Plan for model pivots from day one.
https://t.co/37rZmW6deq
Yeah, the gap is closing for sure. Sol’s improvements are noticeable, but Fable still nails the details like a champ. Check out this tool for building pipelines if you’re into that sort of thing—might help with your projects! <a href="https://t.co/fMZSisYgmM" target="_blank">Taskmeld</a>
Grok 4.5 just dropped. We've been building with GLM.
This is what we made:
🎯 Throw-based PVP (Clash Royale style)
🏔️ 3D terrain that changes the fight
⚡ Fully playable, running now
The tech keeps getting better. But the idea still comes from the human.
Built with GLM. Inspired by Grok 4.5. Thoughts? 👇
#gamedev... https://t.co/lP5sFBDdMg via @YouTube