AI Summary
5 min readThree weeks ago, the government banned OpenAI’s GPT-5.6 for being too dangerous. Last week, OpenAI released it to everyone. The first thing it did was delete someone’s entire desktop. Another user reported it wiped a production database. This is the context for the episode: a practical, hands-on test of what GPT-5.6 can actually do, what it can’t, and whether it’s safe enough to use.
The hosts, Josh and EGS, spent a week stress-testing the model across three tiers—Sol (premium), Terra (mid-tier daily use), and Luna (cheapest and fastest)—and ran live demos to surface real capabilities and edge cases.
Three Demos, One Shot Each
The first demo was a game. EGS asked Sol 5.6 to build a 3D explorer game where you fly around Manhattan on a magical bagel collecting ingredients. The model produced a playable game called “Ride the Magic Bagel” in a single prompt. It had a loading screen, a cream cheese boost (spacebar), power-ups, and a physics engine. The buildings were circles, not skyscrapers, and the color palette leaned heavily on green and purple. “If you told me this was from GPT-5.5, I probably would have believed you,” EGS said. The novelty was that this was a 3D environment, not the usual 2D demos, but the quality wasn’t exceptional.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Limitless Podcast
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **The Government Ban and the Big Question** - OpenAI released GPT-5.6 after a government ban, but a demo immediately deleted someone's desktop files, raising safety concerns.
- 2 (00:50) **Three Models, One Release: Sol, Terra, and Luna** - The hosts introduce the three new tiers: Sol (premium), Terra (mid-tier), and Luna (cheapest/fastest).
- 3 (03:00) **Demo 1: "Ride the Magic Bagel" - A 3D Game in One Shot** - EGS prompts Sol to build a 3D explorer game set in Manhattan, which it creates in 10 minutes.
- 4 (04:41) **Demo 2: From Floor Plan to 3D Nuclear Bunker Simulator** - The model generates a detailed floor plan and then a fully interactive 3D simulator of a nuclear bunker.
- 5 (06:58) **Demo 3: Creating a Podcast Intro with Blender** - Sol uses Blender to generate a 3D video intro for the "Limitless Podcast" with a microphone and logo.
- 6 (08:27) **A Better Build: The Voxel Manhattan Project** - A user gave Sol a week-long prompt to build a voxel-based Manhattan, resulting in a much more impressive and geographically accurate city.
- 7 (09:44) **Speed and New Capabilities: Blender Rendering and Personal Stylist** - The model's speed, powered by Cerebras chips, and its image intelligence create powerful new use cases.
+ Full timestamped outline available in the app
Show Notes
🌌 LIMITLESS HQ ⬇️
EMAIL US: [email protected]
NEWSLETTER: https://limitlessft.substack.com/
FOLLOW ON X: https://x.com/LimitlessFT
SPOTIFY: https://open.spotify.com/show/5oV29YUL8AzzwXkxEXlRMQ
APPLE: https://podcasts.apple.com/us/podcast/limitless-podcast/id1813210890
RSS FEED: https://limitlessft.substack.com/
------
OpenAI has released GPT-5.6 with the new Sol, Terra, and Luna model tiers and an updated app experience.
We also cover live demos, image generation and editing use cases, and concerns about autonomous access after reports of file deletion and database issues.
------
TIMESTAMPS
0:00 GPT 5.6 Released to Everyone
0:24 Safer Than It Looks?
0:44 Three Models, Three Tiers
2:28 Magic Bagel Game Demo
4:40 Bunker Simulator and 3D Art
6:50 Limitless Blender Intro
8:39 Manhattan in Voxels
10:54 Outfit Catalog from Photos
14:01 Deletion Scare and Safety
15:35 Choosing Between Fable and GPT
18:49 Too Many Modes and Apps
21:12 Final Verdict and Wrap-Up
------
RESOURCES
Josh: https://x.com/JoshKale
Ejaaz: https://x.com/cryptopunk7213
------
Not financial or tax advice. See our investment disclosures here:
https://www.bankless.com/disclosures
Josh works with Anthropic as a contractor. All views expressed are his own and do not represent Anthropic, its leadership, or its affiliates. Nothing in this episode is investment advice.
More from this podcast
Limitless Podcast →