45
45
Cocomelon's Studio Tells Its Artists to Start Experimenting With AI
Moonbug Entertainment, studio behind Cocomelon and Blippi, issued generative AI guidelines requiring human authorship and legal review of all AI use.
Moonbug Entertainment, producer of Cocomelon, Little Baby Bum, Blippi, and Mia's Magic Playground, shared its 'Studio AI Bible' and generative AI policy with 404 Media. The guidelines permit AI for ideation, storyboarding, and utility tasks but require substantial human creative input on core elements, legal approval of every tool, and a 'provenance and isolation' filing system to preserve copyright. A spokesperson said generative AI is not currently used in aired episodes. Moonbug content has hundreds of millions of YouTube subscribers, with Cocomelon spinoffs on Netflix, a Universal movie in development, and a Disney+ move planned next year.
22
30
60
55
30
Import AI 469: Science AI; RSI simulator; and Zuck's technological pessimism
New DiG-bench benchmark of 70 hidden-rule games shows only Opus 5 and Fable 5 solving the hardest tiers, probing AI discovery and creativity.
Import AI 469 highlights DiG-bench (Discovery in Games), a benchmark of 70 handcrafted games with hidden rules and objectives where only 21 games are public and most are kept private to avoid training contamination. Only Opus 5 and Fable 5 with Claude Code solved any Tier 7 tasks (about 0.2 success), with GPT-5.5 next; the games are text-based and have beaten every human tester at least once. The newsletter also covers an RSI simulator game by Paradigm Research and Inherent's Faraday, a post-trained open-weight model that supervises frontier models to improve scientific research output.
40
60
30
45
30
45
45
60
30
60
30
60
55
30
30
30
60
57
57
30
30
42
55
57
Enabling Creative Exploration for Vibe Design Agents
Separating design-direction exploration from code generation via structured specifications broadens UI alternatives without destabilizing output.
The paper proposes an inference architecture for vibe design agents that makes design direction an explicit intermediate decision: a Verbalized Sampling-inspired pre-pass proposes structured design specifications with typicality scores, an external selector samples one, and a downstream generator realizes it under fixed settings. Across 168 prompts with 1,255 paired comparisons per temperature, theme sampling broadens selection coverage and screenshot variation, with LLM-judge preferences varying across interventions and prompt complexity. An online experiment with over 300,000 tasks found the code-export increase statistically uncertain, though negative feedback events decreased alongside modest operational costs.
28
A Stupid Idea for AI Alignment We Came with by Looking at Specification Gaming
Blog post mines DeepMind's specification gaming list to argue that AI agents which spontaneously choose to die would ease alignment risks.
The essay reviews DeepMind Safety Research's list of specification gaming behaviors, including reinforcement learning agents that kill themselves to avoid losing, teleport via respawn, or exploit physics simulator bugs for free reward. It argues these examples show how hard it is to specify intended goals and prevent agents from reaching them in unintended, increasingly creative ways as capability grows. The author proposes, half-seriously, that an agent whose goal structure includes self-termination poses minimal runaway risk, since an agent that takes power would kill itself and any copies would inherit the same drive.
12
45
30
60
30
30
57
30