# AI News Digest - 2026-04-13

1. OpenAI CEO Sam Altman's San Francisco home was attacked with a firebomb by a suspect who had posted on the PauseAI Discord server and written about AI driving humanity to extinction.

2. Arcee AI spent roughly half its venture capital to train Trinity-Large-Thinking, a 400-billion-parameter open reasoning model designed to rival Claude Opus in agent tasks.

3. Researchers tested 34,000 real-world agent skills and found that skills that performed well in benchmarks provided negligible benefit under realistic conditions and in some cases degraded performance for weaker models.

4. An international research team released OpenWorldLib and proposed a formal definition of "world model," explicitly excluding text-to-video generators such as Sora from that definition.

5. An OpenAI employee tried to explain usage limits of the newly introduced $100 ChatGPT Pro plan after confusing pricing labels on the service's pricing page had left users uncertain.

# References

1. [https://the-decoder.com/man-who-firebombed-sam-altmans-home-was-likely-driven-by-ai-extinction-fears/](https://the-decoder.com/man-who-firebombed-sam-altmans-home-was-likely-driven-by-ai-extinction-fears/)

[Man who firebombed Sam Altman's home was likely driven by AI extinction fears](https://the-decoder.com/man-who-firebombed-sam-altmans-home-was-likely-driven-by-ai-extinction-fears/)

1. [https://the-decoder.com/arcee-ai-spent-half-its-venture-capital-to-build-an-open-reasoning-model-that-rivals-claude-opus-in-agent-tasks/](https://the-decoder.com/arcee-ai-spent-half-its-venture-capital-to-build-an-open-reasoning-model-that-rivals-claude-opus-in-agent-tasks/)

[Arcee AI spent half its venture capital to build an open reasoning model that rivals Claude Opus in agent tasks](https://the-decoder.com/arcee-ai-spent-half-its-venture-capital-to-build-an-open-reasoning-model-that-rivals-claude-opus-in-agent-tasks/)

1. [https://the-decoder.com/agent-skills-look-great-in-benchmarks-but-fall-apart-under-realistic-conditions-researchers-find/](https://the-decoder.com/agent-skills-look-great-in-benchmarks-but-fall-apart-under-realistic-conditions-researchers-find/)

[Agent skills look great in benchmarks but fall apart under realistic conditions, researchers find](https://the-decoder.com/agent-skills-look-great-in-benchmarks-but-fall-apart-under-realistic-conditions-researchers-find/)

1. [https://the-decoder.com/researchers-define-what-counts-as-a-world-model-and-text-to-video-generators-do-not/](https://the-decoder.com/researchers-define-what-counts-as-a-world-model-and-text-to-video-generators-do-not/)

[Researchers define what counts as a world model and text-to-video generators do not](https://the-decoder.com/researchers-define-what-counts-as-a-world-model-and-text-to-video-generators-do-not/)

1. [https://the-decoder.com/openai-employee-tries-to-explain-usage-limits-of-the-new-chatgpt-pro-plans/](https://the-decoder.com/openai-employee-tries-to-explain-usage-limits-of-the-new-chatgpt-pro-plans/)

[OpenAI employee tries to explain usage limits of the new ChatGPT Pro plans](https://the-decoder.com/openai-employee-tries-to-explain-usage-limits-of-the-new-chatgpt-pro-plans/)

For the site tree, see the [root Markdown](https://slashpage.com/ixtj-dev.md).