EA Forum Podcast (Curated & popular)1 Apr 2025

“Mitigating Risks from Rouge AI” by poppinfresh

Introduction

Misaligned AI systems, which have a tendency to use their capabilities in ways that conflict with the intentions of both developers and users, could cause significant societal harm. Identifying them is seen as increasingly important to inform development and deployment decisions and design mitigation measures. There are concerns, however, that this will prove challenging. For example, misaligned AIs may only reveal harmful behaviors in rare circumstances, or perceive detection attempts as threatening and deploy countermeasures – including deception and sandbagging – to evade them.

For these reasons, developing a range of efforts to detect misaligned behavior, including power-seeking, deception, and sandbagging, among other capabilities, have been proposed. One important indicator, though, has been hiding in plain sight for years. In this post, we identify an underappreciated method that may be both necessary and sufficient to identify misaligned AIs: whether or not they've turned red, i.e. gone rouge.

In [...]

---

Outline:

(00:10) Introduction

(01:42) Historical Evidence for Rouge AI

(02:36) Recent Empirical Work

(03:46) Potential Countermeasure

(03:49) The EYES Eval

(04:55) EYES Eval Demonstration

(05:44) Future Research Directions

(06:46) Conclusion

---

First published:
April 1st, 2025

Source:
https://forum.effectivealtruism.org/posts/uKKoj9iqj2cWKsjrt/mitigating-risks-from-rouge-ai

---

Narrated by TYPE III AUDIO.

---

Images from the article:

Apple Podcasts and Spotify do not show images in the episode description. Try Pocket Casts, or another podcast app.

Det här avsnittet är hämtat från ett öppet RSS-flöde och publiceras inte av Podme. Det kan innehålla reklam.

Avsnitt(250)

“Let’s taboo the V-word” by lincolnq

“How long have you been v*g*n?” This is one of the most common icebreakers at animal protection events. It's a baseline assumption, and it mostly holds true: if you’re out advocating for animals not t...

14 Juli 12min

“Giving What We Can’s first YouTube Video is out now!” by JustinPortela

Hello! I'm Justin Portela. I got hired by GWWC to make YouTube videos after AI in Context did such a kickass job. My channel is using that same cinematic, high-production value beauty to talk about e...

9 Juli 1min

“I’m never satisfied” by Ajeya

Note: This post was crossposted from Planned Obsolescence by the Forum team, with the author's permission. The author may not see or respond to comments on this post. But we get the job done I was twe...

8 Juli 6min

“Maybe do the thing you wish CEA would do” by alejoacelas 🔸

I used AI to fix transcription errors, rerrarange the ideas, and suggest tweaks to the title and some sentences. Three of the most exciting projects to come out of EA in recent years are, in a vague s...

8 Juli 4min

“Mabye do the thing you wish CEA would do” by AlejoAcelas🔸

8 Juli 4min

“Possible mistake EAs are making and shout out to Pause AI UK” by Michelle_Hutchinson

I think right now EAs might be making a significant mistake by paying insufficient attention to the political realm. As EAs we tend to figure out what's most impactful for us to work on and focus hard...

29 Juni 6min

“Coming Around To Political Donations” by Jeff Kaufman 🔸

Five years ago I read a post on the EA Forum arguing that "election campaign contributions might be a way in which you can have a substantial impact as a small donor". It struck me as weird but plausi...

12 Juni 4min

“animal welfare has an evidence problem” by matthes

Why I stopped donating to animal welfare charities but feel more motivated than ever to redirect money and talent to the cause. I have wanted to write this post for a while. It is an uncomfortable thi...

6 Juni 26min