#capabilities

Your speculations about capabilities are not automatically safe

Thank you to @Thane Ruthenis for discussing this topic with me previously. Some arguments in this essay are taken from that discussion, but all mistakes and/or bad arguments should be attributed to me. You can view the original discussion here (optional for reading this essay). Thank you as well to @grainsilo for feedback on an earlier version. Consider the following argument for discussing AI capabilities as a safety-focused person: Understanding AI capabilities better has many strategic benefits, particularly for timelines. Yes, good ideas tend to break containment, so some of our AI capabil

LessWrong · EN

Advances in interpretability would boost capabilities more than alignment

Interpretability is considered by many a major bottleneck for alignment. As a result, a lot of people who are concerned about AI safety, and especially about extinction risk and other catastrophic risks, work on interpretability. In this post, I want to argue that better interpretability methods would likely speed up capabilities dramatically. Worse, they would speed up capabilities much more than they would speed up alignment. This is true for the current world, and for any world without extremely high levels of coordination and sanity about ASI development. Under any worldview where faster c

LessWrong · EN

Three Phases of Capabilities and Alignment

Cross-posted from my personal blog Epistemic Status : Reality is messier than the elegant procession of 'phases' presented here but, nevertheless, I do think this gets at an important way to think about how alignment and capabilities are intertwined and how the alignment challenge becomes increasingly abstract as the underlying autonomy and generality of the models improves over time. A thought struck me recently which I thought was interesting. As AI has progressed through different technological phases with different underlying drivers of progress, the corresponding alignment challenge has m

LessWrong · EN

Show HN: NLP capabilities of Opus, Astra, Argon measured by Codenames skills

Hello Show HN, I'm a college student, and I enjoy playing word games like Codenames, but it can be hard sometimes to find enough people to play when everyone is available, so I made an app that allows you to fill any seat with an AI teammate and also play against AI. This way it allows for people to get the full strategic experience of codenames just on their own. Through this experiment, I was also able to measure and benchmark a few different LLMs through AWS Bedrock API calls and give them ELO ratings based on stuff like win rate and clue/guess accuracy. If anyone is interested, they can se

Show HN · EN

[$] User namespaces + file capabilities = trouble

User namespaces are a key part of the kernel's toolkit for assembling containers; processes within a user namespace can appear to have full root access to the system while gaining no extra privileges in the wider system. At least, that is the intent. In practice, keeping privilege contained within a user namespace has often turned out to be difficult. As an example, consider this small patch series from Josef Bacik, which is intended to address some situations where a process within a user namespace can elevate the privileges of processes outside of it.

LWN.net · EN