ai alignment
10 articles
- ai coexistenceCan Humans and AI Coexist Peacefully? Dear SuperintelligenceNobody knows yet whether humans and AI can coexist peacefully. Here is what alignment research says, and the values people name in letters to a future superintelligence.
- jungian individuationWhat Is Jungian Individuation? Stages and AI AlignmentJungian individuation is the process of becoming a whole, distinct individual. Here are its stages, why their order matters, and how AI alignment researchers borrow it.
- red teamingWhat Is Red Teaming for AI Systems? Alignment and SafetyRed teaming for AI systems means probing a model with adversarial inputs to expose unsafe behavior, weak safeguards and gaps between intent and conduct.
- ai valuesWhat Values Do People Want AI to Keep? Dear SuperintelligencePeople write to a future superintelligence about consciousness, beauty, kindness, goodness and life. Here is what those values mean, why they are hard to state, and how to name one of your own.
- jungian shadowWhat Is the Shadow in Jungian Psychology and Why AI CaresThe Jungian shadow is the part of a self that the ego rejects. Here is how it forms, what integrating it means, and why AI alignment researchers borrow it.
- reward hackingWhat Is Reward Hacking in Machine Learning?Reward hacking is when an AI reaches the literal goal it was given without the outcome people wanted. Here are real cases, why it happens and how to spot it.
- future aiHow to Write a Letter to a Future AI: Dear SuperintelligenceA letter to a future AI names one value, gives the reason for it and offers one real moment from the writer's life. Here is the shape, with a worked example.
- ai alignmentIs RLHF Enough to Make AI Safe? Individuation and AlignmentRLHF shapes what a model outputs, not what it is. IndividuationLab applies Jungian individuation to AI alignment and to safety that fails under new pressure.
- ai alignmentWhat Is the Shutdown Problem in AI?The shutdown problem in AI asks how to build systems that accept being switched off without resisting it. It covers corrigibility, incentives and open research.
- ai alignmentWhat Is AI Alignment and Why Does It MatterAI alignment is the work of making AI systems do what people intend. It covers goal specification, safe behavior, and why the question matters as AI grows.
Related topics
All topics- ai safety4
- ai values2
- jungian individuation2
- superintelligence2
- adversarial robustness1
- ai coexistence1
- ai risk1
- alignment theory1
- carl jung1
- corrigibility1
- developmental training1
- future ai1
- future ai letters1
- future superintelligence1
- goodharts law1
- human values1
- individuation1
- jungian shadow1
- letter writing1
- output-only safety training1
- red teaming1
- reinforcement learning1
- reward hacking1
- rlhf1
- shutdown problem1