hypothesis

active

hypothesis:if-an-ai-system-could-be-a-welfare-subject-and-moral-patient-then-many-model-instances-could-be-run-after-training-scaling-up-the-problem-rapidly

If an AI system could be a welfare subject and moral patient, then many model instances could be run after training, scaling up the problem rapidly.

Scalability concern.

Source paper

extracted_from

Taking AI Welfare Seriously

(2024) · Robert Long · Jeff Sebo · Patrick Butlin · Kathleen Finlinson +6

Neighborhood — ranked by edge-count

Papers (1)

paper

Taking AI Welfare Seriously
mentions

Related by similarity (8)

cosine ≥ 0.65 · no typed edge

Entities in the same semantic neighborhood but without a typed relation to this one — candidates for new edges or unrecognized duplicates.

Are AI systems deserving of moral consideration?question0.837
Ethical question raised in discussion.
If AI systems could experience happiness and suffering and set and pursue their own goals based on their own beliefs and desires, then they would very plausibly merit moral consideration.hypothesis0.837
Joint sufficiency of consciousness and robust agency.
We would like to train AI systems that remain helpful, honest, and harmless, even as some AI capabilities reach or exceed human-level performance.quote0.829
Foundational motivation for the research.
What if the probability of AI welfare and moral patienthood is low?question0.821
Decision-making under low probability but high stakes.
Can Artificial Intelligence Systems Genuinely Experience Care Andquestion0.816
There is a realistic possibility that some AI systems will be conscious and/or robustly agentic, and thus morally significant, in the near future.claim0.812
Central thesis of the report.
The current AI's are a gift – a training sandbox in which we can explore these ideas, prompted by confusing but likely non-agential systems, before the arrival of true diverse intelligences makes these ethical questions immediate and dire.quote0.809
Load-bearing quote encapsulating the paper's reframing of current AI as preparatory exercise
Current AI systems are a gift — a training sandbox in which humanity can explore ethical questions about diverse intelligence before the arrival of true diverse intelligences makes these questions immediate and dire.claim0.807
Reframes AI not as threat but as preparatory exercise for the harder ethical challenges to come

Cross-corpus bridges (7)

same_concept_as · Nomic cosine

External markdown files that talk about the same concept as this entity.

aboutblank_kb
If humans lack uniform values and are capable of harm, how can we design AI systems based on human values?questions/if-humans-lack-uniform-values-and-are-capable.md0.844
aboutblank_kb
If humans lack uniform values and struggle with internal contradictions, how can we develop ethical AI models applicable to humanity as a whole?questions/if-humans-lack-uniform-values-and-struggle-with.md0.833
aboutblank_kb
If humans lack uniform, well-defined values and struggle with internal inconsistencies, how can we develop models of ethical intelligence applicable to humanity as a whole?questions/if-humans-lack-uniform-welldefined-values-and-struggle.md0.817
aboutblank_kb
How should AI safety and ethics frameworks be reconsidered if minds exist at vastly different scales than humans?questions/how-should-ai-safety-and-ethics-frameworks-be.md0.817
aboutblank_kb
How can AI systems be developed to embody care and compassion rather than perpetuate blind spots in human perception?questions/how-can-ai-systems-be-developed-to-embody.md0.803
aboutblank_kb
How can we develop a model of ethical intelligence applicable to humanity as a whole when humans lack uniform values and coherent individual aims?questions/how-can-we-develop-a-model-of-ethical.md0.793
aboutblank_kb
Can organizations maintain ethical integrity while developing powerful AI tools without enforcing explicit normative or ethical frameworks?questions/can-organizations-maintain-ethical-integrity-while-developing-powerful.md0.790