The Generative AI Paradox: "What It Can Create, It May Not Understand"

• URL: http://arxiv.org/abs/2311.00059v1
• Date: Tue, 31 Oct 2023 18:07:07 GMT
• Title: The Generative AI Paradox: "What It Can Create, It May Not Understand"
• Authors: Peter West, Ximing Lu, Nouha Dziri, Faeze Brahman, Linjie Li, Jena D. Hwang, Liwei Jiang, Jillian Fisher, Abhilasha Ravichander, Khyathi Chandu, Benjamin Newman, Pang Wei Koh, Allyson Ettinger, Yejin Choi
• Abstract summary: Recent wave of generative AI has sparked excitement and concern over potentially superhuman levels of artificial intelligence. At the same time, models still show basic errors in understanding that would not be expected even in non-expert humans. This presents us with an apparent paradox: how do we reconcile seemingly superhuman capabilities with the persistence of errors that few humans would make?
• Score: 81.89252713236746
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The recent wave of generative AI has sparked unprecedented global attention, with both excitement and concern over potentially superhuman levels of artificial intelligence: models now take only seconds to produce outputs that would challenge or exceed the capabilities even of expert humans. At the same time, models still show basic errors in understanding that would not be expected even in non-expert humans. This presents us with an apparent paradox: how do we reconcile seemingly superhuman capabilities with the persistence of errors that few humans would make? In this work, we posit that this tension reflects a divergence in the configuration of intelligence in today's generative models relative to intelligence in humans. Specifically, we propose and test the Generative AI Paradox hypothesis: generative models, having been trained directly to reproduce expert-like outputs, acquire generative capabilities that are not contingent upon -- and can therefore exceed -- their ability to understand those same types of outputs. This contrasts with humans, for whom basic understanding almost always precedes the ability to generate expert-level outputs. We test this hypothesis through controlled experiments analyzing generation vs. understanding in generative models, across both language and image modalities. Our results show that although models can outperform humans in generation, they consistently fall short of human capabilities in measures of understanding, as well as weaker correlation between generation and understanding performance, and more brittleness to adversarial inputs. Our findings support the hypothesis that models' generative capability may not be contingent upon understanding capability, and call for caution in interpreting artificial intelligence by analogy to human intelligence.

Related papers

• Can Language Models Learn to Skip Steps? [59.84848399905409]
  We study the ability to skip steps in reasoning. Unlike humans, who may skip steps to enhance efficiency or to reduce cognitive load, models do not possess such motivations. Our work presents the first exploration into human-like step-skipping ability.
  arXiv  Detail & Related papers  (2024-11-04T07:10:24Z)
• "I Am the One and Only, Your Cyber BFF": Understanding the Impact of GenAI Requires Understanding the Impact of Anthropomorphic AI [55.99010491370177]
  We argue that we cannot thoroughly map the social impacts of generative AI without mapping the social impacts of anthropomorphic AI. anthropomorphic AI systems are increasingly prone to generating outputs that are perceived to be human-like.
  arXiv  Detail & Related papers  (2024-10-11T04:57:41Z)
• Exploration with Principles for Diverse AI Supervision [88.61687950039662]
  Training large transformers using next-token prediction has given rise to groundbreaking advancements in AI. While this generative AI approach has produced impressive results, it heavily leans on human supervision. This strong reliance on human oversight poses a significant hurdle to the advancement of AI innovation. We propose a novel paradigm termed Exploratory AI (EAI) aimed at autonomously generating high-quality training data.
  arXiv  Detail & Related papers  (2023-10-13T07:03:39Z)
• A Rubric for Human-like Agents and NeuroAI [2.749726993052939]
  Contributed research ranges widely from mimicking behaviour to testing machine learning methods. It cannot be assumed nor expected that progress on one of these three goals will automatically translate to progress in others. This is clarified using examples of weak and strong neuroAI and human-like agents.
  arXiv  Detail & Related papers  (2022-12-08T16:59:40Z)
• Complementary artificial intelligence designed to augment human discovery [2.7786142348700658]
  We reconceptualize and pilot beneficial AI to radically augment human understanding by complementing rather than competing with cognitive capacity. We use this approach to generate valuable predictions for what materials possess valuable energy-related properties. We demonstrate that our predictions, if identified by human scientists and inventors at all, are only discovered years further into the future.
  arXiv  Detail & Related papers  (2022-07-02T19:36:34Z)
• Humanly Certifying Superhuman Classifiers [8.736864280782592]
  Estimating the performance of a machine learning system is a longstanding challenge in artificial intelligence research. We develop a theory for estimating the accuracy compared to the oracle, using only imperfect human annotations for reference. Our analysis provides a simple recipe for detecting and certifying superhuman performance in this setting.
  arXiv  Detail & Related papers  (2021-09-16T11:00:05Z)
• Inductive Biases for Deep Learning of Higher-Level Cognition [108.89281493851358]
  A fascinating hypothesis is that human and animal intelligence could be explained by a few principles. This work considers a larger list, focusing on those which concern mostly higher-level and sequential conscious processing. The objective of clarifying these particular principles is that they could potentially help us build AI systems benefiting from humans' abilities.
  arXiv  Detail & Related papers  (2020-11-30T18:29:25Z)
• Machine Common Sense [77.34726150561087]
  Machine common sense remains a broad, potentially unbounded problem in artificial intelligence (AI) This article deals with the aspects of modeling commonsense reasoning focusing on such domain as interpersonal interactions.
  arXiv  Detail & Related papers  (2020-06-15T13:59:47Z)
• Dynamic Cognition Applied to Value Learning in Artificial Intelligence [0.0]
  Several researchers in the area are trying to develop a robust, beneficial, and safe concept of artificial intelligence. It is of utmost importance that artificial intelligent agents have their values aligned with human values. A possible approach to this problem would be to use theoretical models such as SED.
  arXiv  Detail & Related papers  (2020-05-12T03:58:52Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.