Gary Marcus

Rank 14 of 47
|
Score 89
3/15/2024, 9:41:21 PM
X
In reply to:
Derya Unutmaz, MD
@DeryaTR_
·
570d
@linasvepstas @GaryMarcus Yes, but LLM-generated content does not change the experimental results. Can you point out specific examples that are erroneous? Or fraudulent scientific data due the use of LLMs?
Linas
@linasvepstas
·
570d
@DeryaTR_ @GaryMarcus The posted examples of LLM-generated content were from materials science and chemistry articles, published by Elsevier. One had all-Chinese authors, the other all-Islamic authors.
Derya Unutmaz, MD
@DeryaTR_
·
571d
@GaryMarcus This is not a major issue in STEM research, as you need to perform actual experiments and generate meaningful data to publish; LLMs can’t do the experiments. However, this could be problematic in social sciences.
Gary Marcus
@GaryMarcus
·
571d
Mayday.

The sudden pollution of science with LLM-generated content, known to yield plausible-sounding but sometimes difficult to detect errors (“hallucinations”) is serious, and its impact will be lasting.

Please share this short essay (link below) with scientists and with…

The statement in question is part of a conversation about the impact of language model-generated content on scientific research. The tone of the statement is critical and suggests skepticism about the peer review process of certain papers, implying that errors could be present due to lack of thorough review. The statement also makes a distinction between fraud and error, indicating that the issue at hand is related to mistakes rather than intentional deceit. The image attached to the conversation contains text that is critical of the quality of some scientific publications, using strong language to describe the situation.

  1. Principle 1:
    I will strive to do no harm with my words and actions.
    The statement is critical but does not seem to intend harm. It aims to highlight potential issues in the scientific review process. [+1]
  2. Principle 2:
    I will respect the privacy and dignity of others and will not engage in cyberbullying, harassment, or hate speech.
    The statement respects the privacy and dignity of others by not singling out individuals but rather commenting on a process. [+1]
  3. Principle 4:
    I will engage in constructive criticism and dialogue with those in disagreement and will not engage in personal attacks or ad hominem arguments.
    The statement engages in constructive criticism by pointing out the need for technical expertise in reviewing scientific papers. [+1]
  4. Principle 6:
    I will use my influence for the betterment of society.
    The statement uses influence to raise awareness about the importance of thorough review in scientific publications, which could be seen as contributing to the betterment of society. [+1]