• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI
    Reaction

    A proposal to reshape research publication around clear, interesting work

    A poster argues that AI has made publication less reliable as a stamp of correctness or prestige. They propose judging papers regardless of provenance and requiring video talks to show human understanding.

    BB
    TF
    3 Sources, ,

    TLDR

    In a reply, a poster proposes that publications collect interesting, well-written work rather than serve as stamps of correctness or prestige. Their proposed rules would permit AI use in submissions and in checking technical claims during review, while leaving judgments of novelty, significance and interest to humans. Papers could be rejected for unclear writing or a poor accompanying video talk, regardless of technical merit.

    Combined views

    16.8K

    3 Sources, first seen 13h ago

    Combined views

    16.8K

    3 Sources, first seen 13h ago

    96 likes
    13h ago
    first seen 13h ago
    96 likes
    7 comments
    40 saves
    19 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    7 comments
    40 saves
    19 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    3 Sources

    @tonylfengPUBLICATION Until now, publication has been a source of prestige and a stamp of correctness. AI has exposed the limitations of the latter function, and is quickly obviating it; AI has also made the former function unreliable. We propose that publication henceforth focus on collecting interesting and well-written work, regardless of provenance; prestige will have to come from elsewhere. This may be difficult for old-timers like me who have developed a psychological attachment to prestigious publications; we’ll have to overcome that for the health of the field. Rejecting based on exposition. “Poor exposition is sufficient grounds for rejection. Papers may be rejected with a one sentence review, regardless of the technical merit of the paper, solely because of exposition which is sufficiently unclear.” AI use should not be penalized. “AI use in the ideas or writing of the submission is permitted, and reviewers may not penalize a paper for AI use as long as it does not inhibit human understanding.” AI use should be allowed for reviewing. “AI assistance should be allowed for review, in checking technical claims. Evaluation of novelty, significance, and interest must be done by humans.” Submissions must have an accompanying video talk. “The point of the talk is to provide evidence of human understanding, through a clear presentation of the merits of the paper, including a discussion of the paper’s importance within the context of prior work and a description of the key technical and conceptual ideas used to achieve its main claims. Reviewers are not required to watch the talk, but are permitted to reject a paper with a poor talk solely due to the quality of the talk with no consideration of the technical merits of the paper.”
    @boazbaraktcsRT @tonylfeng: PUBLICATION Until now, publication has been a source of prestige and a stamp of correctness. AI has exposed the limitations…

    3 Sources

    @tonylfengPUBLICATION Until now, publication has been a source of prestige and a stamp of correctness. AI has exposed the limitations of the latter function, and is quickly obviating it; AI has also made the former function unreliable. We propose that publication henceforth focus on collecting interesting and well-written work, regardless of provenance; prestige will have to come from elsewhere. This may be difficult for old-timers like me who have developed a psychological attachment to prestigious publications; we’ll have to overcome that for the health of the field. Rejecting based on exposition. “Poor exposition is sufficient grounds for rejection. Papers may be rejected with a one sentence review, regardless of the technical merit of the paper, solely because of exposition which is sufficiently unclear.” AI use should not be penalized. “AI use in the ideas or writing of the submission is permitted, and reviewers may not penalize a paper for AI use as long as it does not inhibit human understanding.” AI use should be allowed for reviewing. “AI assistance should be allowed for review, in checking technical claims. Evaluation of novelty, significance, and interest must be done by humans.” Submissions must have an accompanying video talk. “The point of the talk is to provide evidence of human understanding, through a clear presentation of the merits of the paper, including a discussion of the paper’s importance within the context of prior work and a description of the key technical and conceptual ideas used to achieve its main claims. Reviewers are not required to watch the talk, but are permitted to reject a paper with a poor talk solely due to the quality of the talk with no consideration of the technical merits of the paper.”
    @boazbaraktcsRT @tonylfeng: PUBLICATION Until now, publication has been a source of prestige and a stamp of correctness. AI has exposed the limitations…