Will AI avoid exploitation? Artificial general intelligence and expected utility theory.

A simple argument suggests that we can fruitfully model advanced AI systems using expected utility theory. According to this argument, an agent will need to act as if maximising expected utility if they're to avoid exploitation. Insofar as we should expect advanced AI to avoid exploitation, it follo...

Descripción completa

Detalles Bibliográficos
Publicado en:Philosophical Studies Vol. 182; no. 7; pp. 1519 - 1539
Autor principal: Bales, Adam
Formato: Artículo
Publicado: Springer Nature Jul2025
Materias:
Acceso en línea:Ver este registro en EBSCOhost
fields @attributes:
  recordID: 1
pdfLink:
plink: https://search.ebscohost.com/login.aspx?direct=true&db=hlh&AN=186909892&site=ehost-live
header:
  @attributes:
    shortDbName: hlh
    uiTerm: 186909892
    longDbName: Humanities International Complete
    uiTag: AN
  controlInfo:
    bkinfo:
    jinfo:
      jid:
        00318116
        4L8
      jtl: Philosophical Studies
      issn: 00318116
      maglogo: N
    pubinfo:
      dt: Jul2025
      vid: 182
      iid: 7
      pid: 237
      pub: Springer Nature
    artinfo:
      ui:
        186909892
        10.1007/s11098-023-02023-4
      ppf: 1519
      ppct: 20
      formats:
        fmt:
          – @attributes:
              type: T
          – @attributes:
              type: P
              size: 894KB
      tig:
        atl: Will AI avoid exploitation? Artificial general intelligence and expected utility theory.
      aug:
        au: Bales, Adam
        affil: https://ror.org/052gg0110 University of Oxford, England, UK
      su:
        Artificial intelligence
        Utility theory
        Digital technology
        Argument
        General factor (Psychology)
      sug:
        subj:
          Artificial intelligence
          Utility theory
          Digital technology
          Argument
          General factor (Psychology)
      keyword:
        Artificial general intelligence
        Expected utility theory
        Money pump arguments
      ab: A simple argument suggests that we can fruitfully model advanced AI systems using expected utility theory. According to this argument, an agent will need to act as if maximising expected utility if they're to avoid exploitation. Insofar as we should expect advanced AI to avoid exploitation, it follows that we should expected advanced AI to act as if maximising expected utility. I spell out this argument more carefully and demonstrate that it fails, but show that the manner of its failure is instructive: in exploring the argument, we gain insight into how to model advanced AI systems.
      pubtype: Academic Journal
      doctype: Article
      src: R
    language: English
    refInfo:
    copyright:
      @attributes:
        flag: Y
      custom: Philosophical Studies is a copyright of Springer, 2025. All Rights Reserved.
      item: Philosophical Studies
      holder: Springer Nature
      dt:
        @attributes:
          year: 2025
    holdings:
      @attributes:
        islocal: N