Submission 416
Mr Henrik Angeltveit
Submitter group: Individuals.
Position labels are not shown for individual submitters.
An AI model identified the submitter group and wrote the summary, key points and recommendation summaries. Counts do not measure importance or merit.
Summary
Read the original submission for the full text.
Argues that AI poses an extinction risk and advocates banning the development of superintelligence and of data centres used to train new models, legislation making it easy to withhold consent for IP use in AI training, and more funding for AI safety research. It acknowledges enforcement of a ban may be difficult.
Key points
Quotes are the submitter’s own words. Text taken from documents, including scanned pages read by text-recognition software, can differ from the PDF. Check the original. Listed in the same order as on the submission’s page in the interactive guide.
Argues that AI models will be capable of more harm as their abilities improve and that improvement should not continue.
IftheabilitiesofAI modelscontinuetoimprove,theywillbecapableofdoingmoreandmoreharm.Wemustnot allowAImodelstocontinueimproving.
A program corrected obvious text-recognition errors in this quote for display.
Argues that a ban on superintelligence development would cost nothing if superintelligence is impossible and could mean the difference between survival and extinction if it is possible.
Evenif superintelligenceisnotpossibleinthenear-term,banningsomethingimpossiblewouldcost nothing.Andifitispossible,abancouldpotentiallymeanthedifferencebetweensurvivaland extinction
A program corrected obvious text-recognition errors in this quote for display.
Says smarter models facilitate superintelligence potentially developed in countries with weaker regulations, and argues for banning stronger AI development.
buildingsmartermodelsstill makesbuildingsuperintelligenceeasier,andthisfinalstepcouldconceivablybedoneina differentcountrywithweakerregulations.Assuch,thereisagoodargumentforbanningthe developmentofstrongerAIsatall.
A program corrected obvious text-recognition errors in this quote for display.
Argues that legislation making it easy to withhold consent to IP use in AI training is essential.
Creating legislationthatwouldmakeiteasyforpeopletonotconsenttohavingtheirIPusedtotrainAIis essential.
A program corrected obvious text-recognition errors in this quote for display.
Argues safer AI may reduce mental-health dangers, especially for children and teenagers, and the risk of supplying information that enables harm to others.
SaferAImodelsmaybelessdangeroustopeople’smentalhealth(especiallychildren andteenagers),andalsoreducetheriskofAIprovidinginformationthatallowspeopletohurt others.
A program corrected obvious text-recognition errors in this quote for display.
Recommendations
Recommendations are shown beside the submission’s own words. Listed in the same order as on its page in the interactive guide.
Ban the development of superintelligence.
Iadvocateforabanonsuperintelligence
A program corrected obvious text-recognition errors in this quote for display.
Ban data centres used to train new AI models.
abanondatacentresusedtotrainnewmodels
A program corrected obvious text-recognition errors in this quote for display.
Consider banning development of stronger AIs altogether.
Assuch,thereisagoodargumentforbanningthe developmentofstrongerAIsatall.
A program corrected obvious text-recognition errors in this quote for display.
Legislate so people can easily decline consent for their IP to be used to train AI.
Creating legislationthatwouldmakeiteasyforpeopletonotconsenttohavingtheirIPusedtotrainAIis essential.
A program corrected obvious text-recognition errors in this quote for display.
Direct more money into AI safety research.
Instead,moremoneycanbedirectedintoresearchtoimproveAIsafety.
A program corrected obvious text-recognition errors in this quote for display.
Read the original
In some quotes a program corrected obvious text-recognition errors for display.