{"id":"https://openalex.org/W7204237472","doi":"https://doi.org/10.48550/arxiv.2608.23875","title":"AI Finds A Way","display_name":"AI Finds A Way","publication_year":2026,"publication_date":"2026-08-24","ids":{"openalex":"https://openalex.org/W7204237472","doi":"https://doi.org/10.48550/arxiv.2608.23875"},"language":null,"primary_location":{"id":"pmh:oai:arXiv.org:2608.23875","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.23875","pdf_url":"https://arxiv.org/pdf/2608.23875","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"volume":null,"type":"preprint","indexed_in":["arxiv","datacite"],"open_access":{"is_oa":true,"oa_status":"green","oa_url":"https://arxiv.org/pdf/2608.23875","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5097955681","display_name":"Aaron Dharna","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]},{"id":"https://openalex.org/I4210127509","display_name":"Vector Institute","ror":"https://ror.org/03kqdja62","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210127509"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Dharna, Aaron","raw_affiliation_strings":["University of British Columbia","Vector Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of British Columbia","institution_ids":["https://openalex.org/I141945490"]},{"raw_affiliation_string":"Vector Institute","institution_ids":["https://openalex.org/I4210127509"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5148256197","display_name":"Cong Lu","orcid":null,"observed_orcids":[]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Lu, Cong","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5146701629","display_name":"Ryan Sullivan","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I141945490","display_name":"University of British Columbia","ror":"https://ror.org/03rmrcq20","country_code":"CA","type":"education","lineage":["https://openalex.org/I141945490"]},{"id":"https://openalex.org/I4210127509","display_name":"Vector Institute","ror":"https://ror.org/03kqdja62","country_code":"CA","type":"facility","lineage":["https://openalex.org/I4210127509"]}],"countries":["CA"],"is_corresponding":false,"raw_author_name":"Sullivan, Ryan","raw_affiliation_strings":["University of British Columbia","Vector Institute"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of British Columbia","institution_ids":["https://openalex.org/I141945490"]},{"raw_affiliation_string":"Vector Institute","institution_ids":["https://openalex.org/I4210127509"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5053798319","display_name":"Joel Lehman","orcid":"https://orcid.org/0000-0002-9535-1123","observed_orcids":["https://orcid.org/0000-0002-9535-1123"]},"institutions":[{"id":"https://openalex.org/I40120149","display_name":"University of Oxford","ror":"https://ror.org/052gg0110","country_code":"GB","type":"education","lineage":["https://openalex.org/I40120149"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Lehman, Joel","raw_affiliation_strings":["University of Oxford"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"University of Oxford","institution_ids":["https://openalex.org/I40120149"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5052300917","display_name":"Victoria Krakovna","orcid":null,"observed_orcids":[]},"institutions":[{"id":"https://openalex.org/I4210090411","display_name":"Google DeepMind (United Kingdom)","ror":"https://ror.org/00971b260","country_code":"GB","type":"company","lineage":["https://openalex.org/I4210090411","https://openalex.org/I4210128969"]}],"countries":["GB"],"is_corresponding":false,"raw_author_name":"Krakovna, Victoria","raw_affiliation_strings":["Google DeepMind"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Google DeepMind","institution_ids":["https://openalex.org/I4210090411"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5112191507","display_name":"Jeff Clune","orcid":"https://orcid.org/0000-0003-4890-5274","observed_orcids":["https://orcid.org/0000-0003-4890-5274"]},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Clune, Jeff","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":2,"institutions_distinct_count":4,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":null,"has_fulltext":true,"cited_by_count":0,"citation_normalized_percentile":null,"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":null,"last_page":null},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T13904","display_name":"Artificial Intelligence Applications","score":0.5073000192642212,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T13904","display_name":"Artificial Intelligence Applications","score":0.5073000192642212,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10462","display_name":"Reinforcement Learning in Robotics","score":0.2874000072479248,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10883","display_name":"Ethics and Social Impacts of AI","score":0.1371999979019165,"subfield":{"id":"https://openalex.org/subfields/3311","display_name":"Safety Research"},"field":{"id":"https://openalex.org/fields/33","display_name":"Social Sciences"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/reward-hacking","display_name":"reward hacking","score":0.6869999766349792},{"id":"https://openalex.org/keywords/reward-specification","display_name":"reward specification","score":0.3919999897480011},{"id":"https://openalex.org/keywords/foundation-models","display_name":"foundation models","score":0.609000027179718},{"id":"https://openalex.org/keywords/reinforcement-learning","display_name":"reinforcement learning","score":0.5979999899864197},{"id":"https://openalex.org/keywords/ai-safety","display_name":"AI safety","score":0.5600000023841858},{"id":"https://openalex.org/keywords/ai-alignment","display_name":"AI alignment","score":0.39100000262260437},{"id":"https://openalex.org/keywords/scientific-discovery","display_name":"scientific discovery","score":0.6290000081062317},{"id":"https://openalex.org/keywords/computational-creativity","display_name":"computational creativity","score":0.5669999718666077}],"concepts":[{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.6593000292778015},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6489999890327454},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5910000205039978},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.4620000123977661},{"id":"https://openalex.org/C206345919","wikidata":"https://www.wikidata.org/wiki/Q20380951","display_name":"Resource (disambiguation)","level":2,"score":0.4440999925136566},{"id":"https://openalex.org/C188147891","wikidata":"https://www.wikidata.org/wiki/Q147638","display_name":"Cognitive science","level":1,"score":0.4018000066280365},{"id":"https://openalex.org/C18762648","wikidata":"https://www.wikidata.org/wiki/Q42213","display_name":"Work (physics)","level":2,"score":0.33899998664855957},{"id":"https://openalex.org/C136197465","wikidata":"https://www.wikidata.org/wiki/Q1729295","display_name":"Variety (cybernetics)","level":2,"score":0.33059999346733093},{"id":"https://openalex.org/C2780966255","wikidata":"https://www.wikidata.org/wiki/Q5474306","display_name":"Foundation (evidence)","level":2,"score":0.3116999864578247},{"id":"https://openalex.org/C539667460","wikidata":"https://www.wikidata.org/wiki/Q2414942","display_name":"Management science","level":1,"score":0.30550000071525574},{"id":"https://openalex.org/C162027153","wikidata":"https://www.wikidata.org/wiki/Q2264109","display_name":"Artificial general intelligence","level":2,"score":0.28369998931884766},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.27570000290870667},{"id":"https://openalex.org/C108583219","wikidata":"https://www.wikidata.org/wiki/Q197536","display_name":"Deep learning","level":2,"score":0.26179999113082886},{"id":"https://openalex.org/C105409693","wikidata":"https://www.wikidata.org/wiki/Q5937824","display_name":"Human intelligence","level":2,"score":0.26109999418258667},{"id":"https://openalex.org/C11012388","wikidata":"https://www.wikidata.org/wiki/Q170658","display_name":"Creativity","level":2,"score":0.257099986076355},{"id":"https://openalex.org/C157170001","wikidata":"https://www.wikidata.org/wiki/Q4781507","display_name":"Applications of artificial intelligence","level":2,"score":0.2529999911785126}],"mesh":[],"locations_count":2,"locations":[{"id":"pmh:oai:arXiv.org:2608.23875","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.23875","pdf_url":"https://arxiv.org/pdf/2608.23875","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},{"id":"doi:10.48550/arxiv.2608.23875","doi":"https://doi.org/10.48550/arxiv.2608.23875","is_oa":true,"landing_page_url":"https://doi.org/10.48550/arxiv.2608.23875","pdf_url":null,"source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":null,"is_accepted":false,"is_published":null,"raw_source_name":null,"raw_type":"Preprint"}],"best_oa_location":{"id":"pmh:oai:arXiv.org:2608.23875","doi":null,"is_oa":true,"landing_page_url":"https://arxiv.org/abs/2608.23875","pdf_url":"https://arxiv.org/pdf/2608.23875","source":{"id":"https://openalex.org/S4306400194","display_name":"arXiv (Cornell University)","issn_l":"2331-8422","issn":["2331-8422"],"is_oa":true,"is_in_doaj":false,"is_core":false,"listed_in":[],"host_organization":"https://openalex.org/I205783295","host_organization_name":"Cornell University","host_organization_lineage":["https://openalex.org/I205783295"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"text"},"sustainable_development_goals":[],"sustainable_development_goals_aurora":[],"study_designs":[],"awards":[],"funders":[],"has_content":{"pdf":true,"grobid_xml":true},"content_urls":{"pdf":"https://content.openalex.org/works/W7204237472.pdf","grobid_xml":"https://content.openalex.org/works/W7204237472.grobid-xml"},"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Artificial":[0],"Intelligence":[1],"(AI)":[2],"algorithms":[3],"frequently":[4],"learn":[5],"creative":[6],"and":[7,16,85,220,238],"unexpected":[8,87,226],"solutions,":[9],"surprising":[10,128],"even":[11],"expert":[12],"researchers":[13],"who":[14],"develop":[15],"study":[17],"them.":[18,191],"They":[19,109],"often":[20],"astonish":[21],"practitioners":[22],"by":[23],"discovering":[24],"unanticipated":[25],"behavior,":[26],"exploiting":[27],"loopholes":[28],"in":[29,187,230],"reward":[30,164],"signals,":[31],"or":[32,165],"spontaneously":[33],"uncovering":[34],"previously":[35],"unknown":[36],"scientific":[37,205],"phenomena.":[38],"However,":[39,151],"accounts":[40,98],"of":[41,67,76,105,114],"such":[42],"unconventional":[43],"behavior":[44],"across":[45,147],"machine":[46,61],"learning":[47,62,146,198],"are":[48,99],"seldom":[49],"formally":[50],"documented.":[51],"This":[52],"work":[53,66,211],"presents":[54],"26":[55],"curated":[56],"firsthand":[57],"anecdotes":[58,72],"from":[59],"various":[60],"subfields":[63],"representing":[64],"the":[65,74,90,103,111,157,223,234],"over":[68],"100":[69],"researchers.":[70],"These":[71],"showcase":[73],"capability":[75],"modern":[77,231],"AI":[78,107,140],"systems":[79],"to":[80,89,160,203,216,236],"circumvent":[81],"human-imposed":[82],"design":[83],"limitations":[84],"discover":[86],"solutions":[88],"tasks":[91],"we":[92,193,208],"train":[93],"them":[94],"on.":[95],"Furthermore,":[96],"these":[97,183,196],"particularly":[100],"important":[101],"for":[102,242],"safety":[104],"future":[106,218],"systems.":[108],"illustrate":[110],"fundamental":[112,184],"challenge":[113],"aligning":[115],"models":[116,178],"with":[117],"human":[118],"values":[119],"without":[120,130],"diminishing":[121],"their":[122],"creativity,":[123],"so":[124],"they":[125],"can":[126,154,189,200],"make":[127],"discoveries":[129],"producing":[131],"surprising,":[132],"potentially":[133],"harmful":[134],"outcomes.":[135],"The":[136],"paper":[137],"first":[138],"details":[139],"achieving":[141],"superhuman":[142],"success":[143],"through":[144],"reinforcement":[145],"many":[148],"challenging":[149],"domains.":[150],"reward-driven":[152],"optimization":[153],"fail":[155],"when":[156],"model":[158],"learns":[159],"hack":[161],"an":[162],"underspecified":[163],"unarticulated":[166],"constraint.":[167],"We":[168],"then":[169],"present":[170],"case":[171],"studies":[172],"suggesting":[173],"that":[174,195,222],"harnessing":[175],"internet-scale":[176],"foundation":[177],"(FMs)":[179],"has":[180],"not":[181],"resolved":[182],"challenges":[185],"and,":[186],"fact,":[188],"supercharge":[190],"Nevertheless,":[192],"argue":[194],"same":[197],"dynamics":[199],"be":[201],"harnessed":[202],"accelerate":[204],"discovery.":[206],"Finally,":[207],"hope":[209],"this":[210],"provides":[212],"a":[213],"consolidated":[214],"resource":[215],"inform":[217],"research":[219],"demonstrates":[221],"tendency":[224],"toward":[225],"behaviors":[227],"is":[228],"commonplace":[229],"AI,":[232],"highlighting":[233],"need":[235],"anticipate":[237],"manage":[239],"AI's":[240],"capacity":[241],"innovative,":[243],"yet":[244],"unpredictable,":[245],"solutions.":[246],"(abstract":[247],"abridged)":[248]},"counts_by_year":[],"updated_date":"2026-09-30T09:27:27.532063","created_date":"2026-08-27T00:00:00"}