Virtual personal assistants powered by artificial intelligence are becoming ubiquitous across technology platforms, with every major tech firm adding AI to their services and dozens of amasevisi akhethekile tumbling onto the market. While immensely useful, researchers from Google say humans could become too emotionally attached to them, leading to a host of negative social consequences.
Okusha ucwaningo iphepha from Google’s DeepMind AI research laboratory highlights the potential benefits of advanced, personalized AI assistants to transform various aspects of society, saying they “could radically alter the nature of work, education, and creative pursuits as well as how we communicate, coordinate, and negotiate with one another, ultimately influencing who we want to be and to become.”
Lo mthelela wokudlula, vele, ungaba inkemba esika nhlangothi zombili uma ukuthuthukiswa kwe-AI kuqhubeka nokusheshisa ngaphandle kokuhlela okucatshangelwayo.
Ingozi eyodwa ebalulekile? Ukwakhiwa kwamabhondi asondelene ngokungafanele—okungase kube kubi nakakhulu uma umsizi ethulwa ngokumelela okufana nomuntu noma ubuso. "Lawa ma-ejenti okwenziwa angasho ngisho nokuthi athanda umsebenzisi, abeke isisekelo sokuthi abasebenzisi bakhe ubuhlobo besikhathi eside be-AI," kusho leli phepha.
Uma kungahloliwe, okunamathiselwe okunjalo kungaholela ekulahlekelweni kokuzimela komsebenzisi kanye nokulahlekelwa ubudlelwano bomphakathi ngoba i-AI ingangena esikhundleni sokusebenzelana kwabantu.
This risk is not purely theoretical. Even when AI was in a somewhat primitive state, an AI chatbot was influential enough to convince an user to commit suicide after a long chat back in 2023. Eight years ago, an AI-powered email assistant named “Amy Ingram” was realistic enough to prompt some users to send love notes and even attempt to visit her at work.
U-Iason Gabriel, usosayensi ocwaningayo eqenjini lokucwaninga ngokuziphatha le-DeepMind kanye nomlobi wephepha, akazange amphendule. I-Decrypt isicelo sokuphawula.
Ku-tweet, nokho, u-Garbriel uxwayise ngokuthi "izinhlobo zomsizi ezikhulayo zomuntu siqu nezifana nezomuntu zethula imibuzo emisha nge-anthropomorphism, ubumfihlo, ukwethembana kanye nobudlelwano obufanele ne-AI."
Ngenxa yokuthi “izigidi zabasizi be-AI zingasatshalaliswa ezingeni lomphakathi lapho zizosebenzisana khona kanye nabangewona abasebenzisi,” uGabriel uthe ukholelwa esidingweni sokuvikela okwengeziwe kanye nendlela ephelele kulo mkhuba omusha womphakathi.
Iphepha locwaningo liphinde lixoxe ngokubaluleka kokuqondanisa inani, ukuphepha, kanye nokusetshenziswa kabi ekuthuthukisweni kwabasizi be-AI. Noma abasizi be-AI bengasiza abasebenzisi bathuthukise inhlalakahle yabo, bathuthukise ubuhlakani babo, futhi basebenzise isikhathi sabo, ababhali baxwayise ngezingozi ezengeziwe njengokungahlotshaniswa kahle nomsebenzisi nezithakazelo zomphakathi, ukubekwa kwamanani kwabanye, ukusetshenziselwa izinjongo ezinonya, nokuba sengozini. ekuhlaselweni yizitha.
Ukuze kubhekwane nalezi zingozi, ithimba le-DeepMind lincoma ukuthuthukisa ukuhlola okuphelele kwabasizi be-AI nokusheshisa ukuthuthukiswa kwabasizi be-AI abazuzisa umphakathi.
“Njengamanje simi ekuqaleni kwalesi sikhathi sezinguquko kwezobuchwepheshe nezenhlalo. Ngakho-ke sinethuba lokwenza manje—njengabathuthukisi, abacwaningi, abenzi bezinqubomgomo, kanye nababambiqhaza bomphakathi—ukubumba uhlobo lwabasizi be-AI esifuna ukubabona emhlabeni.”
AI misalignment can be mitigated through Reinforcement Learning Through Human Feedback (RLHF), which is used to train AI models. Experts like Paul Christiano, who ran the language model alignment team at OpenAI and now leads the non-profit Alignment Research Center, warn that improper management of AI training methods could end in catastrophe.
“I think maybe there’s something like a 10-20% chance of AI takeover, [with] many [or] most humans dead, ” Paul Christiano wathi on the Bankless podcast last year. “I take it quite seriously.”
eyahlelwa Ryan Ozawa.
Hlala unolwazi lwezindaba ze-crypto, thola izibuyekezo zansuku zonke ebhokisini lakho lokungenayo.
