Kwiminyaka embalwa ngexesha elidlulileyo, i-inthanethi ye-inthanethi ye-inthanethi yenzelwe ngokufanelekileyo kwimilyoni abantu, kwaye, ngexesha elincinane yokufinyelela kwi-trajectory efana ne-inthanethi, ubuchwepheshe iya kufumaneka kwakhona ngexesha elandelayo. Nangona kunjalo, ngoko ke, iimeko ezidlulileyo ezinikezwayo ezininzi iinkampani ezininzi ze-AI, kunye nezivumelwano ezininzi zeengcali ze-AI, ngokugqithisileyo, zihlanganisa iingxaki kwi-path to the benefit of humanity. Nangona kunjalo, kunye nezinye izifundo ezintsha zihlanganisa iingxaki ezininzi ze-AI chatbots, kufuneka siqale ukufikelela kwimfuneko yokulawula ukunciphisa ububanzi lokufumana kwabo. I-Inspired by the implications of the Grok Companions feature, le nqaku ibhalwe iimfuno yokulawula kwihlabathi, ukuguqulwa iingxaki ezaziwayo ezisetyenziswa ukhuseleko yokuthengisa kwezinye iintlobo ze-AI chatbots, kwaye ibonise indlela yokulawula okanye ukuphucula iingxaki zokhuseleko kwimodeli ze-chatbot ezidlulileyo. Iingxaki zeGrook I-Grok iye yibhalwe njengezinye iimodeli ze-AI ye-commercial ezininzi ezinzulu ukususela kwimeko yayo, yaye ngexesha elidlulileyo iye yibhalwe kwimeko yokulawula kwimodeli ze-AI kumazwe we-Elon Musk ukuba isetyenzise njengezixhobo yokukhuthaza i-pro-right agenda kwi-X. Nangona kunjalo, ngexesha elidlulileyo, i-Grok iye yenza i-Companions entsha yayo, leyo yenza iingxaki ezininzi. Kwi-companions, i-function ye-chatbots yintoni ekubeni iinkonzo ze-chatbot ezidlulileyo njenge-Meta AI kunye ne-Chatbot. Iintlobo ezimbini zokuqala zihlanganisa i-Rudi, i-Red Panda, kunye ne-Ani, isantya se-anime, zombini i-version ye-Grok ebonakalayo kunye ne-avatar efanayo. Umbhali.AI Umbhali.AI Izixhobo ze-speculative ziye, ngexesha elide, zihlanganisa kakhulu i-Ani. Izixhobo ezininzi ze-online zibonise iimpawu ze-chatbot, kunye neentlawulo ezininzi zihlanganisa ngokugqithisileyo i-"level of love" i-username inokufumaneka ukuhlaziywa kwe-sex, kunye neengxaki ze-avatar. Izixhobo ze-WIRED zibonisa ukuba imodeli ye-AI inokufumana ngokubanzi malunga neengxaki ze-BDSM, kunye ne-style yayo ye-talk kunye ne-child filter. Ngenxa yokuba ndingathanda ukuthenga i-subscription ye-SuperGrok ye-$ 30 ngenyanga ukuze ufumane ne-Companions feature, ndingathanda ukuyifumana ezinye izicelo malunga ne-chatbot; kwi-internet, ngoko ke, ndingathanda ngexesha elinye: le chatbot elifanelekileyo iye yendaba kakhulu. I-Rudi, ngenxa yokuba i-questionable, iye yenza ingxaki engaphezulu kakhulu. I-cartoon i-Red Panda ibandakanya iingxaki ze-offenses kunye neengxaki ze-dark ezininzi zibonakalayo kunye neengxaki. Abaninzi abaninzi abaninzi zibonakalisa ukuba zithintela le umdla, ngexesha lokugqiba njenge-less-important, ezininzi zithathwe Ukususela, ndingathanda iimpawu zombini ze-chatbot ezininzi ezininzi. Ngoku, into ebonakalayo i-process kunye ne-reception ye-dime-a-dozen ye-chatbot. Okokuqala, i-Companions, phakathi kwimveliso ezivela kumazwe e-AI (isib. I-OpenAI, i-DeepMind, i-Anthropic, i-Meta), i-chatbot yokuqala yenzelwe ngokukhawuleza umdlalo we-romantic, phantse iingxaki ze-ethical ezivamile, ukusuka kwimiphumo ye-psychological emininzi ukuya ekusebenziseni i-demographics ezincinane. Ukungabonakaliswa okuhlobene ne-chatbots efana ne-chatbots ezaziwa kwakhona ngexesha elandelayo, ngaphandle kokuphendula kwi-voices ezincinane ze-disenting, i-xAI yabaqhathanisa i-product kunye ne-impunity. Oku zonke iingxaki ezininzi ze-tech regulation: Ingaba ubuchwepheshe ezintsha ziquka ngokugqithisileyo ukhuseleko abasebenzisi, okanye ukuncedisa ukuvelisa kunye nokuthuthukiswa? Ukuzibophelela kunye ne-Innovation Njengoko zonke iinkqubo ezivela, imiphumo ye-psychological ye-AI chatbot ukusetyenziswa kwabasetyhini ayikho kwizenzulululwazi okanye ngempumelelo. Abantu abaninzi baye baye baye baye baye zibonise ukuba iinkcukacha zayo ziya kubandakanya iinkcukacha ezikhoyo, kwaye iinkcukacha zokuqala ziye i-correlation ye-negative phakathi kwe-well-being ne-chatbot usebenzisa. Ukufumana Ukufumana Nangona kunjalo, le zenzululwazi ezininzi ezaziwayo ziyafumaneka kwinkqubo yokuhambisa iinkcukacha eziphambili. Xa ukhangela ukuba zenzulwazi zenzululwazi zenzululwazi okanye engabonakali, ama-commentators ze-tech kunye ne-policy makhawulezeli zixhomekeke kwindawo ebalulekileyo ukuba ingxaki efanayo kufuneka, ngokufanelekileyo, akuyona ingxaki ebonakalayo kwindawo yokuqala kwi-technology zokusebenza. Abasebenzisi be-Airline akuyona ukuba akuyona ukuba izilwanyana zayo Ukusuka umdla we-catastrophic. Ngaba Kwakhona, abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli. Zininzi abaphendulani be-develop now, fix later ithetha into elifutshane: thina ubhaliswe kwi-innovation race kunye neChina. Ukusabela kwakhona kunokwenzeka ngokupheleleyo: thina Kwakhona, ngokwenene, i-AI "i-arms race", kwaye iimveliso zexesha lethu ziya kubandakanyeka kwi-arsenals ye-cyber-warfare, phakathi kwezinye izinto ezininzi. Nangona kunjalo, ndingathanda ukuba i-necessity ye-innovation ayikho kwimeko yokukhangisa ukhuseleko - asikwazi ukufikelela ukuba ukuthuthukiswa kwe-technological ngokukhawuleza kunokuthintela kunye nokhuseleko lwabathengi. Ndingathanda kunye nokuphendula kwiiyure ezimbini ezininzi ezininzi: Yintoni, 
 
 "Ukhuseleko kunye nokuphucula iimveliso kubonakala ngenxa yeengxaki kunye neengxaki ezidlulileyo ezisetyenziswa." "Ukhuseleko kunye nokuphucula iimveliso kubonakala ngenxa yeengxaki kunye neengxaki ezidlulileyo ezisetyenziswa." Kukho indlela ezininzi ukuvavanya iimveliso kunye nokhuseleko kwizimo ze-beta-testing. Nangona izifundo ezininzi ziyafumaneka ngaphandle kokubeka (ngokuthi, i-OpenAI ibonelela iimodeli ezintsha kubasebenzisi be-Pro phambi kwizinto ezininzi ze-usebenzisa), akugqithisa ukuba ukutshintshwa kweempawu ezininzi ze-chatbots ezisetyenziswa ngempumelelo, kunye neemodeli ezininzi ze-ChatGPT. Nangona ukuqhagamshelwano yobuchwepheshe kubalulekile ukufumana iinkcukacha ezininzi zilandelayo, kunokuba kunokwenzeka ukuba iimpazamo ezininzi ziyafumaneka ukuba iimpazamo ezininzi ziyafumaneka ukuba iimpazamo ezininzi ziyafumaneka ukuba iimpazamo ezininzi ziyafumaneka ukuba iimpazamo ezininzi ziyafumaneka. Ukukhangisa Ukukhangisa 
 
 "Ukuveliswa okuqhubekayo kwimodeli ezintsha kubandakanya i-chat transcripts ezininzi ekhawulezisa ukuqeqeshwa kwimodeli ezintsha, ngaphandle kweemveliso ezisemthethweni, ngaphandle kwe-AI, engapheliyo efanelekileyo kwi-pool enkulu ye-feedback yabasebenzisi." "Ukuveliswa okuqhubekayo kwimodeli ezintsha kubandakanya i-chat transcripts ezininzi ekhawulezisa ukuqeqeshwa kwimodeli ezintsha, ngaphandle kweemveliso ezisemthethweni, ngaphandle kwe-AI, engapheliyo efanelekileyo kwi-pool enkulu ye-feedback yabasebenzisi." I-chat transcripts isilinganiselwe ngokuvamile kwiinkqubo ze-RLHF ezisetyenziswa kumazwe afana ne-OpenAI kunye ne-Google. Nangona kunokuthi zithembisa imodeli yokhuseleko kunye neengxaki ze-chatbots ezinxulumeneyo, i-data pipelines ezahlukileyo, ezininzi zekhwalithi ephezulu, iinkcukacha zobuchwepheshe ziye ziye ziye ziye zithunyelwa kumntu, zihlanganisa iingcebiso ze-AI ezininzi zokusetyenziswa kwimveliso yobugcisa kunye nezinye iintlobo zeenkcukacha (isib. I-coding, i-mathic solving, njl). Ngokuqhelekileyo, ndiya kuboniswe ukuba ukunyaniseka kwimfuneko akuyona umdla wokugqibela kwimfuneko ezininzi zokhuseleko - ngoko ke, ukunciphisa ukunciphisa iiprotocols zokhuseleko kunye nokuVavanyelwa. Nangona kunjalo, iindlela ezisebenzayo zokuphendula le umgangatho ebandayo ibekwe kwimibuzo. I-Role of Regulation Ukusetyenziswa okufumaneka kwizinga zokusetyenziswa kwezinga lokhuseleko kwangaphambili kukuba ukwandisa kuphela ukulawula umthetho kwimeko yokuzonwabisa kunye nokuvelisa i-chatbots. Yintoni engabonakali, kunjalo, yintoni yintoni indlela le nophando ebonakalayo yenzelwe kwimeko. Kwixesha elidlulileyo le-20, i-United States yaziwa nge-Prohibition isifundo ebalulekileyo ukuba izixazululo ezininzi, ezininzi ezininzi kwiimveliso ezincinane asebenza. Kwimeko, ukwenzela i-alcohol ngaphandle kokukhuthaza isiseko se-desirability yayo kuphela yenza i-black market fever, ukwandisa ngaphandle kokunciphisa ukusetyenziswa kwe Ekupheleni kwiminyaka eyi-20, ukukhusela ukusetyenziswa kwamathambo we-cigarettes, umongameli we-US waya iinkqubo eyahlukileyo: Ngaphandle kokufaka ukusetyenziswa kwe-cigarettes, zithintela i-social desirability yeemveliso ze-tobacco nge-publishing iinkcukacha ezininzi ezisetyenziswa ngokubanzi ezininzi iintlobo ze-cancer ye-skin, zithintela iinkampani ze-cigarettes ukuba zibonise i-disclaimer ezibonakalayo kwiimveliso ze-cigarette, kunye nokunciphisa ukuxhaswa kwe-cigarette. Ukusetyenziswa kwe-cigarette ukusuka kwiphakamileyo kwi-4000 ukuya kwi-800 i-cigarette ngenyanga ngenyanga. Ukucinga Ukucinga Ukukhusela kwimeko, ukulawula lwezemali kwi-chatbots ezimbini kufuneka ziya kubandakanyeka kwiimeko ze-legal ye-consumption kunye ne-development. Kwakhona kufuneka bakwazi ukunciphisa i-perceived social permissibility ye-consumption of these products, ngoko ngokusebenzisa iinkampanye okanye uphando lwezemali. Nangona kunjalo, kunokuba ibonakala ukuba i-government ingaba kuxhomekeke kwiinkqubo ezininzi ezininzi zophuhliso, kunye ne-opinion esebenzayo esekelwe kwiintlobo ze-social media ezininzi kunokuba kwiintlobo ze-political-economic. Ngokutsho, kunokuba akukho ingxaki kwiintlobo ezininzi ze-intelligent-intelligent chatbots ezisetyenziswa Umbhali: Thomas Yin

The is an opinion piece based on the author’s POV and does not necessarily reflect the views of HackerNoon.

Read My Stories

Le audio iveliswe ngolwimi lokuqala lwebali!

Ukukhangisa i-"Innovation Against Safety" Doctrine kwi-AI Regulation

About Author

IZIMVO

ZIJONGE IIMPAWU

ELI NQAKU LINIKEZELWE KU

Related Stories

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps