UHlakahlisombulu

Mayelana Wikipedia

Artificial_Intelligence,_AI

uHlakahlisombulu ( AI ) ubuhlakani obuboniswa izinguxa, obuphambene nobuhlakani bemvelo obuboniswa izilwane nabantu. Ucwaningo ngohlakahlisombulu luchasiswa njengomkhakha wemfundo yezinto ezihlakaniphile, okubhekisela kunoma isiphi isimiso esibona indawo esizungezile futhi esithatha isinyathelo esizokwandisa amathuba aso okufinyelela imigomo yaso.

Igama elithi "uHlakahlisombulu" ngaphambilini lalisetshenziselwa ukuchazisa izinguxa ezilingisa amakhono "omuntu" okuqonda ayamaniswa nengqondo yomuntu, afana "nokufunda" kanye "nokuxazulula izinkinga". Abacwaningi abakhulu bohlakahlisombulu sebayinqabile lencasiselo futhi manje bachazisa uhlakahlisombulu ngokwenhluzeko nokwenza Izinto ezihluzekile, okuyincaziso enweba ukuchazwa kobuhlakani. [lower-alpha 1]

Izinhlelozokusebenza zohlakahlisombulu zihlanganisa izinjini zokuhlwaya ezithuthukisiwe zolwebu (isb., Google ), izinhlelo zezincomo (ezisetshenziswa i- YouTube, i-Amazon ne- Netflix ), ukuqonda inkulumo yomuntu (njenge- Siri ne- Alexa ), izimoto ezizishayelayo (isb., Tesla ), ukwenza izinqumo okuhlelelekiwe kanye nokuncintisana ezingeni eliphezulu kakhulu ezinhlelweni zomfeketho wamaqhinga-su (ezifana noThimba nomHaqo ). [2] Njengoba izinguxa ziba nekhono ngokwengeziwe, imisebenzi ebhekwa njengedinga "ubuhlakani" ivamise ukususwa encazelweni yohlakahlisombulu, into eyaziwa ngokuthi umphumela wohlakahlisombulu[3]. [4]

uHlakahlisombulu lwasungulwa njengomkhakha wezemfundo ngo-1956, futhi eminyakeni kusukela lapho uye wabhekana namagagasi ambalwa ethemba, [5] [6] okulandelwa ukudumala nokulahlekelwa uxhaso (okwaziwa ngokuthi " ubusika bohlakahlisombulu "), [7] [8] okwalandelwa izindlela ezintsha, impumelelo kanye noxhaso oluvuselelwe. [6] [9] Ucwaningo lohlakahlisombulu luye lwazama futhi lwalahla izindlela eziningi ezihlukene kusukela ekusungulweni kwalo, okuhlanganisa ukulingisa ubuchopho, ukufanisela ukuxazulula izinkinga zomuntu, ukucabanga okuhlelekile, isizindalwazi esikhulu solwazi kanye nokulingisa ukuziphatha kwezilwane.

  1. Stuart Russell and Peter Norvig characterize this definition as "thinking humanly" and reject it in favor of "acting rationally".[1]

Ezinganekweni[hlela | Hlela umthombo]

Igama elithi "robot" (IsiThununu) ngokwalo laqanjwa nguKarel Čapek emdlalweni wakhe we-1921 othi RUR.

Izidalwa mbulu ezikwazi ukucabanga ziye zavela njengezilekeleli zokuxoxa izindaba kusukela ezikhathini zasendulo, futhi bezilokhu ziyisihloko esiphikelelayo ezinganekweni zesayensi . [3]

Umgqigqo ovamile kule misebenzi waqala nge-Frankenstein kaMary Shelley, lapho umuntu obunjiwe eba usongo kubabumbi bakhe. Lokhu kuyavela nasemisebenzini ofana ne2001: A Space Odyssey ka-Arthur C. Clarke no-Stanley Kubrick (bobabili ngo-1968), futhi kwenziwa nayi-HAL 9000, isiCikizi esibulalayo esilawula iNdizamkhathi ebizwa Discovery One, kanye naku -The Terminator (1984) kanye ne -The Matrix (1999) ). Ngokuphambene, izithununu ezithembekile ziyivelakancane njengoGort ovela ku-The Day the Earth Stood Still (1951) kanye noBishop waku-Aliens (1986). [10]

U- Isaac Asimov wethula Imithetho Emithathu Yezithununu ezincwadini nasezindabeni eziningi aziloba, ikakhulukazi kuchunge lwe-"Multivac" olumayelana nesicikizi esinobuhlakani obuphezulu esibizwa ngalelogama. Imithetho ka-Asimov ivame ukuvela uma kunezingxoxo zemicondozo yezinguxa; [11] nakuba cishe bonke abacwaningi bohlakahlisombulu bejwayelene nemithetho ka-Asimov, ngokuvamile babheka leimithetho njengento engenamsebenzi ngezizathu eziningi, esinye sazo ukuthinziyela kwayo. [12]

I-Transhumanism (ukuhlanganiswa kwabantu nezinguxa) iyahlwaywa ku- Ghost in the Shell ye-manga kanye nakwilando lwenganeko yesayensi i-Dune .

Kunemisebenzi embalwa esebenzisa uhlakahlisombulu ukuze isiphoqe ekubhekaneni nemibuzo eyisisekelo yakuthi kuyini okusenza sibe ngabantu, ngokusibonisa izidalwa mbulu ezikwazi ukuba nemizwa, futhi kanjalo zibhekane nokuhlupheka. Lokhu kuyabonakala ku-RUR ka- Karel Čapek, nemibukelo i-AI Artificial Intelligence kanye ne-Ex Machina, kanye nasencwadini ethi Do Androids Dream of Electric Sheep (ingabe izimomothu ziphupha Ngezimvu Zamazuba?), kaPhilip K. Dick . U-Dick ubheka umbono wokuthi ukuqonda kwethu ukuzithoba komuntu kuguqulwa ubuchwepheshe obudalwe ngohlakahlisombulu. [13]

Umlando[hlela | Hlela umthombo]

I- didrachma yecwebe yaseCrete eveza u- Talos, i- automaton yasendulo kwezinganeko enobuhlakani mbulu

Izidalwa mbulu ezinobuhlakani zavela njengezinsiza zokuxoxa izindaba endulo, [14] futhi zaziwulwangu lwezinganeko, njengaku- Frankenstein kaMary Shelley noma i- RUR kaKarel Čapek. [15] Laba balingisi kanye neziphetho zabo baphakamisa izingqinamba ezifanayo nalezo okuxoxwa ngazo emicondozweni yohlakahlisombulu . [16]

Ukufundwa kokuhluza Izinto kwenguxa noma "okuhlelekile" kwaqala ngezazi zefilosofi nezibalo endulo. Ukucwaninga umhluzamqondo wezibalo kwaholela ngokuqondile kumbono wokucikiza ka- Alan Turing, owasikisela ukuthi inguxa, ngokushudula izimpawu ezilula njengokuthi "0" kanye "no-1", ingafefenyeka noma yisiphi isenzo esingacatshangwa seziphumo zezibalo. Lo mbono wokuthi iziCikizi zezibhangqiwe zingafefenyeka noma iyiphi inqubo yokuhluza okuhlelekile waziwa ngokuthi isihlawumbiselo sika-Church noTuring . [17] Sona, kanye nokuvubukulwa kwe- neurobiology, umbono wokwaziswa, kanye ne- cybernetics, kwaholela abacwaningi ukuba bacabangele ukuthi kungenzeka yini ukwakha ubuchopho be-elekthronikhi. [18]Umsebenzi wokuqala manje owaziwa ngokuthi uhlakahlisombulu kwakungumklamo ihlelekile "weziNzwa ezimbulu" ze-Turing complete owenziwa nguMcCullouch no- Pitts ngowe-1943. [19]

  1. Russell & Norvig (2021), p. 2.
  2. Google (2016).
  3. 3.0 3.1 McCorduck (2004).
  4. Schank (1991).
  5. Crevier (1993), p. 109.
  6. 6.0 6.1 Funding initiatives in the early 80s: Fifth Generation Project (Japan), Alvey (UK), Microelectronics and Computer Technology Corporation (US), Strategic Computing Initiative (US):
  7. First AI Winter, Lighthill report, Mansfield Amendment
  8. Second AI Winter:
  9. Clark (2015b).
  10. Buttazzo (2001).
  11. Anderson (2008).
  12. McCauley (2007).
  13. Galvan (1997).
  14. AI in myth:
  15. McCorduck (2004), pp. 17–25.
  16. McCorduck (2004), pp. 340–400.
  17. Berlinski (2000).
  18. AI's immediate precursors:
  19. Russell & Norvig (2009).