Izisekelo Zobuhlakani Bokwenziwa (AI) Zabasebenzisi be-Linux – Ingxenye Yesibili

Izisekelo Zobuhlakani Bokwenziwa - Ingxenye Yokuqala: Imvelaphi

Yebo Uyithandile yini iposi yethu yokuqala egxile ekufundeni i-Artificial Intelligence (AI) yabasebenzisi be-Linux?Cishe ubusuvele ulinde ingxenye yesibili. Futhi uma ungakayifundi ingxenye yokuqala, sizocacisa ukuthi igxile ekuchazeni kafushane nangokuqondile ukuthi iyini i-AI kanye nezinhlobo ezahlukahlukene esezivele zidaliwe (ezincane noma ezibuthakathaka, ezikhiqizayo, kanye ne-Agent) kanye nalezo okulindeleke ukuthi zidalwe (ezijwayelekile noma eziqinile, kanye ne-Superintelligence). Kule ngxenye yesibili, sizogxila kakhulu ku indlela uBuhlakaniphi Bokwenziwa obuthuthuke ngayo ngokuhamba kwesikhathi futhi buqokomisa ezinye zezinyathelo ezibalulekile kakhulu ezinikeze impilo nokubaluleka kwale nsimu.

Kungani? Ngoba ngaphandle kokungabaza, futhi kungakhathaliseki ukuthi sisebenzisa luphi uhlelo lokusebenza, Ubuhlakani Bokwenziwa buwubuchwepheshe obungafikanga nje kuphela ukuze buhlanganiswe ngokujulile kukho konkekodwa, ngokusongela isikhathi namahora okusebenza, cishe, Kancane kancane, izosusa labo abangaqondi kahle ukuthi ikhona yini kanye nokuphathwa kwayo.Ngakho-ke, akukho okungcono kunokuqala ukuthola umongo wokuthi kuyini, ukuthi kwaqala kanjani, ukuthi kuthuthukiswe kanjani, nokuthi kusetshenziswe kanjani, phakathi kwezinye izinto eziningi.

Ubuhlakani Bokwenziwa buhlobene kakhulu nomhlaba we-Linux, ngakho-ke namuhla sikunikeza le ngxenye yokuqala ukuze ufunde kabanzi ngayo.

Kodwa, ngaphambi kokuba singene kulokhu "Ingxenye yesibili yensimu ethokozisayo nemangazayo yoBuhlakaniphi Bokwenziwa"Futhi ikakhulukazi mayelana nokuhlukaniswa kwayo kwamanje kanye neminye imicimbi ephawulekayo yomlando, sincoma ukuthi uhlole incwadi yangaphambilini neyamuva ehlobene kulolu chungechunge lwezincwadi, lapho eqeda ukuyifunda:

Ubuhlakani Bokwenziwa buyinsimu yesayensi yamakhompyutha ezinikele ekuthuthukiseni izinhlelo ezikwazi ukulingisa izinqubo zobuhlakani bomuntu, njengokucabanga, ukufunda, ukuqonda, kanye nobuciko. Konke lokhu kufezwa ngama-algorithms kanye nenani elikhulu ledatha, elilawulwa yimishini enamandla noma ihadiwe ethuthukisiwe/ekhethekile, ekwazi ukuhlaziya ulwazi, ukuxazulula izinkinga eziyinkimbinkimbi, ukwenza izinqumo ezizimele, kanye nokuthuthukisa ukusebenza kwazo ngokuhamba kwesikhathi.

Isihloko esihlobene:

Izisekelo Zobuhlakani Bokwenziwa (AI) Zabasebenzisi be-Linux - Ingxenye Yokuqala

Izisekelo Zobuhlakani Bokwenziwa - Ingxenye Yesibili: Imvelaphi

Ingxenye Yesibili: Ukuhlukaniswa Kobuhlakani Bokwenziwa Ngesikhathi

Nakuba kuyiqiniso ukuthi i-AI ingahlukaniswa ngezinga layo lamandla okubala noma ubuhlakani (Okwamanje: Kuncane noma Kubuthakathaka, Kukhiqiza Futhi Kunobuntu; Ikusasa: Okujwayelekile noma Okuqinile, kanye Nobunhloli Bokwenziwa), indlela ejwayelekile yokuhlukanisa ibe yile elandelayo: Ubuhlakani Bokwenziwa Obungokomfanekiso kanye Nobuhlakani Bokwenziwa BokuxhumanaNgakho-ke, ngezansi sinikeza imininingwane ethokozisayo ngakho.

Ubuhlakani Bokwenziwa Obungokomfanekiso

Ubuhlakani Bokwenziwa Obungokomfanekiso (“i-AI yakudala” noma i-GOFAI – I-AI Enhle Yakudala) I-AI Engokomfanekiso iyigatsha lesayensi yamakhompyutha elisekelwe esimisweni sokuthi ubuhlakani bungamelwa ngokuphathwa kwezimpawu nemithetho enengqondo. Ngokungafani nokufunda okujulile kwamanje, okufunda emaphethini enanini elikhulu ledatha, i-AI engokomfanekiso Yakhiwe "kusukela phezulu kuya phansi", yethula ulwazi lomuntu ngqo emshinini.

Izici ezivelele

Phakathi kwakhe izici ezivelele Ezinye izibonelo zifaka:

• Ukuzimela kwedatha yakhoAkudingi izigidi zezibonelo ukuze kufundwe; kudinga uchwepheshe ongumuntu ukuze achaze imithetho yomdlalo.
• Ukucabanga okunengqondo okusetshenzisiweIsebenzisa imithetho yokuphetha (uma u-A eyiqiniso, khona-ke u-B kungenzeka). Ithembele ekucabangeni okusemthethweni ukuze ifinyelele iziphetho.
• Uhlobo lokumelwa kolwaziIsebenzisa izimpawu (amagama, izinombolo, noma imiqondo) ukumela izinto zangempela kanye nobudlelwano bazo. Isibonelo: Is_A(Cat, Mammal).
• Ukufundeka kalula kwezinqubo zayo (Ibhokisi Elimhlophe)Ngokungafani namanethiwekhi e-neural anamuhla, "angamabhokisi amnyama," ukucabanga okungokomfanekiso kusobala ngokuphelele. Ungabona kahle ukuthi yimuphi umthetho owasetshenziswa ukuze kufinyelelwe impendulo.

Ubuhlakani Bokwenziwa Bokuxhumana

Uma i-AI engokomfanekiso "iwukucabanga", i-Connectionist Artificial Intelligence "iwukubhayoloji".Le ndlela iphefumulelwe yindlela ubuchopho bomuntu obucubungula ngayo ulwazi. Esikhundleni semithetho eqinile, ehlelwe ngesandla (njengokuthi "uma u-A, bese kuba u-B"), I-Connectionism isebenzisa amanethiwekhi amayunithi alula axhumene (abizwa ngokuthi ama-neuron okwenziwa) afunda ngokuhlangenwe nakho kanye nokulungiswa kwesignali.Ngakho-ke, kungamandla ashukumisayo ngemuva kwalokho esikwaziyo namuhla ngokuthi i-Deep Learning kanye namanethiwekhi e-neural.

Izici ezivelele

Phakathi kwakhe izici ezivelele Ezinye izibonelo zifaka:

• Isimo esikhulayo sezinqubo zasoUkuziphatha okuhlakaniphile "kuvela" ekusebenzisaneni kwamayunithi amaningi alula, hhayi ekuhlakanipheni okuhlanganisiwe.
• Ukufundeka kalula kwezinqubo zayo (Ibhokisi Elimnyama)Ngokungafani ne-AI engokomfanekiso, kunzima kakhulu ukuchaza kahle ukuthi kungani inethiwekhi yezinzwa yenze isinqumo esithile. Ulwazi "lusatshalaliswa" kuzo zonke izisindo zezinombolo eziyizigidi.
• Ukubekezelela Umsindo (Ukungalungi Nokungalungi)Banekhono lokuphatha idatha engaphelele noma engacacile. Uma inethiwekhi yokuxhumana ingenayo iphikseli esithombeni, cishe izoqhubeka nokuyibona into; uhlelo olungokomfanekiso lungase luhluleke ngenxa yokuntuleka komthetho oqondile.

Izinyathelo zomlando ekuHleleni, ekuHlaleni koBuhlakaniphi kanye naseRobhothi kusukela ngo-1925 kuya ku-2025: iminyaka eyi-100 yentuthuko

Ikhulu elidlule (1900)

iminyaka emi-50 yokuqala

• 1842Isazi sezibalo u-Ada Lovelace wayengowokuqala ukubona amandla amakhompyutha ngale kwezibalo.
• 1921UKarel Čapek, umbhali wemidlalo waseCzech, wethule umdlalo wakhe wesayensi oqanjiwe othi “Rossum's Universal Robots”, lapho ahlola khona umqondo wabantu bokwenziwa ayebabiza ngokuthi amarobhothi noma “i-robota” (isigqila).
• 1939Umsunguli kanye nesazi sesayensi uJohn Vincent Atanasoff, kanye noClifton Berry, bakha umshini wokuqala wekhompyutha wedijithali. Le khompyutha yayingahlelwa, kodwa yayingaxazulula izibalo eziqondile ezifika ku-29 ngesikhathi esisodwa, okwenza u-Atanasoff abe nguBaba weKhompyutha.
• 1943Abacwaningi base-US bethula imodeli yabo yama-neuron okwenziwa, abhekwa njengobuhlakani bokuqala bokwenziwa.
• 1949Idokhumenti ethi “Inhlangano yokuziphatha” ishicilelwe, eyasebenza njengesisekelo sokufunda ama-algorithms kumanethiwekhi e-neural okwenziwa (iNebrija University, n.d.).
• 1950Isazi sezibalo saseNgilandi kanye nomsunguli wesayensi yamakhompyutha u-Alan Turing ubuze lo mbuzo: "Ingabe imishini ingacabanga?" Esihlokweni sakhe esithi "Computing Machinery and Intelligence," uTuring waklama lokho okwaziwa ngokuthi iTuring Test, noma umdlalo wokulingisa, ukuze anqume ukuthi umshini uyakwazi yini ukucabanga.

Iminyaka emi-50 edlule

• 1956UJohn McCarthy, uprofesa eDartmouth College, wahlela i-workshop yasehlobo ukuze acacise futhi athuthukise imibono mayelana nemishini yokucabanga, futhi wakhetha igama elithi "Artificial Intelligence" ngenxa yayo. Ingqungquthela yaseDartmouth, ebhekwa kabanzi njengesikhathi esiyisisekelo se-Artificial Intelligence (AI) njengomkhakha wocwaningo, yayihlose ukuthola "indlela yokuthola imishini ukuthi isebenzise ulimi, yakhe izifinyezo kanye nemibono, ixazulule izinkinga manje ezigcinelwe abantu, futhi izithuthukise."
• 1961I-General Motors ifaka irobhothi lokuqala lezimboni elizothatha indawo yabantu emisebenzini yokuhlanganisa.
• 1964UJoseph Wizenbaum usungula uhlelo lokuqala lwekhompyutha lokucubungula ulimi lwemvelo, olulingisa ingxoxo nabantu.
• 1965I-DENDRAL, uhlelo lokuqala lochwepheshe olugxile ekuhlaziyweni kwamakhemikhali ama-molecule, luyakhishwa.
• 1966: Irobhothi lokuqala elihambayo elisebenza ngezinhloso ezijwayelekile elikwazi ukucabanga ngezenzo zalo liqala ukubonakala emakethe.
• 1966: I-ELIZA, i-chatbot yokuqala ekwazi ukulingisa ingxoxo yomuntu, yembulwa; kanye ne-Shakey, irobhothi lokuqala elihambayo elizimele elinohlelo lwemisebenzi.
• 1967Umshini wePerceptron Mark 1 wakhiwe, owaziwa ngokuthi ukuvela kwenethiwekhi ye-neural kaRosenblatt kuhadiwe ekhethekile.
• 1968: I-Mac Hack, uhlelo lokuqala lwe-chess ukufinyelela ezingeni lokuncintisana le-Class C.
• 1969Incwadi ethi "Perceptrons" ishicilelwe, ukugxekwa kwezibalo okwamisa ucwaningo kumanethiwekhi ezinzwa iminyaka eminingi.
• 1972I-WABOT-1, irobhothi lokuqala emhlabeni elinobuntu, elakhiwe eNyuvesi yaseWaseda, lakhiwa.
• 1972: Isofthiwe ye-MYCIN yakhiwe, uhlelo lochwepheshe lokuxilongwa kwezifo zegazi kanye nemithi kadokotela.
• 1974: Ubusika bokuqala be-AI buqala, buchazwa njengenkathi yokuncipha okukhulu kwezimali kanye nenzalo kulandela uMbiko we-Lighthill.
• 1976I-PROLOG, ulimi lokuhlela olusekelwe ku-logic esemthethweni, iyathuthukiswa.
• 1980: XCON / R1, uhlelo lokuqala olukhulu lochwepheshe bezentengiselwano olusetshenziswa yi-DEC.
• 1981Isizukulwane Sesihlanu, iphrojekthi enkulu evela eJapane yokuhola i-AI ngokusebenzisa ikhompyutha ehambisanayo.
• 1982Amanethiwekhi aseHopfield, ukwethulwa kwamanethiwekhi e-neural anememori ehlangene esekelwe ku-physics.
• 1984: Iphrojekthi ye-CYC, Ukuqala kokudalwa kwesisekelo solwazi olujwayelekile olujwayelekile.
• 1984I-WABOT-2, irobhothi elikwazi ukufunda umculo nokudlala isitho sikagesi.
• 1985: I-AARON, uhlelo lokudweba oluzimele lobuciko oluboniswe engqungqutheleni ye-AAAI.
• 1986: Ukusabalalisa emuva, ukwenziwa kwaziwe kwe-algorithm eyisisekelo yokuqeqesha amanethiwekhi ajulile.
• 1986Izimoto Ezizimele: U-Ernst Dickmanns ubonisa imoto yokuqala eyirobhothi ekwazi ukuhamba ngesivinini esingu-55 mph.
• 1987Ubusika Besibili, obubonakala ngenkinga entsha yokuzethemba kanye nokuwa emakethe yehadiwe ekhethekile ye-AI.
• 1992: TD-Gammon, uhlelo lwe-backgammon olufinyelele ezingeni lochwepheshe ngokufunda ukuqinisa.
• 1997I-Deep Blue, ukunqotshwa okungokomlando kwengqwele yomhlaba ye-chess uGarry Kasparov.
• 1997: Ngokwemvelo, isofthiwe yokuqala yokuqaphela inkulumo eqhubekayo yezentengiselwano.
• 1998: URobot Kismet, iNgcali ekuxhumaneni ngokomzwelo phakathi komuntu nerobhothi.
• 1999I-AIBO, irobhothi lokuqala labathengi elinamakhono okufunda.

Isihloko esihlobene:

Amaphrojekthi we-Artificial Intelligence 2023: Amahhala, amahhala futhi avulekile

Ikhulu lamanje (2000)

Ishumi leminyaka lokuqala

• 2000I-Kismet, ikhanda lerobhothi elikwazi ukubona nokudala kabusha imizwa yabantu kanye nezinkomba zenhlalo, yaklanywa. Kwakuwukuhlolwa kwerobhothi yezenhlalo kanye nokubala okunemizwa okwakubonisa amadivayisi okufaka alingisa ukuguquguquka kokubona komuntu, ukuzwa, kanye nokwakheka komzimba.
• 2002: I-Roomba, i-vacuum cleaner yokuqala yerobhothi ekhiqizwe ngobuningi ethengiswa yi-iRobot, ekwazi ukuhamba nokuhlanza ikhaya, yethulwa emakethe.
• 2004I-DARPA Grand Challenge, umncintiswano wokuqala omkhulu wezimoto ezizimele endaweni ehlane, uyaqhubeka.
• 2006Umklomelo weNetflix wavela, umncintiswano owathuthukisa ama-algorithms okuncoma ngokusekelwe kudatha enkulu.

Ishumi leminyaka lesibili

• 2011I-Apple ihlanganisa i-Siri, umsizi obonakalayo one-interface yezwi, ku-iPhone 4S yayo. Uhlelo lwe-IBM lwe-Watson luyavela, luwina umklomelo wokuqala ohlelweni oludumile lwemibuzo lwethelevishini lwaseMelika i-Jeopardy! Futhi i-Google Brain, iphrojekthi enkulu yenethiwekhi yezinzwa ekwazi ukubona amakati ku-YouTube ngaphandle kokugadwa, iyathuthukiswa.
• 2013I-DeepMind Atari yathuthukiswa, i-AI ekwaziyo ukuba yingcweti emidlalweni yakudala ngokufunda okujulile kokuqinisa.
• 2014Uhlelo lokuqala lwekhompyutha luphumelela iTuring Test ngokukholisa ingxenye yesithathu yabahluleli ababambe iqhaza kulolu cwaningo ukuthi lungumuntu.
• 2015Ubuchwepheshe obuvulekile njengeTensorFlow kanye nePyTorch buyethulwa, okuqalisa ukwenziwa kwentando yeningi kwe-AI ngokusebenzisa imitapo yolwazi evulekile yocwaningo olukhulu.
• 2016Lolu hlelo, olusekelwe kunethiwekhi ejulile yezinzwa, luhlula uLee Sodol, iqhawe lomhlaba le-Go, emidlalweni emihlanu.
• 2016I-AlphaGo, ikhompyutha ye-AI, yathuthukiswa eyakwazi ukufeza ukunqotshwa okungokomlando kweqhawe le-Go, ikhombisa ukuqonda okungokweqhinga okwenziwa.
• 2016USophia, irobhothi elifana nomuntu elikwaziyo ukushukumisa nokuphatha izingxoxo ezilula, wembulwa.
• 2017"I-Transformer" yakhiwe, isakhiwo sokunaka esivumele ukucutshungulwa kolimi okukhulu.
• 2017I-AlphaZero yathuthukiswa, i-AI efunde i-chess, i-Shogi ne-Go ngamahora, idlula izinhlelo ezikhethekile.
• 2018I-BERT, imodeli yolimi lwe-Google eyashintsha ukuqonda komongo, yasungulwa.
• 2019I-GPT-2, imodeli yokuqala yokubhala eyinhloko enekhono lokubhala kahle, yembulwa.
• 2019I-AlphaStar iveziwe, i-AI ekwazile ukufinyelela ezingeni le-Grandmaster ku-StarCraft II, iphatha ulwazi olungaphelele.

Iminyaka emi-5 edlule

• 2020Inkampani i-“General Motors” ifaka irobhothi lokuqala lezimboni elizothatha indawo yabantu emisebenzini yokuhlanganisa.
• 2020I-GPT-3, imodeli yokuqala enkulu yolimi ekwazi ukukhiqiza umbhalo ongahlukaniseki nombhalo womuntu, imenyezelwe.
• 2021I-AlphaFold 2, isofthiwe egxile ekuxazululeni ukugoqwa kwamaprotheni, imenyezelwe, eye yakwazi ukusheshisa ezokwelapha ngamashumi eminyaka.
• 2021I-DALL-E, i-AI eyaqala ukukhiqizwa kwezithombe zobuciko ngolimi lwemvelo, ikhishwa emphakathini.
• 2021I-GitHub Copilot ikhishwe, i-AI ehlanganiswe nokuthuthukiswa kwesofthiwe ebhala ikhodi ngokuzenzakalelayo.
• 2022I-Stable Diffusion, i-AI eyaqala ukwenziwa kwezithombe ngentando yeningi ngomthombo ovulekile, ikhishwa emphakathini.
• 2022I-ChatGPT yakhiwe, i-AI eqalwe ngayo yathandwa kakhulu yi-Generative AI emhlabeni wonke kuzo zonke izinhlaka.
• 2023I-GPT-4, imodeli yokuqala ye-multimodal ene-akiling ethuthukisiwe kanye nenkumbulo yomongo enwetshiwe, imenyezelwe.
• 2023: Le nkampani iMeta ikhipha i-Llama, i-AI ekwethulweni kwayo kuqalisa umncintiswano wamamodeli olimi oluvulekile anamandla.
• 2023: I-Gemini 1.0 Ultra yembulwe, imodeli yokuqala ephumelele kakhulu kochwepheshe babantu ekuhlolweni okukhulu kwe-multimodal.
• 2023I-Silent Eight iyethulwa, iphrojekthi enkulu yokuthunyelwa kwe-AI yokuthobela imithetho yezezimali kanye nokutholakala kobugebengu ngezinga elikhulu.
• 2024I-Sora, imodeli ye-OpenAI ekwazi ukukhiqiza amavidiyo angokoqobo angafika kumzuzu owodwa ngombhalo, ikhishwa emphakathini.
• 2024Iphrojekthi yesishayamthetho ethi "EU AI Act" imenyezelwe, umthetho wokuqala ophelele emhlabeni wonke wokulawula ukuthuthukiswa nokusetshenziswa kwe-AI.
• 2024I-Stability AI yembule i-StableLM 2, imodeli yolimi olukhulu oluvulekile oluthuthukisiwe, kanti i-Google ikhiphe i-Gemini 1.5 ku-beta elinganiselwe.
• 2024Imodeli ye-OpenAI o3 ikhishwe, imodeli ethuthukisiwe enekhono lokucabanga elijulile kanye namakhono okuphetha anengqondo.
• 2025I-OpenAI ikhipha i-GPT-5, i-Google yethula i-Gemini 2.0 kanye ne-Nano Banana.
• 2025Imodeli ye-DeepSeek evulekile ye-AI iyembulwa, ivula indlela yamanye amamodeli avulekile anamandla futhi angabizi kakhulu.
• 2025Ukuthuthukiswa kwe-AI PC Gen 2 kuvela, okubonisa ukuhlanganiswa kwamakhompyutha nama-chip e-NPU ukuze kusebenze i-AI yendawo ethuthukisiwe.
• 2025Lokhu kuphawula ukuqala kwenkathi eyaziwa ngokuthi "Abakhi Be-Agent", ebonakala ngokufinyelela ochungechungeni lwamapulatifomu ajwayelekile ukudala ama-ejenti azimele akwazi ukwenza izinqubo zebhizinisi.

Isihloko esihlobene:

Ukusatshalaliswa kwe-GNU/Linux ngobuchwepheshe obuhlanganisiwe be-Artificial Intelligence obuhle kakhulu ukuzama ngo-2026

Isifingqo

Ngamafuphi, sithemba ukuthi lokhu "Ingxenye yesibili yensimu ethokozisayo nemangazayo yoBuhlakaniphi Bokwenziwa" ngokugcizelela ku- I-Generative Artificial Intelligence (GAI)Sengathi ngabe inikele kakhulu ekuqiniseni nasekukhuliseni ulwazi lwakho namakhono kulo mkhakha wesimanje nowezinto ezintsha. Ngendlela yokuthi ikuvumele ukuthi kancane kancane ukuze kuhlangatshezwane nabanye asebethuthukile kakade maqondana nolwazi nokuqeqeshwaOkusho ukuthi, ekusetshenzisweni kwazo kwangempela. Maduze nje, sizoqhubeka nokuthunyelwe okwengeziwe kulolu chungechunge ukuze abasebenzisi abaningi nabafundi abanothando bendawo yonke ye-Linux bakwazi ukufunda kabanzi nge-AI ngokujwayelekile, ikakhulukazi nge-Generative AI.

Okokugcina, khumbula vakashela wethu «ikhasi lasekhaya» ngeSpanishi. Noma, nganoma yiluphi olunye ulimi (ngokwengeza nje izinhlamvu ezi-2 ekupheleni kwe-URL yethu yamanje, isibonelo: ar, de, en, fr, ja, pt kanye no-ru, phakathi kwezinye eziningi) ukuze ufunde okuqukethwe kwamanje okwengeziwe. Ukwengeza, sikumema ukuthi ujoyine yethu Isiteshi socingo esisemthethweni ukufunda nokwabelana ngezindaba eziningi, imihlahlandlela kanye nezifundo ezivela kuwebhusayithi yethu.