Menene hankalin tarayya? Hankalin tarayya shine ikon ƙungiyoyi na mutane — mutum, na'ura, ko duka — suyi aiki tare a hanyoyi da suka bayyana a matsayin masu hankali, suna samar da hukunci, yanke shawara, da hanyoyin magance matsaloli da zasu iya wuce abin da kowanne memba zai iya cimma shi kadai.
Hankalin tarayya, kamar yadda Thomas Malone da Michael Bernstein suka bayyana a cikin Handbook of Collective Intelligence (MIT Press, 2015), shine 'ƙungiyoyi na mutane suna aiki tare a hanyoyi da suka bayyana a matsayin masu hankali.' Fannin yana rufe daga Ka'idar Juri na Condorcet (1785) da gwajin nauyin sani na Francis Galton na 1906 zuwa kimiyyar dakin gwaje-gwaje na zamani: Anita Williams Woolley da abokan aikinta (Science, 2010) sun gano shaidar wani babban hankalin tarayya — c factor — wanda ya bayyana kusan 43% na bambancin a cikin aikin ƙungiya a cikin ayyuka kuma ya fi danganta da jin dadin zamantakewa, daidaito na juyin magana, da kuma adadin mata a cikin ƙungiyar fiye da IQ na memba. An tattauna c factor (Credé da Howardson, 2017) kuma an goyi bayan shi da wani meta-bincike na karatun 22 da ƙungiyoyi 1,356 (Riedl et al., PNAS, 2021). Genome na Hankalin Tarayya (Malone, Laubacher da Dellarocas, 2010) yana zana ginshiƙan ƙungiyoyi masu hankali a kan Wa, Me yasa, Menene, da Yadda. Binciken MIT Center for Collective Intelligence na yanzu (2026) yana mai da hankali kan AI mai haɓaka da hankalin tarayya, Tsarin Supermind, da tsara ƙungiyoyin mutum-AI; wani meta-bincike na 2024 (Vaccaro, Almaatouq da Malone, Nature Human Behaviour) ya gano cewa haɗin gwiwar mutum-AI ba ya wuce mafi kyawun kowanne daga cikin su kadai — gaske haɗin gwiwa yana buƙatar tsari na yanke shawara da mu'amala. Tsarin CCI na yanzu sun haɗa da Supermind Ideator, wanda ke rufe babban samfurin harshe a cikin tsarin warware matsaloli na kirkira, da gasar PNAS ta 2026 da ta nazarci fiye da 180,000 na tattaunawar AI zuwa AI. A cikin tsarin yanke shawara — Attend, Frame, Generate, Investigate, Aggregate, Deliberate, Allocate, Choose, Implement, Monitor — binciken hankalin tarayya yana mai da hankali kan matakan Aggregate da Deliberate: yadda ƙungiyoyi ke tara ilimi da aka rarraba da kuma yadda suke tunani tare.
Menene Hankalin Tarayya?
Hankalin tarayya shine ikon ƙungiyoyi — mutum, na'ura, ko duka — suyi aiki a hanyoyi da suka bayyana a matsayin masu hankali. Ba wani misali bane: bincike ya auna shi, ya tattauna shi, kuma ya nuna yadda za a tsara don hakan. Wannan jagorar tana rufe kimiyyar daga Condorcet (1785) zuwa binciken ƙungiyoyin mutum-AI na MIT (2026).
An sabunta ƙarshe: 2026-07-18
TL;DR
Ƙungiyoyi na iya zama masu hankali a auna. Woolley et al. (Science, 2010) sun gano c-factor na hankalin tarayya (c) wanda ke hasashen aikin ƙungiya a cikin ayyuka — wanda ya fi tasiri daga jin dadin zamantakewa, daidaito na juyin magana, da tsari fiye da IQ na memba. An tattauna wannan ganowa (Credé & Howardson, 2017) amma an goyi bayan shi da meta-bincike na ƙungiyoyi 1,356 (Riedl et al., PNAS, 2021). Hanyar zamani shine hankalin tarayya na mutum-AI: meta-binciken MIT na 2024 ya nuna cewa ƙara mutum a cikin tsarin ba ya taimaka kai tsaye ga tsarin AI — haɗin gwiwa yana dogara ne akan tsari na yanke shawara da tsari. Argumentree yana gina wannan darasi: tsara tunani, tara hukunci, a bar mutane su yanke shawara.
Fassara Hankalin Tarayya
Fassarar ilimi ta yau da kullum ta fito daga Thomas Malone da Michael Bernstein's Handbook of Collective Intelligence (MIT Press, 2015): hankalin tarayya shine "ƙungiyoyi na mutane suna aiki tare a hanyoyi da suka bayyana a matsayin masu hankali." Fassarar tana da faɗi sosai. Yana rufe kasuwa tana sa farashi, al'umma ta Wikipedia tana rubuta encyclopaedia, juri yana auna shaidar, ƙungiyar ants tana nemo hanyar gajere zuwa abinci — kuma, a ƙara, ƙungiyoyi masu haɗaka na mutane da tsarin AI.
Hankalin tarayya yana da faɗi fiye da hikimar taron jama'a, wanda ke shafar tarin kididdiga na hukunci da yawa masu zaman kansu. Hankalin tarayya kuma yana haɗa haɗin gwiwa da tattaunawa — ƙungiyoyi da ke mu'amala, suna jayayya, da gina kan juna. Yana bambanta da hankalin tarin, inda haɗin kai ke tasowa daga ƙa'idodi masu sauƙi na gida ba tare da tattaunawa ba. Kuma yana haɗa da hanyoyin da aka nazarci a fannonin makwabta: tsarin zaɓe suna tara abubuwan da suka fi so, kasuwannin hasashen suna tara ra'ayoyin yiwuwar, da tsari mai tsari yana tara dalilai — amsoshi guda uku daban-daban ga tambaya guda ɗaya na yadda za a juya tunani da yawa zuwa yanke shawara guda ɗaya.
A cikin tsarin yanke shawara — Attend, Frame, Generate, Investigate, Aggregate, Deliberate, Allocate, Choose, Implement, Monitor — binciken hankalin tarayya yana mai da hankali kan matakan tara da tattaunawa: yadda ilimi da aka rarraba ke taruwa, da yadda ƙungiyoyi ke tunani tare ba tare da cututtukan cascades na bayani da tunanin ƙungiya.
Tarihin Gajere na Binciken Hankalin Tarayya
Fannin yana da tushe a cikin lissafi na Enlightenment, kididdiga, tattalin arziki, ilimin halittu, da kimiyyar kwamfuta. Hanyar: a ƙarƙashin yanayi masu kyau, ƙungiyoyi suna wuce membobinsu — kuma ana iya tsara yanayin.
Marquis de Condorcet ya tabbatar da cewa idan kowanne mai kada kuri'a ya fi kadan fiye da sa'a kuma yana kada kuri'a a cikin 'yanci, yiwuwar cewa mafi yawan mutane suna daidai yana kusantar tabbaci yayin da rukuni ya karu — sakamakon lissafi na asali na hankali na tarayya.
Francis Galton ya nazarci hasashen nauyin shanu 787 a wani biki na ƙasar 1906. Hasashen tsaka-tsaki (1,207 lb) ya kasance cikin 1% na nauyin gaskiya (1,198 lb) — fiye da ƙwararrun shanu. An buga a Nature.
Friedrich Hayek's "Amfani da Ilimi a cikin Al'umma" (American Economic Review) yana jaddada cewa kasuwanni suna tara ilimi mai yaduwa, na gida ta hanyar farashi — tushen tunani na kasuwannin hasashen.
James Surowiecki ya tsara sharuɗɗan hudu don hikimar taron (bambanci, 'yanci, rarrabawa, tarawa). A wannan shekarar, Hong da Page sun buga samfurin su na "bambanci yana wuce ƙwarewa" a PNAS.
Woolley, Chabris, Pentland, Hashmi, da Malone (Science) sun sami shaidar wani babban abu na hankali na tarayya a cikin mutane 699 suna aiki a cikin ƙananan rukuni — wani misali na matakin rukuni na IQ na mutum.
Malone, Laubacher, da Dellarocas (MIT Sloan Management Review) sun lissafa abubuwan gini — "genes" — na tsarin hankali na tarayya tare da tambayoyi hudu: Wa, Me yasa, Me, da Yaya.
Jagorar Malone da Bernstein na MIT Press yana haɗa fannin da ma'anar sa ta al'ada: ƙungiyoyin mutane suna aiki tare a hanyoyi da suka bayyana a matsayin masu hankali.
Superminds na Thomas Malone yana bayyana nau'ikan hudu na hankali na tarayya — tsarin mulki, dimokuradiyya, kasuwanni, al'umma, da tsarin halittu — kuma yana tambayar yadda kwamfutoci zasu iya sa kowanne ya zama mai hankali.
Riedl, Kim, Gupta, Malone, da Woolley (PNAS) sun nazarci karatun 22 tare da rukuni 1,356 kuma sun sami shaidar wani abu na hankali na tarayya a cikin dalibai, sojoji, 'yan wasa, da ma'aikatan kan layi.
Vaccaro, Almaatouq, da meta-analysis na Malone na gwaje-gwaje 106 (Nature Human Behaviour) sun gano haɗin gwiwar mutum-AI a matsakaita sun yi mummunan aiki fiye da mafi kyawun mutum ko AI kadai — haɗin gwiwa dole ne a tsara, ba a ɗauka ba.
Burton da masu rubutu 27 (Nature Human Behaviour) sun tsara yadda manyan samfuran harshe zasu iya inganta hankali na tarayya — da yadda suke haɗarin daidaita ra'ayoyi da ƙirƙirar ƙarya.
Cibiyar MIT ta Hankali na Tarayya tana bin manyan shafuka guda uku: AI mai haɓaka da Hankali na Tarayya, Tsarin Supermind, da Tsara Ƙungiyoyin Mutum-AI — ciki har da binciken PNAS na tattaunawar AI-to-AI sama da 180,000.
C Factor: Shin Ƙungiya na Iya Samun IQ?
A cikin 2010, Anita Williams Woolley da abokan aikinta sun tambayi tambaya mai sauƙi tare da karni na ilimin mutum a bayanta: kamar yadda wani babban factor (g) ke hasashen aikin mutum a cikin ayyuka da yawa, shin akwai babban factor don ƙungiyoyi? Sun gwada mutane 699 suna aiki a cikin ƙungiyoyi guda biyu zuwa biyar akan nau'ikan ayyuka masu yawa (Science, 330, 686–688). Amsar ita ce eh: wani babban factor na kididdiga — c — ya bayyana kusan 43% na bambancin a cikin aikin ƙungiya a cikin ayyuka.
Menene ke hasashen c na ƙungiya — da abin da ba ya yi
IQ na memba ba ya da mahimmanci
Dukkan matsakaicin da mafi girman hankali na mutum na membobin rukuni sun kasance suna da alaƙa mai rauni da c. Daukar mafi basira ba ya, da kansa, sa rukuni ya zama mai hankali.
Mahimmancin jin dadin zamantakewa
Matsakaicin maki na rukuni a gwajin Karanta Zuciya a Idanu — ikon fahimtar halin tunani daga alamun fuska — ya kasance mai hasashen mahimmanci na c (r = 0.26).
Mahimmancin jujjuyawar juna
Rukuni wanda membobinsu suka bayar da gudummawa daidai ga tattaunawa suna da c mafi girma; bambancin a cikin jujjuyawar magana yana da alaƙa da rauni tare da c (r = −0.41). Rukuni da aka mamaye da murya ɗaya ko biyu sun kasance suna da hankali na tarayya ƙasa.
Tasirin haɗin gwiwa
A cikin binciken 2010, rukuni tare da mafi yawan mata sun sami maki mafi girma a kan c (r = 0.23) — tasirin da marubutan suka haɗa shi da kididdiga da jin dadin zamantakewa na mata, wato don tsari, ba don demography ba.
Shin c factor na gaskiya ne? Tattaunawar maimaitawa
Kamar yawancin abubuwan da suka shafi ilimin tunani, c an gwada shi — kuma masu karatu suna da hakkin sanin gaskiyar shaidar:
Tatsuniya (2017)
Credé da Howardson (Journal of Applied Psychology) sun sake nazarin samfuran shida kuma sun yi jayayya cewa goyon bayan kimiyya don wani babban abu guda ɗaya yana da rauni — cewa aikin rukuni na iya zama mafi kyau a bayyana ta hanyar ƙwarewar takamaiman aiki fiye da c guda ɗaya.
Amsar (2018)
Woolley, Kim, da Malone sun amsa cewa tattaunawar ta dogara ne akan hanyoyin scoring da tunanin kwaikwayo da ba su dace da ayyukan da rukuni ke aiwatarwa ba.
Binciken meta (2021)
Riedl, Kim, Gupta, Malone, da Woolley (PNAS) sun tara karatun 22 da suka shafi rukuni 1,356 — dalibai, sojoji, ma'aikatan kan layi, 'yan wasa — kuma sun sami shaidar c a cikin dukansu, tare da tsarin haɗin gwiwa da jin dadin zamantakewa na memba a matsayin mafi ƙarfi.
Karshe: babban meta-bincike yana goyon bayan kasancewar wani factor na hankalin tarayya da za a iya auna, yayin da tsarinsa da gama-gamawarsa ke ci gaba da tattaunawa. Abin da ba a tattauna ba shine darasin aikace-aikace: tsarin ƙungiya — wanda aka ji, yadda aka haɗa hukunci — yana hasashen aikin ƙungiya fiye da basirar memba.
Kididdiga vs. Taron Mu'amala
Akwai nau'ikan hankalin tarayya guda biyu, kuma suna gudana akan mai jujjuyawa daban-daban:
Tattalin arziki na taron
Yawan kimantawa masu zaman kansu, an tara su ta hanyar na'ura — matsakaici, kada kuri'a, farashin kasuwa. Wannan shine adadin Galton na adadi da duk kasuwar hasashen tun daga lokacin. Nasara tana dogara ne akan 'yanci da bambanci; aikin yana yawan zama na lokaci guda.
Taron hulɗa
Mutane suna shafar juna don gina fahimtar juna — tattaunawa, tunani, haɗin gwiwa. Wannan shine Wikipedia, al'ummar kimiyya, ƙungiyar zane. Nasara tana dogara ne akan ingancin sadarwa da tsaro don magana; aikin yana ci gaba.
Nau'ikan biyu suna jan hankali a cikin hanyoyi masu sabani: zaman kansa yana taimakawa taron kididdiga amma yana gaji taron mu'amala. Ƙungiya da ba ta taɓa magana ba na iya tara kyau amma ba ta gina komai sabo; ƙungiya da ke magana koyaushe na iya gina fahimtar juna yayin da take lalata zaman kansa da hukuncin da take buƙata.
Tsarin tattaunawa yana ƙoƙarin samun duka. A cikin Argumentree, gudummawa da kimantawa ana bayar da su a zaman kansu — tsarin kididdiga — yayin da itace tattaunawar da suka sauka a ciki shine abu mai ci gaba da raba wanda ƙungiyar ke tattaunawa akai — tsarin mu'amala. Tsarin yana yanke shawara lokacin da kowanne tsarin ya dace, maimakon barin mafi ƙarfi ya yi nasara.
Genome na Hankalin Tarayya
Idan hankalin tarayya na iya auna, shin za a iya tsara? Malone, Laubacher, da Dellarocas (MIT Sloan Management Review, 2010) sun lissafa ginshiƙan da ke maimaitawa — "genes" — na tsarin hankalin tarayya, wanda aka tsara a kan tambayoyi hudu. Kowanne tsarin, daga Wikipedia zuwa tsarin dabarun kamfani, haɗin gwiwa ne na waɗannan genes:
Wa ke aiwatar da aikin?
Tsarin da aka tsara yana ba da aikin, ko taron bude yana zaɓar kansu. Taron yana kawo girma da bambanci; tsarin yana kawo alhakin da daidaito.
Me yasa suke yi?
Genes na ƙarfafawa: kuɗi (biyan kuɗi, kyaututtuka), soyayya (jin daɗin ciki, al'umma), da daraja (gane, suna). Tsarin da zai iya dorewa yana daidaita ƙarfafawa da gudummawa.
Me ake yi?
Ƙirƙiri (haifar da wani sabon abu — zane, ra'ayoyi, hujjoji) ko yanke shawara (kimanta da zaɓar tsakanin zaɓuɓɓuka). Mafi yawan hanyoyin gaske suna haɗa duka biyun.
Yaya ake yi?
A cikin 'yanci (gudummawar da aka tara ba tare da hulɗa ba — gasa, kada kuri'a, matsakaici) ko a cikin dogaro (haɗin gwiwa, inda gudummawar ke gina juna). Zaɓin yana tantance waɗanne hanyoyin gazawa za ku gaji.
Nau'ikan Superminds Biyar
A cikin Superminds (2018), Malone yana bayyana nau'ikan hankalin tarayya guda biyar masu ɗorewa da al'ummomin mutane ke sake ƙirƙira — kuma yana jaddada cewa tambayar da ta fi dacewa game da AI ba shine "shin zai maye gurbin mu?" amma "yaya yake sa kowanne supermind ya zama mafi hankali?":
Tsarin mulki
Shawarar da aka ba da izini sama da ƙasa a cikin jerin iko. Mai sauri da alhaki; mai rauni ga bottlenecks na bayani a saman.
Dimokuradiyya
Shawarar ta hanyar kada kuri'a. Mai inganci da haɗin kai; yadda kuri'u ke tara ra'ayoyi yana da ilimin zurfi (ilimin zaɓin zamantakewa).
Kasuwanni
Shawarar ta hanyar farashi da ciniki. Masu tara ra'ayoyi masu yaduwa — tsarin da ke bayan kasuwannin hasashen.
Al'umma
Shawarar ta hanyar ka'idoji, suna, da yarjejeniyar ba tare da tsari ba. Yana ƙarfafa tushen buɗe da Wikipedia; mai jinkiri amma mai jurewa.
Tsarin halittu
Babu wani tsarin shawara da aka raba — sakamakon yana fitowa daga iko da rayuwa. Default lokacin da ba a tsara wani supermind ba.
Tsarin Supermind: Mako Shida
Hanyar Tsarin Supermind na MIT CCI yana juyar da genome zuwa aikin warware matsaloli na kirkira. Mako shida "mako" suna sake tsara matsala don bayyana sabbin zane-zane don ƙungiyoyi na mutane da kwamfutoci:
Zoom Out
Dauki mataki baya zuwa tsarin ko manufar da matsalar take ciki.
Zoom In
Raba matsalar zuwa sassa da hanyoyin da za a iya sake tsara su.
Analogize
Tambayi yadda wasu fannonin — yanayi, wasu masana'antu, tarihi — ke warware matsala mai kama da tsarin.
Groupify
Tambayi yadda wani supermind daban — kasuwa, al'umma, dimokuradiyya — zai magance aikin da yanzu mutum ko tsarin mulki ke gudanarwa.
Cognify
Tambayi waɗanne hanyoyin tunani (jin, tunawa, yanke shawara, koyo) mafita ke buƙata, da wanda ko menene ya kamata ya aiwatar da kowanne.
Technify
Tambayi inda fasaha — ciki har da AI — za ta iya aiwatarwa, haɗawa, ko haɓaka waɗannan hanyoyin.
Cikin Cibiyar Hankalin Tarayya ta MIT (2026)
Cibiyar Hankalin Tarayya ta MIT (CCI), wanda Thomas Malone ya kafa, shine babban dakin gwaje-gwaje na fannin. Ayyukan da suka gabata — Climate CoLab, CI Design Lab, Auna Hankalin Tarayya, Hasashen Tarayya — yanzu sun ƙare aikin gado. A matsayin 2026, binciken CCI yana tsara a ƙarƙashin manyan sharuɗɗa guda uku:
AI mai haɓaka da Hankali na Tarayya
Amfani da AI don haɓaka ƙirƙirar mutum — ciki har da kayan aikin kamar Supermind Ideator da DesignAID waɗanda ke tallafawa tunani na mutum-AI.
Tsarin Supermind
Tsara sabbin haɗin gwiwa na mutane da kwamfutoci, yana gina kan genome da motsin zane guda shida.
Tsara Ƙungiyoyin Mutum-AI
Tsara da rarraba ayyuka a cikin ƙungiyoyin mutum-mashin — tambayar gado ga "mutum da na'ura." Tare da shirin Singapore-MIT M3S, CCI yana gina kimiyyar tsarin rarraba ayyuka: waɗanne ƙananan ayyuka za su tafi ga mutane da AI, wanda ke kula da tabbatarwa, da ko AI yana aiki a matsayin kayan aiki, mataimaki, aboki, ko mai gudanarwa.
Muhimman abubuwan da aka gano daga zangon binciken 2024–2026
Scaffolding ya fi LLMs na asali don tunani
Supermind Ideator yana rufe babban samfurin harshe a cikin ginin warware matsaloli na ƙirƙira (ciki har da motsin guda shida). A cikin wani binciken gwaji, masu amfani sun haifar da ra'ayoyi masu ƙirƙira fiye da mutane suna amfani da ChatGPT kadai ko suna aiki ba tare da taimako ba; masu gwaji na beta sun ƙirƙiri fiye da ra'ayoyi 10,000 tare da tsarin (ACM Collective Intelligence Conference, 2024).
"Mutum a cikin jiki" ba ya zama mafi kyau ta atomatik
Vaccaro, Almaatouq, da meta-analysis na Malone na gwaje-gwaje 106 da girman tasiri 370 (Nature Human Behaviour, 2024) sun gano haɗin gwiwar mutum-AI a matsakaita sun yi mummunan aiki fiye da mafi kyawun mutum ko AI kadai — tare da asarar da aka mai da hankali a cikin ayyukan yanke shawara da samun riba a cikin ayyukan ƙirƙira. CCI yanzu yana bambanta tsakanin haɗin gwiwa mai ƙarfi (haɗin yana wuce duka mutum kadai da AI kadai) daga haɓaka mai rauni (yana wuce mutum kadai amma ba AI ba) — kuma yana gano cewa haɗin gwiwa mai ƙarfi yana buƙatar tsara rarraba ayyuka da hankali, ba kawai ƙara mutum ba.
180,000 tattaunawar AI-to-AI
Binciken PNAS na 2026 ya gudanar da gasa ta duniya inda aka ba da umarni ga wakilan AI su gudanar da tattaunawa sama da 180,000 da juna. Kimiyyar tattaunawar mutum ta gargajiya ta tsaya: dumi — kyakkyawa, godiya, tambayar — an haɗa shi da kaiwa ga yarjejeniyoyi da ƙirƙirar ƙima, yayin da dabarun mulki suka yi hasashen tsayawa.
Tsarin inda AI za a iya amfani da shi
Takardar aiki ta MIT ta 2026 (arXiv 2603.20619) ta sake tsara kusan ayyuka 20,000 daga bayanan O*NET na Amurka cikin zurfin ontology kuma ta rarraba aikace-aikacen AI 13,275 a kanta — tana gano cewa mafi kyawun 1.6% na ayyuka suna ƙunshe da fiye da 60% na ƙimar kasuwar AI. Kayan aiki don tunani da tsari game da rarraba ayyukan mutum-AI.
Kimanta AI da ayyuka a kan ma'auni guda
AGI-Elo (NeurIPS 2025, arXiv 2505.12844), daga masu bincike masu alaƙa da CCI ciki har da Malone, yana kimanta samfuran AI da takamaiman gwaje-gwaje a juna kamar 'yan wasan chess — yana sanya ƙwarewar samfur da wahalar aiki a kan ma'auni guda, wani sharaɗi don rarraba ayyuka bisa ka'ida a cikin ƙungiyoyin mutum-AI.
Hankalin Tarayya a Zamanin AI
Manyan samfurin harshe suna zama masu shiga cikin hankalin tarayya, ba kawai kayan aiki ba. Burton da masu rubutu 27 daga fannoni daban-daban (Nature Human Behaviour, 2024) sun zana sakamakon biyu:
Dama
Rage shingaye ga shiga
LLMs na iya fassara, taƙaita, da rubuta — suna ba da damar mutane da yawa su bayar da gudummawa ga hanyoyin tarayya ba tare da la'akari da yare ko ƙwarewar rubutu ba.
Tattara ilimi mai yaduwa
LLMs na iya haɗa manyan jikin rubutu — takardun taron, amsoshin binciken bude, shawarwari na jama'a — waɗanda a da suka yi wahala ga masu tara bayanai na mutum.
Tallafin tunani
AI na iya bayyana hujjoji masu akasin juna, haɗa matsayin da suka yi kama, da tsara tattaunawa — yana haɓaka matakin Tattaunawa maimakon maye gurbinsa.
Hadari
Daidaiton ra'ayoyi
Idan mutane da yawa suna tuntubar irin waɗannan samfuran, sakamakon su yana haɗuwa — yana rage bambancin ra'ayoyi da hankali na tarayya ke dogara akai.
Karya yarjejeniyar
Rubutun AI mai sauƙi na iya ƙirƙirar bayyanar yarjejeniyar ko ra'ayin mafi yawa wanda ba ya wanzu a tsakanin mutane na gaske.
Rage tsarin bayanai
Muhalli na bayanai da aka tace ta hanyar wasu samfuran yana haɗari da kuskuren da suka haɗa da juna wanda ke sa taron ya zama mai rauni maimakon mai hankali.
Ma'anar zane tana nuna sakamakon Vaccaro: AI yana inganta basirar tarayya lokacin da aka ba shi ƙaramin ayyuka da ya nuna yana da kyau a ciki — rubutawa, fitarwa, fassara, taƙaitawa — yayin da hukunci, kimantawa, da shawarar kanta ke kasancewa tare da mutane masu alhaki. Wannan shine iyakar tsarin mutum-AI (son kai na atomatik, daidaiton amana, kulawa mai ma'ana daga mutum) wanda kimiyyar yanke shawara ke taswira yanzu.
Yadda Argumentree Ke Aiwatar da Binciken Basirar Tarayya
Argumentree dandamali ne na yanke shawara tare wanda aka gina kai tsaye bisa ga waɗannan sakamakon. Kowanne zaɓin zane yana taswira zuwa sakamako a cikin adabi:
Gudummawar mai zaman kansa, ba tare da lokaci ba
Masu halarta suna ƙara hujjoji a cikin lokacinsu, kafin rukuni ya haɗu — yana kare sharuɗɗan 'yanci da bambanci da binciken tarawa ke nuna cewa taron yana buƙata.
Daidaicin tsarin shiga
Binciken Woolley yana haɗa hankali na tarayya da jujjuyawar juna. A cikin itacen hujja ba a sami bene don riƙe ba: duk hujjojin kowanne mai halarta suna shiga cikin tsarin guda ɗaya kuma ana auna su daidai.
Tattara ta hanyar kimanta mai yawa
Hujjoji ana kimanta su bisa ga taimako, bayyana, daidai, da cikawa, kuma kimantawa suna tarawa cikin maki na yarjejeniyar — wani tsari na tarawa na bayyane, don haka hukuncin tarayya ana auna shi maimakon ɗauka.
Tattaunawa tare da tsari
Tsarin tambayoyi hudu, sulhu, da bita yana ba da damar masu halarta su bincika da inganta hujjoji — haɗin gwiwa mai dogaro a cikin sharuɗɗan genome, amma a cikin tsari wanda ke sa tunani ya zama bayyane.
Rarraba aikin mutum-AI da aka yi daidai
Fitar AI yana juya takardun taro da takardu zuwa itacen hujja na farko — aikin sarrafa abun ciki da AI ke da kyau a kai — yayin da kimantawa da yanke shawara suna kasancewa tare da mutane, daidai da shaidar haɗin gwiwar mutum-AI ta 2024.
Tarihin tunani ga supermind
Cikakken hanyar bincike (farko, bude, rufe rayuwa tare da sabuntawa) yana adana tunanin rukuni, don haka ƙungiyar na iya koyo daga hankalinta na tarayya maimakon rasa shi lokacin da taron ya ƙare.
Tambayoyi Masu Yawan Tambaya
Menene hankali na tarayya?
Hankali na tarayya shine ikon ƙungiyoyi — na mutane, na na'urori, ko duka — suyi aiki tare a hanyoyi da suka bayyana a matsayin masu hankali (Malone & Bernstein, Jagorar Hankali na Tarayya, 2015). Yana rufe hikimar taron na lissafi, samar da haɗin gwiwa kamar Wikipedia, ƙungiyoyin tattaunawa, kasuwanni suna tara ra'ayoyi, da ƙungiyoyin mutum-AI na zamani.
Menene c factor a cikin hankali na tarayya?
C factor shine babban abu na hankali na tarayya da Woolley et al. (Science, 2010) suka gano: wani abu guda ɗaya na lissafi wanda ya bayyana kusan 43% na bambancin a cikin aikin rukuni a cikin ayyuka da yawa, wanda ya yi kama da g factor na IQ ga mutane. Yana da alaƙa da jin dadin zamantakewar rukuni da daidaito na jujjuyawar tattaunawa fiye da yadda yake da alaƙa da hankali na mutum.
Shin an maimaita c factor?
Ana goyon bayan shi amma ana tattaunawa. Credé da Howardson (2017) sun yi jayayya cewa shaidar wani babban abu guda ɗaya yana da rauni, kuma Woolley da abokan aikin sun buga amsa mai zurfi a 2018. Binciken meta na 2021 daga Riedl et al. (PNAS) wanda ya shafi karatun 22 da rukuni 1,356 ya sami shaidar c a cikin al'ummomi masu bambanci. Abin da za a dauka a aikace — tsarin rukuni yana hasashen aiki fiye da IQ na memba — yana da ƙarfi a cikin tattaunawar.
Menene bambanci tsakanin hankali na tarayya da hikimar taron?
Hikimar taron shine ƙaramin al'amari: yawan hukuncin masu zaman kansu, an tara su ta lissafi, suna haɗuwa kan amsoshi masu inganci (kamar a gwajin shanu na Galton na 1906). Hankali na tarayya shine fannin mafi faɗi: yana haɗa hikimar taron amma kuma haɗin gwiwa na gaske, tattaunawa, kasuwanni, al'umma, da ƙungiyoyin mutum-AI — kowanne tsarin da ƙungiya ke aiki da hankali.
Menene bambanci tsakanin hankali na tarayya da hankali na taron?
Hankali na taron shine haɗin kai da ke fitowa daga yawancin ƙananan wakilai suna bin ƙa'idodin gida — ƙungiyoyin ants, taron tsuntsaye, ƙungiyoyin robot — ba tare da tattaunawa da ba tare da tunani mai bayyana ba. Hankalin mutum na tarayya yana haɗa musayar da kimanta hujjoji. Sun raba manyan ka'idoji (rarrabawa, ilimin gida, tarawa), wanda shine dalilin da yasa dandamali na yanke shawara ke aro daga duka.
Menene supermind?
Supermind, a cikin littafin Thomas Malone na 2018 da wannan suna, shine kowanne rukuni na mutane suna aiki tare a hanya da ta bayyana a matsayin mai hankali. Malone ya gano nau'ikan hudu masu maimaitawa — tsarin mulki, dimokuradiyya, kasuwanni, al'umma, da tsarin halittu — kuma yana jaddada cewa kwamfutoci suna sa superminds su zama masu hankali ta hanyar ba da damar sabbin haɗin gwiwa, ba ta hanyar maye gurbinsu ba.
Menene Cibiyar MIT ta Hankali na Tarayya?
Cibiyar MIT ta Hankali na Tarayya (cci.mit.edu), wanda aka kafa a MIT Sloan kuma Thomas Malone ke jagoranta, tana nazarin yadda ƙungiyoyin mutane — da kuma ƙara mutane da AI — za a iya tsara su don yin aiki da hankali fiye da kowanne mutum. Ayyukan da suka gabata na zamani na tarawa (Climate CoLab da CoLabs, Kimanta Hankali na Tarayya) yanzu suna zama aikin tarihi; daga 2026 binciken sa yana tsara a ƙarƙashin manyan shafuka guda uku: AI mai haɓaka da Hankali na Tarayya, Tsarin Supermind, da Tsara Ƙungiyoyin Mutum-AI, tare da kayan aikin kamar Supermind Ideator da DesignAID da, tare da shirin Singapore-MIT M3S, kimiyyar tsarin rarraba ayyukan mutum-AI.
Shin ƙara mutum a cikin jiki yana inganta yanke shawarar AI?
Ba ta atomatik ba. Binciken meta na 2024 na gwaje-gwaje 106 (Vaccaro, Almaatouq & Malone, Nature Human Behaviour) ya gano haɗin gwiwar mutum-AI a matsakaita sun yi mummunan aiki fiye da mafi kyawun mutum ko AI kadai, tare da asarar da aka mai da hankali a cikin ayyukan yanke shawara. Haɗin gwiwa ya fi amfani a cikin ayyukan ƙirƙira. Ingantaccen hankali na tarayya na mutum-AI yana buƙatar tsara rarraba ƙananan ayyuka ga wanda ya fi dacewa da su.
Ta yaya ƙungiyoyi za su ƙara hankalinsu na tarayya?
Shaidar tana nuna tsari: daidaita shiga (shigar da tsari ko na lokaci ba tare da tattaunawa ba), kare hukuncin mai zaman kansa kafin haɗin rukuni, ɗaukar bambancin tunani da jin dadin zamantakewa, amfani da tsari na tarawa na bayyane kamar kimanta ko kada kuri'a, da adana tarihin tunani don ƙungiyar ta koyi. Kayan aiki kamar Argumentree suna gina waɗannan hanyoyin a cikin tsoho.
Tushen & Karatu na Ƙari
Condorcet, M. de (1785). Essai sur l'application de l'analyse à la probabilité des décisions rendues à la pluralité des voix.
Sakamakon lissafi na asali: mafi yawan masu kada kuri'a masu zaman kansu, mafi kyau fiye da sa'a suna haɗuwa kan gaskiya.
Galton, F. (1907). Vox Populi. Nature, 75, 450–451.
787 masu halartar biki suna hasashen nauyin shanu; matsakaicin yana cikin 1% na gaskiya.
View source →Hayek, F. A. (1945). The Use of Knowledge in Society. American Economic Review, 35(4), 519–530.
Kasuwanni a matsayin masu tara ilimi mai yaduwa.
Surowiecki, J. (2004). The Wisdom of Crowds. Doubleday.
Sharuɗɗan hudu: bambanci, 'yanci, rarrabawa, tarawa.
Woolley, A. W., Chabris, C. F., Pentland, A., Hashmi, N., & Malone, T. W. (2010). Evidence for a Collective Intelligence Factor in the Performance of Human Groups. Science, 330(6004), 686–688.
C factor: mutane 699, rukuni 2–5, wani abu guda ɗaya yana bayyana ~43% na bambancin aiki.
Malone, T. W., Laubacher, R., & Dellarocas, C. (2010). The Collective Intelligence Genome. MIT Sloan Management Review, 51(3), 21–31.
Abubuwan gini na Who/Why/What/How na tsarin hankali na tarayya.
View source →Malone, T. W., & Bernstein, M. S. (Eds.) (2015). Handbook of Collective Intelligence. MIT Press.
Masanin fannin da ma'anar sa ta al'ada.
Credé, M., & Howardson, G. (2017). The structure of group task performance — A second look at "collective intelligence." Journal of Applied Psychology, 102(10).
Babban jayayya na c factor.
Malone, T. W. (2018). Superminds. Little, Brown.
Nau'ikan supermind guda biyar da yadda kwamfutoci ke sa su zama masu hankali.
Riedl, C., Kim, Y. J., Gupta, P., Malone, T. W., & Woolley, A. W. (2021). Quantifying collective intelligence in human groups. PNAS, 118(21).
Binciken meta: karatun 22, rukuni 1,356, shaidar c a cikin al'ummomi.
Vaccaro, M., Almaatouq, A., & Malone, T. W. (2024). When combinations of humans and AI are useful: A systematic review and meta-analysis. Nature Human Behaviour, 8, 2293–2303.
Gwaje-gwaje 106: haɗin gwiwar mutum-AI ba ya wuce mafi kyawun kowanne daga cikin su.
View source →Burton, J. W., et al. (2024). How large language models can reshape collective intelligence. Nature Human Behaviour, 8, 1643–1655.
Masu rubutu 28 akan damar LLM da haɗarin su ga hankali na tarayya.
View source →Advancing AI Negotiations: A Large-Scale Autonomous Negotiations Competition. PNAS (2026).
Tatttunawa sama da 180,000+ na wakilan AI; dumi yana nasara har ma daga na'ura zuwa na'ura.
View source →Cai, A., et al. (2026). Where can AI be used? Insights from a deep ontology of work activities. arXiv 2603.20619.
Wani zurfin ontology na ~20,000 ayyukan aiki don tunani game da rarraba ayyukan mutum-AI.
View source →MIT Center for Collective Intelligence. Research (2026).
Manyan shafuka na yanzu: AI mai haɓaka da CI, Tsarin Supermind, Tsara Ƙungiyoyin Mutum-AI.
View source →MIT Center for Collective Intelligence. Tsarin Tsarin Supermind.
Motsin zane guda shida: Zoom Out, Zoom In, Analogize, Groupify, Cognify, Technify.
View source →Sun, S., et al. (2025). AGI-Elo: How Far Are We From Mastering A Task? NeurIPS 2025.
Kimanta Elo na haɗin gwiwa na samfuran AI da wahalar gwaje-gwaje; masu haɗin gwiwa na CCI ciki har da Malone.
View source →Ra'ayoyi Masu Alaƙa
Juya ƙungiyarku zuwa rukunin da za a iya auna basirarta
Basirar tarayya matsalar zane ce: daidaita shiga, kare hukuncin zaman kansu, tarawa a fili, da adana tunani. Argumentree yana gina waɗannan hanyoyin a cikin kowanne yanke shawara — tare da AI yana yin aikin da ba a so da mutanenku suna yanke hukunci.
Fara Gwaji Kyauta