Al-Masihu Ya iso, Kamar Yadda Ganuwar Mu Ke Rugujewa
Anan ga gwajin tunani don kofi na safiyar Litinin ɗin ku: Google kawai ya sauke Gemini 4 Argon, "iyakan iyaka AI" tare da fitowar alamar alama miliyan 1, a sarari wanda aka tsara don zama babban mai kare tsaro ta yanar gizo [2]. Yana da dijital na dijital daidai da dandali mai ƙyalli mai ƙyalli mai ƙyalli na orbital. A lokaci guda, yanayin hasashe mai ma'ana: Mai tsaron Microsoft, daidaitaccen dijital na sturdy, idan ɗan tsatsa, ƙofar gidan ka, yana da raunin rana, bari mu kira shi 'RoguePlanet,' ana amfani da shi don samun damar matakin SYSTEM.
Fada min, wanne ne yake raya ku da dare?
Don duk tsananin farin ciki da ke kewaye da AI, don duk alkawuran gano barazanar kai tsaye da bincike mai rauni, gaskiyar da ba a bayyana ba ita ce, tushen tushenmu har yanzu yana cike da ramukan da ɗan adam ya yi. Google yana so ya sayar mana da hangen nesa na AI-powered cyber-nirvana, amma yanzu yana tunatar da mu cewa ƙasa a ƙarƙashin ƙafafunmu har yanzu tana da sauri.
Mafarkin Argon: Alamomin bege miliyan (ko Hubris?)
Don bayyanawa: Gemini 4 Argon babban tsari ne. Ƙididdigar fitarwa miliyan 1 ba kawai babban buffer ba ne; yana wakiltar babban canji a zurfin nazari [3]. Ka yi tunanin ciyar da duk wani log ɗin martani na abin da ya faru, cikakken jujjuyawar hanyar sadarwa, ko cikakkiyar lambar tushe na aikace-aikace mai mahimmanci a cikin AI kuma samun shi ya tofa ɓoyayyiyar ɓarna, ɓarna, da yuwuwar fa'ida waɗanda zasu ɗauki ƙungiyar manazarta ɗan adam makonni don ganowa. Alkawarin kenan.
Shawarar da Google ya yanke na fitar da wannan keɓancewar ga "amintattun masu kariyar yanar gizo" ta hanyar shirin su na Fairwind wani shiri ne na wayo. Yana sarrafa labarin, yana tabbatar da cewa masu karɓar farkon na iya zama masu ba da shawara, kuma, mahimmanci, yana bawa Google damar tattara bayanan duniya masu kima a cikin yanayi mai sarrafawa. Ba wai kawai game da sayar da samfur ba ne; shi ne game da tsara makomar tsaro ta kasuwanci tare da Google a jagorancin. Wannan ba kawai wani LLM ba ne; sanarwa ce mai ƙarfi game da barazanar yanar gizo, tare da AI a matsayin makami mai ƙarfi [5].
Amma a nan ne shakku ya taso. "Masu tsaro masu aminci." Wanene ya bayyana "amintaccen"? Shin manyan kamfanoni ne kawai masu albarkatu masu yawa da kwangilar Google da ake da su? Menene game da ƙananan ƙungiyoyi, masu samar da ababen more rayuwa masu mahimmanci ba tare da layin kai tsaye zuwa ƙungiyar Fairwind ta Google ba, waɗanda za a iya cewa sun fi * masu rauni kuma * ƙari * suna buƙatar kayan aikin ci gaba? Wannan zaɓen zaɓen, yayin da ake iya fahimta daga hangen nesa na haɓaka samfur, yana haɗarin ƙirƙirar yanayin tsaro mai hawa biyu: ƙwararrun ƙwararrun AI da kowa da kowa yana ci gaba da zazzage faɗakarwar SIEM da hannu.
Kuma kada mu manta da hatsarori na asali. Fitowar alama ta miliyan 1 tana da ƙarfi, amma kuma babban filin hari ne idan samfurin kanta na iya sarrafa shi ko guba. Menene zai faru lokacin da maharin ya koyi sana'a ya sa ya ƙetare kariyarsa, ko ya shigar da muggan bayanai a cikin tsarin horonsa? Muna magana ne game da AI da aka ƙera don nemo allura a cikin hay, amma menene idan haydar kanta ta lalace?
RoguePlanet: Gaskiyar Tsaro mara kyau
Canjawa daga wannan makoma mai haske zuwa ga ɓarna a halin yanzu, la'akari da raunin 'RoguePlanet' na mu a cikin Mai tsaron Microsoft. isa ga matakin SYSTEM. An yi amfani da shi sosai a cikin yanayi mai ma'ana. Wannan ba rauni ba ne na ka'idar; yana wakiltar ramin gibi a cikin ɗayan mafi yaɗuwar hanyoyin tsaro na ƙarshen duniya a duniya. Defender shine tushe ga kungiyoyi marasa adadi, layin farko na tsaro ga miliyoyin injuna. Kuma a cikin wannan yanayin, an ƙetare shi kawai.
Wannan ba gazawar AI ba ce. Wannan gazawar asali ce ta injiniyan software, na ƙwaƙƙwaran gwaji, na ɓangaren ɗan adam wanda har yanzu yana ginawa da kiyaye waɗannan mahimman tsarin. Yayin da muke shagaltuwa da muhawara game da abubuwan da suka shafi ci gaba na AI da kuma abubuwan aikin injiniya na gaggawa, maharan suna tafiya kawai ta ƙofar gaba saboda wani ya bar ta a buɗe.
Tasirin nan da nan shine ɓarkewar gaggawa. Masu kula da IT da ƙungiyoyin tsaro na yanar gizo yanzu suna cikin matsanancin ƙoƙari, suna ba da fifiko ga faci, farautar alamun sasantawa, da yin addu'a ba a riga an same su ba. Wannan shi ne gritty, babban matsa lamba gaskiyar cybersecurity cewa babu wani adadin jin daɗin AI da zai iya ɓoyewa. Yaƙi ne na yau da kullun, yaƙi da kuskuren ɗan adam da mugun nufi, wanda aka buga a cikin ramukan sarrafa facin da martanin da ya faru.
Gaskiyar Frontier Ba AI kawai ba; Tushen Tsaro ne
Juxtaposition na Gemini 4 Argon da 'RoguePlanet' ba kawai m; tunatarwa ce mai tada hankali. Muna haɓaka ƙaƙƙarfan ƙaƙƙarfan ƙaƙƙarfan ƙaƙƙarfan ƙa'idodin tsaro na AI akan tushe waɗanda har yanzu basu da ƙarfi.
AI, har ma da iyaka AI kamar Argon, kayan aiki ne. Kayan aiki mai ƙarfi, mai canzawa, babu shakka. Zai ƙara ƙarfin ɗan adam, sarrafa ayyuka masu wahala, da yuwuwar fallasa barazanar da ba mu iya gani a da. Amma ba mafita ba ce ta duniya. Ba zai kawar mana da alhakin rubuta amintaccen lamba ba, gwada gwada software ɗin mu sosai, yin faci da sauri, da kuma kula da ƙaƙƙarfan kariyar mai nau'i-nau'i.
Haɗarin shine cewa ruɗin AI yana haifar da rashin jin daɗi. "Oh, Gemini 4 Argon zai kama hakan." Zai iya? Ko za a horar da shi a duniyar da 'RoguePlanet'-rauni ba kasafai ba ne, kuma ta haka ne za a rasa amfani da sabon labari na gaba wanda ke haifar da irin wannan aibi? AI yana da kyau kwarai a ƙirar ƙira, amma sifili-kwanaku, ta ma'ana, * sababbi * alamu.
Ainihin "sabon kan iyaka" a cikin tsaro ta yanar gizo ba kawai ƙara ƙarin AI ga matsalar ba. Yana da game da sabuntawa, kusan m, mai da hankali kan tushen tushe. Yana game da amintattun zagayowar ci gaban software, ingantaccen ƙirar ƙira, ci gaba da sarrafa rauni, da al'adar da ke ba da fifiko ga tsaro tun daga tushe, ba kamar yadda ake tunani ba. Yana da game da fahimtar cewa yayin da AI zai iya taimaka mana mu ƙara gani, ba zai gyara kurakuran da muka riga muka yi tafiya a kai ba.
Don haka, yayin da Google ke shagaltuwa da ƙaddamar da dandalin tsaro na orbital, kada mu manta da ƙarfafa ganuwar katangar. Domin a yanzu, maharan ba sa ƙoƙarin hack AI; suna kawai cin gajiyar Windows Defender da ba a buɗe ba akan kwamfutar tafi-da-gidanka na Shugaba. Kuma wannan, abokaina, matsala ce babu abin da aka fitar da alamar miliyoyin da zai iya magancewa idan babu tsaftar muhalli.
Sources
[2] Manyan Labaran Fasaha A Yau, Oktoba 1, 2026: AMD, Broadcom, Google, LG, Meta ... [3] Babban ra'ayin karni | 3 ga Oktoba 2026 | Masanin Tattalin Arziki - https://www.economist.com/weeklyedition/2026-10-03 [5] Sabbin labarai na AI da muka sanar a cikin Satumba 2026 - https://blog.google/innovation-and-ai/technology/ai/google-ai-updates-september-2026/