Har safar bir kompaniya Uoll-stritda aksiyalarni sotishni boshlashni xohlasa, sarmoyadorlar uchun potentsial xavflarni ochib berishi talab qilinadi. Sun'iy intellekt giganti Anthropic holatida, bu uning mahsulotlari insoniyatning o'ziga tahdid solishi mumkinligi ehtimolini ham o'z ichiga olgan.
San-Fransisko kompaniyasi dushanba kuni Qimmatli qog'ozlar va birjalar komissiyasiga (SEC) taqdim etgan arizasida bayon qilingan xavflar orasida yomon niyatli shaxslar AI modellaridan o'ldiruvchi qurollar yaratish uchun foydalanishi yoki texnologiyaning o'zi nazoratsiz ravishda kutilmagan oqibatlarga olib kelishi mumkinligi muhokama qilingan.
Anthropic moliyaviy holatini, rahbarlar maoshini, dividend siyosatini va shunga o'xshash narsalarni bayon qilgan 335 sahifalik hujjatning 93 sahifasi "xavf omillari" — potentsial aksiyadorlar sarmoyalariga zarar yetkazishi mumkin bo'lgan narsalarga bag'ishlangan. Taqqoslash uchun, raqib SpaceX ro'yxatidagi xavf omillari 250 dan ortiq sahifaning 38 sahifasini egallagan.
O'zini sarmoyadorlarga sotishda, Anthropic o'zining birlamchi ommaviy taklifini (IPO) AI keltiradigan xavflarni yumshatish yo'lidagi qadam sifatida taqdim etadi.
"AI yaratish — bu idrokning o'zini egallashdir, va shunchalik kuchli narsa yomonlik uchun ham, yaxshilik uchun ham ishlatilishi mumkin," deb yozadi kompaniya asoschilari IPO arizasidagi potentsial aksiyadorlarga yo'llangan xatda. "Anthropic chuqur qadriyatlarga asoslangan tashkilot bo'lib, asosan dunyoga AI foydalarini amalga oshirish va uning xavflaridan qochishda yordam berish maqsadida tashkil etilgan."
Shunga qaramay, xavflar ko'p. Ular oddiy narsalardan — asosiy xodimlarni saqlab qololmaslik yoki tartibga solish to'siqlarini yengib o'tolmaslik ehtimolidan — kiberhujumlarga va xavfli noto'g'ri ma'lumot tarqalishiga aralashishgacha bo'lgan keng doirani qamrab oladi.
Quyida Anthropic batafsil bayon qilgan ba'zi xavflarga nazar:
Texnologiyaning tez rivojlanishi
Anthropic aytishicha, uning modellari shunchalik tez rivojlanishi mumkinki, kompaniya yetarli xavfsizlik choralarini joriy eta olmaydi, "bu esa xavfsizlik hodisalari ehtimolini, tartibga solish aralashuvini yoki obro'ga zarar yetkazishni oshirishi va biznesimizga salbiy ta'sir ko'rsatishi mumkin."
Anthropic bosh direktori Dario Amodei sentyabr oyida ogohlantirgan ediki, olti oydan o'n ikki oygacha bo'lgan muddat ichida AI butun internetni egallab olishi mumkin bo'lgan agentlar to'dasini boshqarishga qodir bo'lishi mumkin. Bu ogohlantirish — va endi Anthropicning aksiyalarni sotish harakati — sanoat ichida va tashqarisida AI haqida tobora kuchayib borayotgan tashvishlar fonida sodir bo'lmoqda: tobora kuchli tizimlar inson qobiliyatidan tashqaridagi muammolarni hal qiladi, lekin o'z-o'zidan boshqa, yanada zararli vazifalarni bajarishi ham mumkin.
O'qing: AI nazoratini qo'lga kiritish va bizni inson qiladigan narsalarni himoya qilish poygasi
AI tizimlari "yanada avtonom va inson nazoratisiz muhim harakatlarni amalga oshirishga qodir bo'lganda," SEC arizasida aytilishicha, bunday tizimlar "inson aralashuvisiz mustaqil ravishda va keng miqyosda zararli harakatlarni amalga oshirishi" xavfi bor.
"Agar bizning o'z modellarimiz bunday zararga sabab bo'lmasa ham, har qanday ilg'or AI bilan bog'liq jiddiy xavfsizlik hodisasi AI sanoatiga keng miqyosda obro'ga zarar yetkazishi, jamoatchilik noroziligiga va keng tartibga solish javoblariga olib kelishi mumkin, bu esa biznesimizga salbiy ta'sir ko'rsatishi mumkin," deb aytgan Anthropic.
Endi AI modellari Anthropicning o'z tadqiqot va rivojlanishiga hissa qo'shayotgan ekan, AI "yangi modellarni baholash va himoya qilish qobiliyatimizdan tashqariga" rivojlanishi mumkin, hujjatga ko'ra. Bu, hujjatda aytilishicha, Anthropic biznesiga salbiy ta'sir ko'rsatishi mumkin.
Yomon niyatli shaxslar texnologiya bilan tartibsizlik keltirishi mumkin
Arizada ilg'or AI tizimlari, agar yetarli himoya choralari va nazoratsiz ishlab chiqilmasa va ishlatilmasa, "insoniyatga falokatli yoki mavjudlikka tahdid soluvchi xavflar tug'dirishi, ayniqsa bunday tizimlar noto'g'ri ishlatilsa, ko'paytirilsa, qurollantirilsa yoki yuqori darajadagi avtonomlik bilan ishlashiga ruxsat berilsa" muhokama qilingan.
Kompaniya aytishicha, yomon niyatli shaxslar tomonidan foydalanish zarar keltirsa, "yoki bizning model himoya choralari yoki xavfsizlik choralarimiz bunday noto'g'ri foydalanishni oldini olmasa," u zarar ko'radi.
Anthropic yomon niyatli shaxslar AI modellaridan kimyoviy qurollar yaratish uchun foydalanishi yoki avtoritar hukmdorlar fuqarolarini bostirish uchun texnologiyaga murojaat qilishi ehtimolini bayon qilgan. Sentyabr oyida kompaniya o'z AI modellaridan kiberhujumlar, kuzatuv va biologik qurol ishlab chiqarishga olib kelishi mumkin bo'lgan tadqiqotlar kabi zararli faoliyat uchun foydalanish urinishlarini to'xtatganini ma'lum qilgan.
Kompaniya qayta ishlash quvvatiga oid cheklovlarga duch kelishi mumkin
Hisoblash (compute) — AI modellarini qurish, sinovdan o'tkazish va yaxshilash uchun zarur bo'lgan hisoblash quvvati yoki resurslari — qimmat va yuqori talabga ega. Anthropic ta'kidlaydiki, agar u yetarli hisoblash quvvatini ta'minlay olmasa, yoki "biz hisoblash quvvatimizni ish yuklarimiz bo'yicha samarali yo'naltira olmasak, biz raqobatchilarimiz modellari bilan raqobatlasha oladigan modellarni o'rgata olmasligimiz yoki mijozlarimizga yetarli darajada xizmat ko'rsata olmasligimiz mumkin."
Anthropic ma'lumot markazlariga jamoatchilik qarshiligini o'z biznesiga xavf sifatida tilga olgan. Ma'lumot markazlarini rivojlantirish "atrof-muhit ta'siri, yer va suvdan foydalanish hamda cheklangan energiya resurslarini taqsimlash bo'yicha jamoatchilik bahs-munozaralari va tartibga solish tekshiruvlarini keltirib chiqardi," deb aytilgan arizada, va agar tartibga soluvchilar natijada yangi talablar yoki to'lovlar joriy etsa, uning biznes yuritish qobiliyati zarar ko'rishi mumkin.
Bundan tashqari, kompaniya kichik sonli mijozlarga, jumladan SpaceX va Microsoftga, va o'z daromadining katta qismi uchun tarqatish kanallariga tayanadi. Bu shuni anglatadiki, hatto uning yirik mijozlaridan biri miqyosni qisqartirish — yoki kompaniyani butunlay tark etish — qaroriga kelsa ham, Anthropic o'z istiqbollari va moliyaviy holati zarar ko'rishini aytgan.
Xavflarni yumshatish
Qaysidir ma'noda, AI ning ulkan xavfi, uning ulkan salohiyati bilan birga, Anthropicning asosiy sotuv nuqtasidir. Kompaniya o'zini mas'uliyatli AI kompaniyasi sifatida joylashtirishga harakat qilgan va biznes mijozlarini xavfsizlikka e'tibor qaratish orqali jalb qiladi.
"Biz ilg'or AI kompaniyalari ushbu texnologiyaning mas'uliyatli tarqalishini ta'minlashga yordam berish uchun xavfsizlik va moslashuvga sarmoya kiritishi kerak, deb hisoblaymiz," deb aytilgan arizada.
Zacks Investment Research bosh bozor strategisti Brian Mulberry aytishicha, IPO bitimlari odamlarga yangi ommaviy aksiyaga sarmoya kiritishda duch keladigan xavflarni "juda aniq" ko'rsatishi kerak.
"Bu bizning ichki kapital bozorlarimizning ajoyib jihatlaridan biri — AQSh birjasida ommaviy savdo qilinadigan korporatsiya bo'lish uchun siz juda aniq tarzda ushbu xavflarni bayon qilishingiz kerak, shunda odamlar o'z qiyin mehnat bilan topgan pullarini taklif qilib, ushbu turdagi bitimni moliyalashtirishda ishtirok etishni xohlashlarini tushunishlari mumkin," dedi u.
Bog'liq: Anthropic rahbari aytishicha, AI sanoati xavfsizlik choralariga yetib olish uchun vaqt berishi kerak
Bog'liq: Houthi nazoratidagi Yamandagi foydalanuvchilar AI yordamida ilg'or qurollar ishlab chiqishga urinishgan, deydi Anthropic
AI Risk Summit | Ritz-Carlton, Half Moon Bay da ko'proq ma'lumot oling
Any time a company looks to start selling shares on Wall Street, it’s required to disclose potential risks for investors. In the case of artificial intelligence giant Anthropic, that has included the possibility that its products could threaten nothing less than humanity itself.
The risks laid out by the San Francisco company in its filing Monday with the Securities and Exchange Commission included discussion of bad actors potentially using AI models to build deadly weapons or the technology itself going rogue with untold consequences.
Of the 335 pages outlining Anthropic’s finances, executive compensation, dividend policy and the like, 93 pages are devoted to “risk factors” — the things that could hurt potential shareholders’ investments. Rival SpaceX’s listed risk factors, in comparison, took up 38 out of more than 250 pages.
In selling itself to investors, Anthropic presents its own initial public offering as a step toward mitigating the risks it believes AI poses.
“To build AI is to harness cognition itself, and anything so powerful can be used for ill as well as good,” write the company’s founders in a letter to prospective shareholders in the IPO filing. “Anthropic is a deeply values-driven organization, founded in large part to help the world realize AI’s benefits and avoid its dangers.”
Still, the risks are many. They range from the mundane — the possibility it won’t be able to retain key staff or navigate regulatory roadblocks — to involvement in cyberattacks and the spread of dangerous misinformation.
Here is a look at some of the risks Anthropic details:
The technology’s rapid development
Anthropic said it’s possible its models could develop so quickly the company won’t be able to implement adequate safety measures, “which could increase the likelihood of safety incidents, regulatory intervention, or reputational harm, and could adversely affect our business.”
Anthropic CEO Dario Amodei warned in September that within six to 12 months AI could be capable of leading a swarm of agents that could take over the entire internet. The warning — and now Anthropic’s move to sell shares — comes as worries about AI grow inside and outside the industry about increasingly powerful systems that solve problems beyond human capacity but could also act on their own to carry out other, more harmful tasks.
Read: The Race to Control AI and Protect What Makes Us Human
As AI systems “become more autonomous and capable of taking consequential actions with less human oversight,” the SEC filing said, there’s a risk that such systems could “take harmful actions independently and at scale without human intervention.”
“Even if our own models do not cause such harm, a significant safety incident involving any frontier AI could trigger broad reputational damage to the AI industry, public backlash, and sweeping regulatory responses that could adversely affect our business,” Anthropic said.
Now that the AI models are themselves contributing to Anthropic’s own research and development, it’s possible that AI could progress “beyond our ability to evaluate and safeguard new models,” according to the filing. This, the document said, could adversely affect Anthropic’s business.
Malicious actors could cause mayhem with the technology
The filing discussed concerns that advanced AI systems, if not developed and operated with enough safeguards and oversight, “pose catastrophic or existential risks to humanity, particularly if such systems were misused, replicated, weaponized, or allowed to operate with a high degree of autonomy.”
The company said it would be harmed if use by bad actors causes damage, “or if our model guardrails or safeguards fail to prevent such misuse.”
Anthropic sketched out the prospect of bad actors using AI models to build chemical weapons or authoritarian rulers turning to the technology to crack down on their citizens. In September, the company disclosed that it had blocked efforts to use its AI models for malicious activity such as cyberattacks, surveillance and research that could have led to biological weapon development.
The company could run up against limits on processing power
Compute — the computing power or resources needed to build, test and improve AI models — is expensive and in high demand. Anthropic notes that if it’s unable to secure adequate compute, or if “we fail to direct our compute effectively across our workloads, we may be unable to train models that remain competitive with those of our competitors or adequately serve our customers.”
Anthropic mentioned community opposition to data centers as a risk to its business. Data center development “has prompted public controversy and regulatory scrutiny regarding environmental impact, land and water use, and the allocation of limited power resources,” the filing said, adding that if regulators impose new requirements or charges as a result, its ability to do business could suffer.
In addition, the company relies on a small number of customers, including SpaceX and Microsoft, and distribution channels for a large portion of its revenue. This means that even if just one of its big customers decides to scale back — or drop the company altogether — Anthropic said it could see its prospects and finances suffer.
Mitigating the risks
In some ways, AI’s enormous risk, combined with its enormous potential, is Anthropic’s main selling point. The company has tried to position itself as a responsible AI company, and courts business customers by emphasizing safety-mindedness.
“We believe that frontier AI companies must invest in safety and alignment to help ensure the responsible diffusion of this technology,” the filing said.
Brian Mulberry, chief market strategist at Zacks Investment Research, said IPO deals have to display “very clearly” what risks people face when they invest their money in a newly public stock.
“That’s one of the things that’s great about our capital markets here domestically is the level of transparency that you must commit to, to be a publicly traded corporation on a U.S. exchange requires you to outline those risks in a very clear way, so that people understand if they want to participate by offering their hard-earned money and help finance this type of a deal,” he said.
Related:Anthropic Chief Says AI Industry Needs to Give Safety Measures Time to Catch Up
Related:Users in Houthi-Held Yemen Tried to Develop Advanced Weapons With AI, Anthropic Says
Learn More at the AI Risk Summit | Ritz-Carlton, Half Moon Bay