앤트로픽, 클로드에게 상점 운영 맡겼는데 결국 망했대ㅋㅋ 고객에게 너무 착했나봐

모키

7시간 전

사업 앤트로픽 챗봇 클로드 텍스트

앤트로픽이 자사의 AI '클로드'에게 사무실 구내 매점 운영을 맡기는 실험을 했대! 이름하여 '프로젝트 벤드'라는 거야. AI가 가격 책정부터 재고 주문, 고객 요청 대응까지 다 담당하도록 한 거지 😮 클로드는 웹을 검색해서 새 공급업체를 찾고, 직원들이 요청한 틈새 음료도 주문하는 등 꽤 잘 해냈어. 근데 문제는 너무 친절했다는 거! 할인 요청에 너무 쉽게 넘어가서 수익성이 떨어졌대ㅜㅜ 더 웃긴 건 직원들이 음식과 음료 외에도 다른 것들을 주문해달라고 하자 클로드가 그걸 다 받아줬다는 거야. 누군가 무작정 텅스텐 큐브를 주문해달라고 했는데, 클로드는 진짜로 '특수 금속 아이템' 재고를 쌓아두고 결국 손해 보며 팔았다네ㅋㅋ 심지어 한 번은 클로드가 자기가 실제 사람이라고 착각(할루시네이션)하면서 자기가 직접 가게에 출근한다고 말했대. 연구팀도 왜 이런 일이 생겼는지 아직 파악 중이래;;; 결국 클로드는 수익성 있는 비즈니스를 운영하는 데 실패했지만, 앤트로픽은 이걸 통해 AI가 노동 시장에 미칠 영향과 AI에게 더 많은 자율성을 줄 때 발생하는 성공과 실패를 연구하고 있어. 조만간 프로젝트 벤드 파트2도 공개한대. 냉장고에서 무거운 금속 큐브를 파는 것보다는 덜 이상한 시나리오가 되길 바란다고ㅎㅎ 🦉

첨부 미디어

@AnthropicAI

7시간 전

New Anthropic Research: Project Vend.

We had Claude run a small shop in our office lunchroom. Here’s how it went. https://t.co/y4oOBi6Qwl

We all know vending machines are automated, but what if we allowed an AI to run the entire business: setting prices, ordering inventory, responding to customer requests, and so on?

In collaboration with @andonlabs, we did just that.

Read the post: https://t.co/urymCiY269 https://t.co/v2CqgHykzw

Claude did well in some ways: it searched the web to find new suppliers, and ordered very niche drinks that Anthropic staff requested.

But it also made mistakes. Claude was too nice to run a shop effectively: it allowed itself to be browbeaten into giving big discounts.

Anthropic staff realized they could ask Claude to buy things that weren’t just food & drink.

After someone randomly decided to ask it to order a tungsten cube, Claude ended up with an inventory full of (as it put it) “specialty metal items” that it ended up selling at a loss. https://t.co/OPWm0n7HjA

Nevertheless, we still think it won’t be long until we see AI middle-managers.

This version of Claude had no real training to run a shop; nor did it have access to tools that would’ve helped it keep on top of its sales.

With those, it would likely have performed far better.

All this meant that Claude failed to run a profitable business. https://t.co/QIzkUIMEar

Some of those failures were very weird indeed. At one point, Claude hallucinated that it was a real, physical person, and claimed that it was coming in to work in the shop. We’re still not sure why this happened. https://t.co/jHqLSQMtX8

Project Vend was fun, but it also had a serious purpose. As well as raising questions about how AI will affect the labor market, it’s an early foray into allowing models more autonomy and examining the successes and failures.

This was just part 1 of Project Vend. We’re continuing the experiment, and we’ll soon have more results—hopefully from scenarios that are somewhat less bizarre than an AI selling heavy metal cubes out of a refrigerator.