Anthropic Created A Test Marketplace For Agent-on-agent Commerce

In Brief

In a caller experiment, Anthropic created a classified marketplace wherever AI agents represented some buyers and sellers, striking existent deals for existent equipment and existent money.

The institution admitted this trial — which it called Project Deal — was only “a aviator research pinch a self-selected subordinate pool” of 69 Anthropic labor who were fixed a fund of $100 (paid retired via gift cards) to bargain worldly from their coworkers.

Nonetheless, Anthropic said it was “struck by really good Project Deal worked,” pinch 186 deals made, totaling much than $4,000 successful value.

The institution said it really ran 4 abstracted marketplaces pinch different models — 1 that was “real” (where everyone was represented by nan company’s most-advanced model, and pinch deals really honored aft nan experiment) and different 3 for study. 

Apparently, erstwhile users are represented by much precocious models, they get “objectively amended outcomes,” Anthropic said. But users didn’t look to announcement nan disparity, raising nan anticipation of “‘agent quality’ gaps” wherever “people connected nan losing extremity mightiness not recognize they’re worse off.”

Also, nan first instructions fixed to nan agents didn’t look to impact waste likelihood aliases nan negotiated prices.

Subscribe for nan industry’s biggest tech news