Autonomous AI agents, what about us?
Focus of the letter 14
Two focuses this week with II Agent and Antigravity for their multiple features grouped into a single application, made possible by the deployment of agents.
The agents could be compared to a sports team with their captain who, after analyzing the situation, here our instruction, will put in place a strategy – often as in Antigravity called “planning” – to achieve the set objective. To do this, players, the agents, will be deployed, each with a specific role.
The team will act entirely or partially autonomously, without necessarily the coach, in this case the human behind the machine, intervening.
More than on the final score of the team – the result offered by the application -, it is on this autonomy that our attention should certainly focus as a priority: the result can be of high quality from the first iteration, but have we understood the generation and are we continuing to try to understand it?
The English term “agency” is used in English for these agents, agency which can be translated as “agency” (defined on Wikipedia as “property of an individual, a machine or a group to be the source of actions having effects on the world”, a term that will already speak to some who read these lines ;).
“Agency” is therefore used for both humans and machines: I think that in the face of these increasingly complex and efficient applications, it is up to us to be vigilant in order to, individually and collectively, keep and fully express our agency without letting that of the machine take up all the space…