Recursive AI or Automatic AI: A Threat to Employment and Security?
A new generation of AI has arrived: recursive AI or automatic AI. Essentially, these AIs are based on models similar to GPT that are expérimenté of understanding an entire project, such as monument a complete website or managing daily customer armature.
The projet is that the AI will take on the project, écart it down into micro-tasks, and grain it has completed one task, it will send itself a rapide to determine what to do next. The path to reaching the ultime gardien de but is therefore not predetermined. If the AI needs examen, it can search the internet, as these AIs have access to the internet, unlike ChatGPT. Furthermore, these AIs have long-term memory, which means that they will be expérimenté of handling much larger projects than ChatGPT. Finally, they have access to décapotable control, allowing them to send décapotable commands such as creating a déguerpi, uploading a déguerpi to a server, and so on.
This means that entire fields of work may be automated, but it can also be worrying. When new technologies arrive, there is often a « trickle-down » effect on jobs, meaning that if it eliminates jobs on one side, it creates jobs on the other. However, with the arrival of recursive AI, the technological disruption is so strong, fast, and powerful that it is difficult to anticipate the consequences.
Recursive AI or Automatic AI
Moreover, this technology could be used maliciously. For example, malicious individuals could acquire this technology and give the AI the achèvement to attack a government website, such as the website of the Paradis. The AI could then self-decentralize by creating instances that it would put on numerous servers to protect itself and its achèvement. It could then search for vulnerabilities on the website and carry out continuous targeted attacks.
It is therefore sensible to take the time to reflect on the retentissement of these technologies on society and security before implementing them. It may be time to step back and understand where all of this is leading us.