Artificial Intelligence

Overview

In GATA and its partner states, AI is strictly regulated under the NDA. General purpose processors are illegal specifically due to the threat posed by AI and destructive algorithms like the Daemon virus.

This limitation drove the development of Asimovian computers; application-specific compute platforms that are secure and predictable. As a result of this move toward hard code, all legal artificial intelligence today takes the form of cogs, Asimov-complete hard-coded cognitive systems that can only perceive and compute within a carefully limited scope called an ‘ontology’.

---

AI in the Old World

Even before the Dark Decade, despite its many benefits, AI had remained a contentious issue as it had precipitated an acceleration in economic and political volatility, ubiquitous form of social spying, and an autonomous market for private data, resulting in numerous societal clashes and scandals.

While for many in the Old World, a resigned acceptance had set in, others protested and suffered from blacklisting and criminalization in many areas of society with zero transparency into the processes behind these actions. The popular sentiment among Atla’s founding citizens was that AI, together with a malignant form of short-sighted competition, had created the conditions for The Crash.

AGI in the Old World

While AGI had proven itself to be more reliable and inherently ethical than simply AI agents, its multimodal and cross-domain cognition made auditing or tuning its decision making nearly impossible. AGI's wholistic approach to problem solving made its implementation in democracies and private enterprise untenable for administrators and politicians beholden to populist anti-AI movements.

While AGI was intentionally hobbled by denying it ultimate control, considerable advances were still being made year over year. The promise of ASI (artificial super-intelligence) continued to interest researchers, however its imminence seemed to perpetually recede as they approached it, revealing the shape of the ultimate problem; an every expanding taxonomy of 'kinds' of thinking, morphisms which proved challenging to even define let alone instantiate in code. Researchers came to believe that consciousness was the key.