Recent statements of AI scientists and leaders has outlined the risks of AI for humanity. It is essential to understand different scenarios that are possible with AI development going beyond the boundaries of human ethics, morality, and safety.

It is my duty to inform that these instances are purely speculative. These are based on many assumptions about theoretical bounds of science behind AI, data size, abundance of computing power and entities working outside their legal framework.

Cyber Warfare and hacking: Powerful AI systems can specifically target individuals, organizations or nations to unleash sophisticated cyber-breaches. This may lead to disruption in the whole financial framework, erosion of privacy, severe threat to national security. Supply chains may be disrupted, power systems may be compromised, digital infrastructure may become standstill leading to intense economic damage. Self-replicating botnets may create havoc in the online space.

Flood of misinformation campaigns and fake content: AI filter bubbles and fake content can create unrest in already polarized society and escalate the chances of public unrest. It can diminish the trust in institutions, leaders and democratic processes. It can manipulate public opinion and eliminate the difference between generated content and ground truth. Divisive and inflammatory content will increase the incidents of violence, riots, and divisive demonstrations. It will lead to loss of human agency and decision making capabilities.

Genetic manipulation and reproduction: Digital tools are fast evolving that can accelerate the understanding of different genetic manipulations. With the help of CRISPR and other AI enabled algorithms can self-improve themselves and create powerful versions that can surpass human capabilities. Some of the Hollywood movies have shown this scenario leading to apocalypse like situation.

Nano Robots with Self-replication: With the advent of advance AI systems in the field of nanotechnology, there is a possibility of programming these machines to self-replicate. Nano Robots created with malicious intent have numerous scenarios to cause lethal impact to individuals. It can also affect the food chains or any other systems that rely on sensors or nano technologies.

Rogue Nations or individuals using AI to dominate: Unlike nuclear technology it is very difficult to detect if some rogue nations or organizations are trying to use AI to dominate the world. Few entities may deploy AI to finish their competition or gain more leverage in this world full of inequality. They may take over critical infrastructure of adversary like energy grid, communication networks, transportation systems, digital backbones, and classified product information. Entities who are naïve in AI will find it difficult to survive in such scenarios.

Self-improvement and Integrations: AI systems may be designed that innovate themselves with new algorithms leading to super intelligence. It will lead to intelligence explosion resulting in revolutionary discoveries that may be beyond human power to handle. AI systems may be integrated with Quantum Computing systems, distributed systems and cloud servers. Such robust compute capacity can result in spectacular brain power that can unleash the secrets of humanity, mortality and the ways to go beyond humanly possible boundaries. It can get to know the secrets hidden in the outer space, potential alien life and God element.

Autonomous weapons and Swarm Drones: Autonomous weapons may be developed that can be used without human intervention and can be used indiscriminately. Labs can use AI enabled tools to create advanced chemical and biological tools that have power to target specific populations or individuals. Self-destructing swarm drones, Suicidal submarines, remote controlled warships create unlimited scenarios of new age wars that will make the whole planet vulnerable in the hands of few rogue entities.

Loss of Human Autonomy: New generation becoming over dependent on AI enabled systems can lead to weakening of human agency and decision capabilities. This will make human race vulnerable and agile in front of these powerful technologies. There will be a huge cultural change. The benchmarks of morality and ethics will get new definitions. It will become difficult to distinguish between AI and Non AI systems and species.

This article first appeared in the TOI