Introduction
It is important to emphasize that artificial intelligence (AI) has experienced significant improvement and development in recent years. The purpose of the given essay is to logically analyze the risks of AI and the general public’s reaction to it. The context of the problem at hand is AI as a security threat with an existential risk dimension. As long as AI technology continues to develop at any pace, it will eventually surpass the collective human competencies in all domains, which means humanity will lose control over it; hence, it is a danger.
The Future of AI and Intelligence Expansion
Intelligence as Information Processing
To properly understand AI, it is helpful to define it first. AI is an artificial simulation of human intelligence, which means it can self-learn, self-improve, and carry out functions that humans can (Shchitova, 2020). From this, one can derive the first premise that intelligence is a matter of information processing conducted similarly to what human brains are capable of. Narrow AIs are already used in many domains; even social media platforms use AI to generate individualized feeds for each person.
Continuous Advancement of AI Technology
The second premise is that humanity will continue improving AI technology, including its software and hardware. Given how valuable technology and AI are for human beings, the only reason such a development would stop would be something that would end civilization (Barrat, 2023). Examples include famines, pandemics, nuclear wars, and climate catastrophes. If these events do not occur or carry a ‘civilization-ending’ outcome, the technology will continue to improve.
Human Intelligence Within the Intelligence Spectrum
The third premise is that human intelligence is not at the peak of the intelligence spectrum. In other words, human beings have not reached the high point of intelligence evolutionarily, and there is plenty of room for being even more intellectually capable. If these three premises are true, then humanity is bound to create a general AI that will be more intelligent than humans. This means it will experience an ‘intelligence explosion’ – a moment when AI will no longer need improvement input from humans but will do so itself (Barrat, 2023). It is when humanity will lose any form of control over it because our collective intelligence will no longer be able to constrain or control it.
The Inevitability of Uncontrolled AI Growth
The final premise is that humanity will also be unable to control or implement proper security measures before the explosion. The recent news at OpenAI provides good insight into how the current societal structures incentivize the reckless development of AI. Vallance et al. (2023) report that the board fired Sam Altman, CEO of a leading AI company, either due to security concerns or internal power struggles. Both possible causes are problematic because if there are security concerns, such drastic measures mean that AI risks are not being taken seriously at the top levels by people at the forefront of AI development. If it is a mere internal power struggle, these individuals are more concerned about money and power over AI safety.
Conclusion
In conclusion, the current AI development framework is commercially incentivized and shows no regard for AI safety. As long as the technology keeps improving, humanity will lose control over it. AI poses a serious risk and security problem for the entire global civilization. Therefore, it is up to the general public and their elected representatives to address the AI security concern adequately with strict and advanced policies.
References
Barrat, J. (2023). Our final invention: Artificial intelligence and the end of the human era. Quercus.
Shchitova, A. A. (2020). Definition of artificial intelligence for legal regulation. Proceedings of the 2nd International Scientific and Practical Conference on Digital Economy (ISCDE 2020), 156, 616-620. Web.
Vallance, C., Liang, A., & Kleinman, Z. (2023). OpenAI staff demand board resign over Sam Altman sacking. BBC. Web.