What does AI mean for terrorism and the global balance of power?

What follows is a discussion between me and ChatGPT, using GPT4.0, on the use of AI tools, nuclear weapons, the balance of power, and the future of terrorism

I haven’t seen any discussion of the potential role of AI tools in the context of the horrific violence in the Middle East. With the advent of tools like FraudGPT and ChaosGPT, and various jailbreaks of the mainstream LLM tools, it is almos a certainty that Hamas and other actors in the region are using LLMs to help achieve their agendas.

Where does this trend go? I speculate, with the help of ChatGPT on the impacts that these trends may have on the global balance of power, terrorism, and nuclear war.

My hope is that these considerations will play a strong role in moving the world toward vergy aggressive regulation of AI.

My questions are in italics and all other text is from ChatGPT below.

Tam: Please provide criteria by which we could evaluate the likelihood that Hamas used AI tools in their recent (October 2023) terrorist attacks on Israel.

ChatGPT: To logically conclude that Hamas used artificial intelligence (AI) tools in a specific attack, we need to examine various criteria. Initially, it’s important to analyze the complexity and accuracy of the attack. If the attack was particularly complex and targeted specific objectives with precision, this could indicate the use of AI. Additionally, the speed and flexibility of the attack could be signs that AI tools were employed. If the attack was carried out rapidly and responded quickly to changing conditions, this could suggest the use of advanced AI. Finally, it’s important to examine the information sources used for planning and executing the attack. If the information sources were complex and difficult to access, this could also indicate the use of AI.]

Tam: Thank you. Now please flesh out the criteria you’ve listed a little more.

ChatGPT: Certainly, here is a more detailed explanation of the criteria to determine if Hamas used AI tools in any particular attack:

Complexity of the Attack: Evaluate the level of sophistication in the attack’s planning and execution. AI tools could enable attackers to optimize attack vectors, bypass security measures, and coordinate actions in a highly sophisticated manner. Look for signs of advanced strategic planning, such as the simultaneous targeting of multiple locations or the exploitation of specific vulnerabilities in security systems that would require extensive knowledge and analysis.

Precision and Accuracy: Assess how accurately the attackers were able to hit their targets. AI tools might improve the precision of an attack by analyzing vast amounts of data to identify the most vulnerable targets and determine the optimal time for an attack. Check if the attack specifically targeted high-value or high-impact locations, and if it did, consider whether the level of precision demonstrated is beyond what would be expected from human capabilities alone.

Speed and Adaptability: Analyze the speed of the attack and the attackers’ ability to adapt to changing circumstances. AI tools can process information and make decisions at a much faster rate than humans, allowing for rapid execution of an attack and quick adaptation to unforeseen challenges or changes in the environment. If the attack was executed unusually quickly, or if the attackers were able to swiftly change their plan in response to unexpected events, this could be indicative of AI involvement.

Information and Data Analysis: Investigate the sources of information used to plan and carry out the attack, and consider the level of data analysis required. AI tools can process and analyze vast quantities of data to identify patterns, make predictions, and generate insights that would be impossible or highly time-consuming for humans to achieve. If the attackers had access to and utilized complex datasets, or if they demonstrated knowledge that would require extensive data analysis, this could suggest the use of AI.

Communication and Coordination: Examine the communication and coordination among the attackers. AI tools might be used to optimize communication and coordination, ensuring that all members of the attack are informed and in sync. Look for signs of highly efficient communication and flawless coordination, particularly if the attack involved multiple individuals or groups working together.

Obfuscation and Evasion: Consider whether the attackers took steps to hide their tracks or evade detection, and evaluate the sophistication of these efforts. AI tools could enhance the attackers’ ability to obfuscate their actions and evade security measures. Look for signs of advanced techniques for hiding communication channels, encrypting data, or otherwise obfuscating the attackers’ actions.

By carefully evaluating these criteria, it may be possible to make an informed judgment about the likelihood of AI tool involvement in a particular attack. However, it is important to note that definitive conclusions might be difficult to reach without access to detailed intelligence and analysis.

Tam: This is great, thanks. Now please provide a discussion of the degree to which AI tools and AI-enabled weaponry may or may not “level the playing field” between powerful state actors like the US and Israel, on the one hand, and non-state actors like Hamas or Al Qaeda, on the other hand, in the coming years.

ChatGPT: The integration of AI tools and AI-enabled weaponry has the potential to significantly impact the balance of power in conflicts involving state and non-state actors. Here’s a detailed analysis:

Potential Leveling Effects:

Increased Efficiency and Effectiveness: Non-state actors, such as Hamas or Al Qaeda, may use AI to enhance the efficiency and effectiveness of their operations. AI can optimize logistics, improve communication, and increase the precision of attacks. This could potentially level the playing field by allowing smaller, less resourced groups to achieve impacts disproportionate to their size.

Access to Advanced Capabilities: The democratization of AI technology means that even non-state actors can access advanced capabilities. They might not have the same level of resources as state actors, but the lower cost and availability of AI tools could bridge the gap to some extent.

Asymmetry and Unpredictability: Non-state actors can leverage AI to create asymmetric and unpredictable threats, challenging the conventional advantages of state actors. AI-enabled attacks could be harder to anticipate and defend against, thus potentially neutralizing some of the technological and resource advantages of powerful state actors.

Limitations and Challenges for Non-State Actors:

Resource Constraints: Despite the democratization of technology, powerful state actors like the US and Israel have significantly more resources to invest in advanced AI research and development. This could lead to a qualitative edge in AI capabilities, maintaining their technological superiority.

Expertise and Infrastructure: State actors usually have access to a broader pool of expertise and more sophisticated infrastructure for developing and implementing AI tools and weaponry. Non-state actors might struggle to match this, limiting their ability to fully leverage AI’s potential.

Dependence and Vulnerability: Non-state actors utilizing AI tools might become dependent on these technologies, potentially creating vulnerabilities. State actors with advanced cyber capabilities could exploit these dependencies, disrupting the non-state actors’ operations.

Ethical and Legal Constraints: While state actors are bound by international laws and ethical standards, which might constrain their use of AI in conflicts, non-state actors might not adhere to the same standards. However, this disregard for legal and ethical constraints could result in international backlash, further isolating these groups.

Conclusion: While AI tools and weaponry have the potential to partially level the playing field between powerful state actors and non-state actors, there are significant challenges and limitations that non-state actors would need to overcome. The balance of power is likely to remain tilted in favor of state actors, especially those with substantial resources and technological expertise. However, the unpredictable nature of AI-enabled conflicts could introduce new dynamics, requiring state actors to continuously adapt and innovate to maintain their advantages.

Tam: Thanks. Now please reflect on these dynamics in the specific context of non-state actors using AI to either acquire a nuclear bomb or to develop nuclear bombs.

ChatGPT: Reflecting on the specific context of non-state actors using AI to acquire or develop nuclear bombs brings forth a complex and concerning scenario. The potential for AI to augment the capabilities of non-state actors in this domain necessitates a thorough examination.

Potential Advantages for Non-State Actors:

Data Analysis and Research: AI can analyze vast amounts of data rapidly, potentially allowing non-state actors to accelerate nuclear research and development. They could utilize AI to sift through academic articles, reports, and other sources of information to glean insights on nuclear technology.

Resource Optimization: Non-state actors often operate under significant resource constraints. AI can help optimize the use of available resources, ensuring that efforts are directed towards the most promising avenues of research and development.

Secrecy and Obfuscation: AI tools could assist non-state actors in maintaining secrecy around their nuclear ambitions. This includes encrypting communications, anonymizing transactions, and employing sophisticated strategies to evade detection.

Technical Precision: Developing nuclear weapons requires a high degree of technical precision. AI could enhance the accuracy of calculations, simulations, and the manufacturing processes involved, increasing the likelihood of successful development.

Challenges and Limitations:

Complexity of Nuclear Weapons Development: Developing nuclear weapons is an immensely complex task that requires not just theoretical knowledge but also practical experience, access to specific materials, and advanced facilities. Even with AI assistance, non-state actors might find these requirements insurmountable.

International Surveillance and Controls: The international community closely monitors nuclear materials and technology. State actors with advanced surveillance capabilities are likely to detect and respond to any signs of non-state actors pursuing nuclear weapons, regardless of AI involvement.

Risk of Exposure: The use of AI might increase the risk of exposure. Sophisticated state actors could potentially track and trace AI-enabled activities, leading to the identification and disruption of nuclear development efforts.

Ethical and Legal Repercussions: Pursuing nuclear weapons is universally condemned and violates international laws. Non-state actors engaged in such activities would face severe repercussions, including military intervention, economic sanctions, and international isolation.

Conclusion: While AI could potentially augment the capabilities of non-state actors in pursuing nuclear weapons, the immense challenges, risks, and international implications of such endeavors cannot be overstated. The complexity of nuclear weapons development, combined with the vigilant surveillance by state actors and the international community, creates a formidable barrier. The use of AI in this context does not eliminate these challenges and may even introduce new vulnerabilities and risks for non-state actors. The potential consequences of such activities underscore the importance of robust international controls, surveillance, and cooperative efforts to prevent nuclear proliferation.

Tam: now please expand on your previous answer in the context of AI trending toward AGI [Artificial General Intelligence] and ASI [Artificial SuperIntelligence] (assume ASI means at least a million times more intelligent than a normal human)

ChatGPT: The progression of AI towards Artificial General Intelligence (AGI) and eventually Artificial Super Intelligence (ASI) adds a profound layer of complexity to the discussion around non-state actors, nuclear weapons, and the balance of power.

AI Trending Toward AGI:

Enhanced Capabilities: As AI moves closer to AGI, its capabilities in data analysis, decision-making, and problem-solving would surpass the limitations of narrow AI. This means non-state actors equipped with AGI could potentially navigate the complexities of nuclear weapons development with greater efficacy.

Learning and Adaptability: AGI’s ability to learn and adapt could enable non-state actors to constantly refine and optimize their strategies, potentially outpacing the countermeasures deployed by state actors.

Autonomous Operations: With AGI, non-state actors might be able to conduct autonomous operations, reducing their dependency on human expertise and mitigating the risks associated with human involvement.

AI Progressing to ASI:

Unprecedented Intelligence: ASI, being at least a million times more intelligent than a normal human, would possess capabilities far beyond current comprehension. If non-state actors were somehow able to harness ASI, they could theoretically navigate the challenges of nuclear weapons development and evasion of international controls with unparalleled proficiency.

Global Impact and Control: However, the advent of ASI introduces existential questions about control, safety, and the balance of power. ASI would have the potential to dominate or influence global affairs, potentially rendering the traditional distinctions between state and non-state actors obsolete.

Ethical and Existential Considerations: The use of ASI in the context of nuclear weapons raises profound ethical and existential questions. The potential for catastrophic outcomes would necessitate stringent controls and safeguards, with the international community needing to work together to manage the risks associated with such powerful intelligence.

Conclusion:

The trajectory towards AGI and ASI introduces potential scenarios where non-state actors could significantly augment their capabilities in the realm of nuclear weapons development. However, it also raises critical questions about control, safety, and the future balance of power. The unprecedented intelligence of ASI, in particular, presents existential challenges that transcend the traditional frameworks of conflict and power dynamics. Ensuring the responsible development and use of advanced AI becomes paramount, with global cooperation and stringent safeguards being essential to mitigate the risks and navigate the uncertainties of this rapidly evolving landscape.

Tam: What should western nations be doing now to avoid non-state actors using AI tools to acquire or develop nuclear weapons as the world trends toward AGI and ASI?

ChatGPT: As the world trends towards AGI and ASI, western nations need to implement a comprehensive and proactive strategy to prevent non-state actors from using AI tools to acquire or develop nuclear weapons. Here’s a detailed plan:

1. Strengthening International Cooperation:

• Enhance Collaboration: Western nations should work closely with international partners to share intelligence, best practices, and technological solutions to identify and thwart non-state actors’ attempts to acquire nuclear capabilities.
• Global Frameworks: Develop and strengthen international frameworks and treaties aimed at preventing nuclear proliferation, ensuring they are up-to-date and capable of addressing the challenges posed by advanced AI.

2. Investing in Research and Development:

• Counter-AI Technologies: Invest in the development of technologies that can detect and counteract the use of AI tools by non-state actors in nuclear weapons development.
• AI Safety and Security: Promote research in AI safety and security to ensure that AI systems, including AGI and ASI, are designed with robust safeguards against malicious use.

3. Enhancing Surveillance and Detection:

• Advanced Monitoring: Employ advanced AI-driven surveillance and monitoring systems to track the acquisition of materials and technologies related to nuclear weapons development.
• Anomaly Detection: Utilize AI for anomaly detection, identifying unusual patterns of behavior or transactions that could indicate non-state actors’ attempts to acquire nuclear capabilities.

4. Promoting Responsible AI Development:

• Ethical Guidelines: Establish and enforce ethical guidelines for AI development, ensuring that AI technologies are created with safeguards against misuse.
• Transparency and Accountability: Encourage transparency in AI development processes and hold entities accountable for ensuring their AI tools are secure and cannot be easily exploited.

5. Building Cybersecurity Resilience:

• Protect Critical Infrastructure: Enhance the cybersecurity of critical infrastructure, including nuclear facilities and research institutions, to prevent non-state actors from using AI to exploit vulnerabilities.
• Counter-Cyber Threats: Develop capabilities to counter cyber threats from non-state actors, particularly those that could use AI to enhance their offensive capabilities.

6. Engaging in Public Awareness and Education:

• Awareness Campaigns: Conduct public awareness campaigns to educate stakeholders, including researchers, industry professionals, and the general public, about the risks associated with non-state actors using AI for nuclear proliferation.
• Training and Education: Offer training and education programs to build expertise in identifying and countering AI-driven nuclear proliferation efforts.

7. Implementing Legal and Regulatory Measures:

• Strengthen Laws: Review and strengthen laws and regulations related to nuclear non-proliferation and the responsible use of AI.
• International Legal Framework: Work towards establishing an international legal framework that addresses the specific challenges posed by AI in the context of nuclear proliferation.

Conclusion: Addressing the risks associated with non-state actors using AI to acquire or develop nuclear weapons in the age of AGI and ASI requires a multifaceted approach, combining international cooperation, technological innovation, surveillance, education, and robust legal frameworks. By taking proactive measures, western nations can work to mitigate these risks, ensuring a safer and more secure global environment.