HUMAN INTELLIGENCE AND
ARTIFICIAL INTELLIGENCE: A
SYMBIOTIC RELATIONSHIP OR A
RIVALRY?
Perceptions and Attitudes Toward Machines
The 4th Industrial Revolution (4IR) is a digital revolution that merges physical, digital, and biological technologies. Key innovations like AI, robotics, and IoT are driving us toward a highly automated world. This shift has raised concerns about automation replacing human labor, leading to widespread fears that “AI will replace humans.” As a result, the potential for increased inequality between low-skilled and highly-skilled workers has become a major societal concern[1].
Currently, discussions and workshops are focusing on the 5th Industrial Revolution (5IR). Unlike the 4IR, the 5IR emphasizes collaboration over competition. It recognizes the unique strengths of both humans and machines, promoting their complementary use for greater efficiency and effectiveness[2]. This perspective challenges the idea that technology will replace humans and suggests that the 5IR aims to address the shortcomings of the 4IR by putting humans back in the driving seat.
The Human and Artificial Intelligence Relationship
Despite the 5IR’s emphasis on collaboration, the hype around AI’s ability to perform tasks traditionally done by humans continues to fuel negative perceptions and attitudes toward machines. This hype often overshadows AI’s limitations and the necessity for human collaboration. For instance, while open-source AI models can outperform top mathematicians on some problems, expensive generative AI models still struggle with basic tasks, such as accurately generating a human hand with five fingers. A viral social media post even highlighted criminals using prosthetic fingers to make surveillance footage appear AI-generated and thus inadmissible in court.[3]
Human and artificial intelligence each excel in different areas. AI can process data and provide insights rapidly, but human intelligence offers empathy, which is crucial for building trust, especially in vulnerable situations like conflict settings. Therefore, a symbiotic relationship between the two is essential.
In 2000, financial technology company PayPal was losing up to $10 million monthly to credit fraud. The volume of transactions made it impossible for humans to detect fraudulent activity alone. Initially, PayPal used an algorithm to tackle the problem, but fraudsters adapted to new tactics. Ultimately, PayPal adopted a hybrid approach where computers flagged suspicious transactions for human operators to review. This strategy helped PayPal achieve its first quarterly profit after a $29.3 million loss the previous year[4]. This example demonstrates the benefits of a symbiotic relationship between human and artificial intelligence.
In my experience building applications using large language models (LLMs) like GPT-4 and LLaMA3, I have witnessed the importance of human intelligence working alongside AI. A prime example is crafting prompts to guide an LLM to produce the desired output. According to W3Schools, prompt engineering involves creating input (usually text) to instruct the generative AI to generate a desired response[5]. This definition illustrates how human intelligence guides AI towards desirable outcomes.
As far as our current technology is concerned, actionable insights still rely heavily on human intelligence.
Layers of the Human Intelligence and Artificial Intelligence Symbiotic Relationship
To determine the various layers of the HINT — AI symbiotic relationship we have to look at tasks and the elements of a task. In the context of a conflict setting, these elements include the
- complexity of a task,
- the requirement for creativity and intuition,
- ethical implications,
- need for empathy and social interaction,
- adaptability to new or unforeseen situations,
- and consequences of errors.
Complexity of the Task: How complicated the task is given the number of variables, interactions, and the level of difficulty involved in completing the task. For example, automating document analysis and summarizing positions of different parties has low complexity, while negotiations involve understanding multiple perspectives, balancing competing interests, and addressing various political, social, and economic factors[6] has high complexity.
Requirement for Creativity and Intuition: refers to the need for innovative thinking and the ability to make decisions based on experience and instinct rather than just data. Creating programs that effectively heal divisions within a community requires innovative approaches that consider cultural nuances and historical contexts[7] while sorting meeting emails on the same does not.
Ethical Implications: The degree to which the task involves moral judgments and the potential impact on ethical standards and principles. For instance, decisions about who receives aid and how it is distributed involve ethical considerations, such as prioritizing the most vulnerable and ensuring fairness[8], unlike routine data processing.
Need for Empathy and Social Interaction: the necessity for understanding, responding, and sharing the feelings of others, and meaningfully interacting with people. Counseling requires high empathy, the ability to build trust, and social interaction[9], while data analysis does not.
Adaptability to New or Unforeseen Situations: refers to how well can the task be handled when unexpected changes and new conditions occur. Rapidly changing security situations require flexible and adaptive responses to protect civilians and manage the fallout[10]. The Paypal fraud detection case above is also a good example of how to go about a task that needs to be adaptive and flexible to new circumstances.
Importance of Contextual Understanding: the requirement for grasping the broader context and nuances of a situation to make informed decisions. A seasoned peace-building and conflict specialist understanding a region or country background to tailor interventions requires contextual understanding, unlike automated report generation, maps, and analysis of local power structures using data.
Consequences of Errors: refers to the potential impact and severity of mistakes or errors made in the task. Tasks where errors can have significant negative consequences typically require human oversight to mitigate risks. Incorrectly handling information can exacerbate tensions, expose the already vulnerable, and lead to violence, making accuracy and careful management crucial. Whereas minor typos in a casual email do not.
The above elements are essential for optimizing the effectiveness and ethical integrity of human intelligence and artificial intelligence collaboration. Having explained the elements the following are the possible layers of human intelligence and artificial intelligence collaboration in the execution of a given task:
1. AI Only — indicates tasks where AI can effectively perform without any human intervention across all elements.
2. AI led with Human Intelligence guidance — demonstratest tasks where human intelligence guides AI across the various elements.
3. Human Intelligence with AI assistance — tasks where AI supports human decision-making under human leadership across the various elements.
4. Human Intelligence Only — indicates tasks where only human intelligence is required for task execution, without any necessary AI involvement.
Below is a matrix developed to visualize the application of this symbiotic relationship.
Table 1: Human Intelligence and Artificial Intelligence Collaboration Layers
Conclusion
The collaboration between HINT and AI in task execution is multifaceted, each bringing distinct advantages. While AI excels with low to medium complexity and predictable outcomes, HINT creativity, empathy, and adaptability shine in contexts requiring nuanced decision-making and social interaction. By recognizing and leveraging the strengths of both, we can achieve a symbiotic relationship that optimizes efficiency, effectiveness, and ethical integrity. Through careful consideration of task elements and detailed collaborative frameworks as to how HINT guidance or AI assistance comes into play, we pave the way for HINT and AI to work seamlessly together to tackle complex challenges and bring about positive impact.
References
[1] World Economic Forum (2016)
[2] Stephanie M. Noble et al. (2022) The Fifth Industrial Revolution: How Harmonious Human–Machine Collaboration is Triggering a Retail and Service [R]evolution
[3] Alexander Gounares (2023) Our legal system is about to get overhauled.
[4] Peter Thiel (2014) Zero to One: Notes on Startups or How to Build the Future
[5] W3Schools Generative AI Prompt Writing Introduction
[6] William Zartman (2007) Peacemaking in International Conflict: Methods & Techniques
[7] Popova, Z (2009) The role of social capital for post-ethnic-conflict reconstruction
[8] Hugo Slim (2015) Humanitatian Ethics: A Guide to the Morality of Aid in War and Disaster
[9]Judith Herman (1992) Trauma and Recovery
[10] Arnold Howitt & Herman Leonard (2009) managing Crises: Responses to Large-Scale Emergencies
