Human Expertise in the Age of ChatGPT

ChatGPT has generated an unprecedented level of excitement, captivating individuals and showcasing the remarkable capabilities of large language models. One company, Edtech Chegg, saw its shares drop more than 40 percent after the company says ChatGPT is killing its business.

However, while ChatGPT is indeed fascinating and has sparked a surge of public interest, we should be cautious not to idealize its abilities or attribute overly human-like qualities to it. While its responses may appear natural, it is crucial to maintain a clear distinction between the virtual and the real.

Indeed, although ChatGPT represents a significant advancement in improving the accessibility, utility, and refinement of AI models, it is essential not to forget that:

1. It is not a revolutionary technology breakthrough, but it builds upon known techniques that have been incrementally optimized over the years.
2. Language models operate using probabilities and can be seen as stochastic parrots; they lack planning capabilities or consciousness, relying solely on statistical patterns in data.
3. ChatGPT excels at generating smooth and engaging responses, aiming to please users.

A Human Revolution

While ChatGPT undoubtedly represents a significant milestone, it is crucial to maintain a realistic understanding of its capabilities and limitations. By doing so, we can appreciate its contributions while avoiding potential misconceptions and confusion.

It is also important to address public perceptions and shed light on the revolution unfolding—a revolution driven by human actions more than technological advancements. It is not the technologies themselves that reshape societies but rather our reappropriation and utilization of them that enable societal evolution. By excessively focusing on potential threats, we risk overlooking the incredible opportunities for personal and collective growth.

According to research, individuals who derive value from AI models report an enhanced sense of competence and autonomy. Moreover, human-algorithm synergies have been shown to outperform standalone algorithms and even human experts relying solely on their judgment, primarily due to mutual enrichment.

These observations call for a shift away from debates centred on the notion of competition between humans and machines. Such debates often arise from fictional narratives that blur the lines between intelligence and dominance.

Rather, our focus should be on the concept of distributed cognition, investigating the evolution and indispensability of human expertise in an augmented world. Embracing the transformative potential of AI models and recognizing the collaborative relationship between humans and machines can enable us to shape a future that leverages the best of both worlds, amplifying human capabilities while reaping the benefits of technological advancements.

Asking the Right Questions

Generative AI primarily operates in a responsive manner. While it is fascinating to marvel at the awe-inspiring answers generated by ChatGPT, a more insightful perspective lies in examining the input provided.

In other words, the precision and informativeness of the answers are contingent upon the context and the quality of the questions asked. Regrettably, we often possess a limited ability to formulate the most effective questions.

Traditionally, expertise has been defined as the accumulation of an individual's knowledge. However, the rise of popular AI models paints a different picture—a world where the importance lies not solely in what we know, but rather in what we desire to know.

Today and in the future, human expertise derives its essence and form from our critical thinking abilities and our aptitude for problem structuring. In other words, these models act as powerful tools that can assist us in our quest for knowledge and problem-solving, but they rely on our ability to frame and structure the questions accurately. Ultimately, human intellect and discernment continue to play a crucial role in leveraging AI effectively and uncovering meaningful insights.

Epistemic curiosity, which refers to the innate desire for knowledge that drives individuals to learn new ideas, overcome informational gaps, and solve intellectual problems, is, therefore, crucial for our adaptation and expertise. By demonstrating intellectual humility and metacognition—being aware of both our knowledge and our limitations—and by asking pertinent questions while critically and reflectively assessing the answers, we can effectively utilize AI models and foster our personal growth and performance.

For example, research conducted by MIT researchers, which examines the impact of ChatGPT on the performance of skilled workers in writing tasks, revealed several notable findings. The study demonstrated that ChatGPT allowed individuals to:

1. complete tasks more efficiently
2. generate content that was assessed to be of superior quality in terms of writing style, substance, and originality
3. enhance worker satisfaction with task completion

Artificial Intelligence Essential Reads

Using AI to Enhance Your Email Communication as a Leader

The Rise of the Techno-Energy Complex

Additionally, the tool proved to be particularly beneficial not only for highly skilled workers, who were able to accomplish tasks more swiftly but also for individuals with initially lower capacities, allowing them to enhance the quality of their output. As a result, the performance gap between the least and most proficient workers was reduced.

What's more, although the future replacement of human activities by AI models is uncertain, it is evident that individuals who possess the skills to effectively utilize AI will gain a competitive advantage. For example, new marketplaces are emerging where individuals can buy and sell "prompts." This trend has given rise to the potential profession of "prompt engineers" who specialize in designing and crafting prompts for AI models, such as ChatGPT—and that can pay up to $375k, according to some reports.

ChatGPT has shaken our practices and pushed our intellectual and creative boundaries. While this development may seem harsh to the non-expert public, it should not generate fears due to scientific ignorance or media misinformation.

As we move forward, nurturing our epistemic curiosity, practising intellectual humility, and honing our metacognitive abilities are essential. By doing so, we can engage with AI models in an informed and purposeful manner, leveraging their capabilities to facilitate our personal and professional growth.

Instead of comparing themselves to the machine or wrongly banning it, everyone should be curious and build a useful mental model of their abilities: How can AI models complement my practice? How can they help me on a daily basis? How can I enrich them? What are their biases and limits?

ChatGPT has opened up extraordinary opportunities for everyone to access knowledge and is intended to be an exceptional learning lever. It is up to us to transform these assets into concrete and beneficial actions. This is our share of humanity and expertise.