Cognitive Bias, Bull5hit Asymmetry Principle, and AI

AI is all over the place. You noticed it, of course! Even, yesterday, my dentist was sharing his thoughts about AI. Everyone has an opinion about AI and advice on how to use AI... Dunning-Kruger's at play! But there's worse: people who have a captive audience and share their thin knowledge with assertiveness and a form arrogance. Dunning-Kruger, peppered with Brandolini's law.

Knowing the principles of Brandolini's law and the Dunning-Kruger effect provide critical lenses through which we can evaluate the hype and challenges associated with emerging technologies like AI, generative AI, and quantum computing. My goal here is to make you smile and empower you with a foundation for applying critical thinking to the discourse surrounding these technologies.

Brandolini's Law in the Realm of Emerging Technologies

Brandolini's law, or the bullshit asymmetry principle, underscores a significant challenge in the digital age: the disproportionate effort required to refute baseless information compared to the ease of producing it. This principle is particularly relevant in the context of artificial intelligence and other emerging technologies, where sensational claims and misinformation can spread rapidly through social media and other digital platforms. Jevin West's lecture on the pitfalls and power of generative AI illustrates this challenge vividly, noting the difficulties in dispelling misinformation propagated by AI technologies themselves.

In the workplace and technology sectors, Brandolini's law manifests through the rapid spread of misleading interpretations of market trends, overhyped modern technologies, and simplistic solutions to complex challenges. Leaders and professionals must navigate these waters by promoting a culture of verification and critical thinking, thereby countering the spread of misinformation and fostering an environment where truth and clarity thrive.

The Dunning-Kruger Effect and Technological Overconfidence

The Dunning-Kruger effect, which describes the phenomenon where individuals with limited knowledge overestimate their understanding, is particularly pertinent in discussions about AI, generative AI, and quantum computing. In these domains, where the complexity and nuances of the technology are often understated, this cognitive bias can lead to overconfidence in simplistic solutions or the capabilities of emerging technologies. The result is a landscape where critical and informed discussions are overshadowed by overly optimistic or uninformed viewpoints.

If, on one side, technologies are widely discussed, very hype, and look "easy" and accessible on the outset, they're highly technical and complex, and require some deeper technical knowledge to be well understood and used. Not even mentioning ethical and unbiased usage. In all cases, it's a long tail game that has started over 70 years ago with the seminal paper from Alan Turing and will continue in the years to come. As it took over 30 years for the internet to mature (commercially), it may take years for current AI technologies to mature as well.

Countering Hype with Critical Thinking

The amount of energy needed to refute bullsh1t is an order of magnitude larger than to produce It.

The challenge of countering hype in our technological fast-paced world, as discussed in the context of Brandolini's law, serves as a pertinent example of the broader issue at hand. Experts face an uphill battle in providing nuanced analysis against grandiose but unsupported claims about the potential of emerging technologies (IMHO one of the most "incoherent babbling strewn with scientific terms" comes from quantum healing, led by gurus like Deepak Chopra, but many others act in the current fields of Gen AI and Quantum Computing). This dynamic emphasizes the necessity of critical evaluation and the importance of credible expertise in steering public discourse towards a more informed and realistic understanding of technology's capabilities and limitations.

Empowering Critical Evaluation in Technological Discourse

To navigate the complexities of emerging technologies effectively, individuals and professionals alike must:

• Foster a culture of skepticism and verification, ensuring claims are critically evaluated and backed by credible evidence. For instance, the rapid adoption of AI and machine learning tools in various sectors has led to bold claims about their capabilities. A practical step towards fostering skepticism would involve implementing peer reviews and audits of AI-driven projects, akin to how AI ethics boards review the deployment of new algorithms. This process helps ensure that claims about an AI system's efficiency, fairness, and safety are not just taken at face value but are empirically tested and validated against established benchmarks.
• Encourage the development of critical thinking skills, enabling the assessment of technologies beyond surface-level hype. An example of this in action is the educational initiatives aimed at demystifying blockchain technology. Initially surrounded by considerable hype, blockchain has been subject to inflated claims regarding its potential to revolutionize virtually every industry. Through workshops, courses, and seminars that unpack the technical workings, limitations, and realistic use cases of blockchain, individuals are better equipped to assess its actual value and applicability to different domains. Another example is coding bootcamps and computer science curriculums incorporating units on ethical programming and the social implications of technology, thus preparing students to critically evaluate the broader impact of their work.
• Promote transparency and openness in discussions about technological capabilities, limitations, and ethical considerations. For example, the discourse around quantum computing often oscillates between extreme optimism and skepticism. By promoting transparent communication from researchers and companies about the current state of quantum technology, its near-term potentials, and the significant hurdles that remain, a more accurate picture emerges. This includes acknowledging the "quantum supremacy" milestones while also clarifying that practical, widespread applications of quantum computing are still years away. Similarly, AI developers sharing models' limitations and error rates helps demystify AI's capabilities, fostering a healthier, more informed public dialogue.
• Engage with and support credible sources of information, prioritizing expert analysis and evidence-based perspectives over sensationalized media coverage. This can be seen in the initiative by many academic institutions and tech companies to make research findings and datasets publicly available, encouraging independent verification and further research. Additionally, platforms like ArXiv and Google Scholar allow free access to pre-published research papers, facilitating a broader peer review process and enabling enthusiasts, scholars, and professionals to stay informed with the latest credible advancements in fields like AI, machine learning, and other emerging technologies. Encouraging the use of fact-checking websites and databases (like Snopes.com or AFP Fact Check, amongst many others) that track the accuracy of public statements made by technology companies and leaders also plays a crucial role in distinguishing fact from hype.

By applying the principles of Brandolini's law and the Dunning-Kruger effect to the discourse surrounding AI, generative AI, quantum computing, and other emerging technologies, we can cultivate a more informed, critical, and nuanced understanding of their potential and challenges. This approach not only guards against misinformation and overconfidence but also encourages a more responsible and ethical engagement with the technologies that are shaping our future.