Explaining it All: How We Became the Center of the Universe (Published 2011) and Self-Taught AI Shows Similarities to How the Brain Works | Quanta Magazine

Explaining it All: How We Became the Center of the Universe (Published 2011) and Self-Taught AI Shows Similarities to How the Brain Works | Quanta Magazine

In "Explaining it All: How We Became the Center of the Universe (Published 2011)," the author explores the idea that the European Enlightenment may have been the pivotal event not just in the history of the West or of human beings, but of the entire universe. The introduction of the scientific method allowed us to concoct and evaluate new hypotheses, giving us a sense that we could do anything. This mastery of knowledge and the ability to apply it to transform the world is what ultimately shapes our reality. The author argues that our ability to create new explanations is the uniquely significant thing about humans.

In "Self-Taught AI Shows Similarities to How the Brain Works | Quanta Magazine," the focus shifts to the field of artificial intelligence and how it relates to the functioning of the human brain. Traditional AI models rely on supervised learning, where algorithms are trained using human-labeled data sets. However, animals, including humans, learn by exploring their environment independently. Computational neuroscientists have begun exploring self-supervised learning algorithms that require little or no human-labeled data. These algorithms have been successful at modeling human language and image recognition, and they align more closely with brain function than their supervised-learning counterparts.

The similarities between the European Enlightenment and self-supervised learning algorithms are striking. Both emphasize the importance of exploration, the ability to evaluate and invent new ideas, and the transformative power of knowledge. The Enlightenment allowed humans to break free from irrational beliefs and embrace rationality, leading to progress and innovation. Similarly, self-supervised learning algorithms enable AI systems to gain a rich and robust understanding of the world by exploring and filling in the gaps in data.

One key insight from the comparison is that the brain primarily learns through self-supervised learning. Around 90% of what the brain does is driven by this process. The brain constantly tries to predict and anticipate what will happen next, allowing for efficient learning and adaptation. This aligns with the idea that our ability to create new explanations is what sets us apart as humans. We are constantly trying to predict and understand the world around us, just like self-supervised learning algorithms.

However, there are limitations to self-supervised learning and AI models. The brain is full of feedback connections that current models lack. To truly understand brain function, future research needs to incorporate these feedback connections and compare the activity of artificial neurons to individual biological neurons. Additionally, there is a need to explore highly recurrent networks in AI models to better match brain activity.

In conclusion, the European Enlightenment and self-supervised learning algorithms share common points regarding the importance of exploration, the power of knowledge, and the ability to create new explanations. Both have led to transformative changes in our understanding of the world. Three actionable pieces of advice that can be derived from this comparison are:

1. Embrace a mindset of exploration and curiosity. Just as the Enlightenment thinkers questioned established beliefs and sought new explanations, we should constantly seek to expand our knowledge and challenge existing ideas.

2. Be open to new ideas and embrace rationality. The Enlightenment emphasized the importance of rational thinking and scientific scrutiny. Similarly, in the age of AI, we should evaluate ideas based on evidence and critical thinking.

3. Foster a culture of self-supervised learning. Encourage independent exploration and curiosity in education and research. By allowing individuals and AI systems to learn through self-discovery, we can unlock new insights and innovation.

By combining the insights from these two articles, we can gain a deeper understanding of the transformative power of knowledge and the parallels between human intellectual progress and AI development. The European Enlightenment and self-supervised learning algorithms both highlight the importance of exploration, curiosity, and the creation of new explanations. These principles can guide us in our pursuit of knowledge and innovation in the modern world.