How We The Humans Build

👷‍♂️✈️🖥️

🧮 Computers find it easier to solve complex math than identify a fruit in a basket.
🧠 Calculators were developed earlier than computer vision–capable machines.
👁️‍🗨️ Humans can do computer vision tasks much better than complex math.

🌍 Species evolve out of nature, geography, and the need to replicate.
🛠️ Humans build things out of necessity.
⏩ The evolution of what we build is much faster (centuries/decades/months) than biological evolution (thousands of years).

🐟 A submarine doesn’t look like a fish, and an airplane doesn’t flap its wings.
✈️ Both are much better at the core thing we want: travel in the ocean or air.
⚙️ Maybe not as efficient as nature, but definitely faster and more powerful.
🧬 Nature achieved efficiency over thousands of years through necessity.
🔧 What we build outperforms nature on other benchmarks—because it’s built for our needs.

🤖 The AI we want to build—or whatever humans may evolve into—doesn’t need to resemble the humans that came before.
Moravec’s Paradox: Moravec’s paradox is the observation in the fields of artificial intelligence and robotics that high-level reasoning tasks, such as performing calculations, playing chess, or passing intelligence tests, are comparatively easy for computers to achieve at adult human levels, whereas low-level sensorimotor skills—like perceiving the environment, grasping objects, or maintaining balance while walking—that infants acquire with minimal effort require immense computational resources and remain exceptionally difficult for machines. As articulated by roboticist Hans Moravec in his 1988 book Mind Children: The Future of Robot and Human Intelligence, “It is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers and chess, but difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility.”