OpenAI solves eighty-year conjecture and seeks to justify its valuation

An OpenAI artificial intelligence has managed to refute the Erdős conjecture on distances between points, a mathematical problem that had remained unsolved for eight decades. The finding, reviewed by experts, consists of a set of points in a high-dimensional space. However, OpenAI's true goal is not science, but to justify its trillion-dollar valuation and attract investments, presenting a result with no practical application for the average citizen.

abstract mathematical space with scattered glowing points forming a counterexample to the Erdős distance conjecture, a holographic projection of high-dimensional point cloud rotating slowly, neural network visualization overlays connecting points with thin luminous lines, a sleek black server rack in background with cooling vents and status LEDs, dramatic low-angle cinematic shot, photorealistic technical illustration, deep blue and electric cyan color palette, volumetric lighting through dust particles, subtle chromatic aberration, ultra-detailed semiconductor textures, futuristic research laboratory atmosphere

The mathematical trick behind the smoke of artificial intelligence 🧠

The example that refutes the Erdős conjecture is a set of points in a high-dimensional space, an abstract concept with no real-life utility. The AI required intensive human supervision to interpret the result, and the expert review was paid for by OpenAI and published in a restricted-access journal. These types of problems are toys to demonstrate computational power, while the same AI continues to hallucinate answers in everyday applications like chatbots or assistants.

The AI solves an 80-year-old problem and can't add two plus two 🤖

While OpenAI sells mathematical smoke, its artificial intelligence continues to hallucinate answers in basic tasks, such as calculating change for a purchase or explaining a joke. The average citizen does not understand the Erdős problem, but the headline makes them feel amazed and more inclined to trust AI for important decisions. So now you know: AI can refute 80-year-old conjectures, but don't trust it to remember your shopping list.