{"id":34006,"date":"2026-07-01T17:00:00","date_gmt":"2026-07-01T22:00:00","guid":{"rendered":"https:\/\/www.ecoticias.com\/en\/?p=34006"},"modified":"2026-07-01T06:40:39","modified_gmt":"2026-07-01T11:40:39","slug":"they-hid-the-number-zero-from-an-artificial-intelligence-system-in-the-hope-that-it-would-discover-it-on-its-own-what-happened-teaches-us-an-important-lesson-about-the-future-of-ai","status":"publish","type":"post","link":"https:\/\/www.ecoticias.com\/en\/they-hid-the-number-zero-from-an-artificial-intelligence-system-in-the-hope-that-it-would-discover-it-on-its-own-what-happened-teaches-us-an-important-lesson-about-the-future-of-ai\/34006\/","title":{"rendered":"They hid the number zero from an artificial intelligence system in the hope that it would discover it on its own; what happened teaches us an important lesson about the future of AI"},"content":{"rendered":"\n

Can a machine learn mathematics so deeply that it invents a missing idea on its own? A new Princeton University study puts that question into one of the simplest places imaginable, basic arithmetic<\/a> without the number zero.<\/p>\n\n\n\n

The researchers trained language models<\/a> based on the GPT-2 architecture on single-digit arithmetic problems that did not include zero, then tested whether those models could handle zero when it appeared later. The answer was a reality check for AI research<\/a>. The models did not independently make the leap, although they learned much faster once they were shown a small number of examples.<\/p>\n\n\n\n

Why zero matters<\/h2>\n\n\n\n

Zero may look simple. It is the number on a keyboard that means nothing, the balance in an empty account, the temperature mark we notice on a freezing morning, or the answer when nothing is left.<\/p>\n\n\n\n

In mathematics, though, zero is not just a blank space. It works as a number, a placeholder, and a concept that changes how arithmetic behaves. That is why it makes a surprisingly sharp test for artificial intelligence.<\/p>\n\n\n\n

The study, titled \u201cNothing from Something: Can a Language Model Discover 0?\u201d asked whether a model could move beyond what it had been shown. In other words, could it infer a new mathematical structure from nearby examples, or would it need direct help?<\/p>\n\n\n\n

A test built to be small<\/h2>\n\n\n\n

The setup was intentionally modest. The team used simple arithmetic rather than advanced algebra or proof writing, because a smaller problem makes the failure or success easier to measure.<\/p>\n\n\n\n

\n
\n
\n

Read More: The enormous mass of warm water in the Pacific that has just \u201cwoken up\u201d and could influence storms, wildfires, monsoons, and temperatures through 2027<\/a><\/h3>\n<\/div>\n<\/div><\/div>\n<\/div>\n\n\n\n

The models were based on GPT-2-size systems, with 12 layers, 12 attention heads, and about 124 million parameters. Some models had language pretraining, while others did not, giving the researchers a way to test whether exposure to ordinary text helped with mathematical generalization.<\/p>\n\n\n\n

That matters because many AI systems look impressive when they solve problems similar to their training data. The harder question is whether they can step outside that familiar zone. That is where the zero test comes in.<\/p>\n\n\n\n

\n