Even Babies Are Still Way Better at Learning Than AI Models
"We still have to burn down a forest and scrape the entire sum of all human knowledge to re-create this milestone that happens in our living rooms over the course of a year." The post Even Babies Are Still Way Better at Learning Than AI Models appeared first on Futurism .
Despite the impressive strides made by AI chatbots in producing sentences that appear human-like, they still trail far behind in terms of reasoning and productivity. While experts marvel at the rapid progress made in the industry, they also acknowledge that a significant gap remains between AI and human learning capabilities. In a recent discussion on large language model efficiency, MIT Technology Review spoke with several professionals who revealed a stark contrast in the way infants and AI models learn language.
Cognitive scientist Michael C. Frank from Stanford University pointed out that while AI has made "amazing" advancements, the industry still needs to "burn down a forest and scrape the entire sum of all human knowledge" to match the language learning achievements of a child within a year. Infants possess an incredible ability to adapt to language, rapidly absorbing linguistic rules and vocabulary.
In contrast, AI models demand astronomical amounts of language data to achieve even rudimentary proficiency, far surpassing what a typical human would need. This inefficiency has become a major challenge for the AI industry, particularly for companies striving to create truly human-level AI through scaling language models. Experts are skeptical that this goal will ever be attainable, as the disparities between the learning processes of humans and AI continue to widen.
Written by urgent.news from Futurism's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
