Erik J. Larson, who has written for us over the years, has a funny article at Nautilus: Is the advanced chatbot (Large Language Model or LLM) that OpenAI is releasing next year – ChatGPT5 – as intelligent as a five-year-old?
The machine, he says, “will supposedly be able to solve complex logical and multi-step problems like a human with a PhD. What it probably won’t do, however, is know much about how the world works.” So says Larson, author of The myth of artificial intelligence (Harvard 2021) compares his abilities to those of the five-year-old in several areas, including these two:
COMMON SENSE AND UNDERSTANDING OF CONTEXT
The LLM:
LLMs can make logical connections based on patterns in training texts, such as understanding that “the ground gets wet when it rains.” However, they may not understand that bringing an umbrella is a practical next step if it is not explicitly mentioned in the text.
They are also able to generate contextually correct responses, provided those contexts are clearly defined in their data. LLMs may struggle with ambiguous phrases such as “Can you pass me the salt?”, but usually know when to say “thank you” after a compliment.
The 5-year-old:
Five-year-olds instinctively know not to stack books on a cake, although they may try it just to see what happens. May say “thank you” after a compliment, but may also stick out their tongue.
They understand the concept of gravity when their ice cream falls, but may still argue that “gravity is to blame” when they trip over their own feet.
May understand that bedtime means pajamas and brushing teeth, but may insist that wearing a superhero cape to bed is a necessary part of the routine, just in case they need to fly in their dreams.
If you ask them about space, you will hear about aliens with ice cream planets.
IMAGINATION AND CREATIVITY
The LLM:
Can cobble together a tale of dragons and knights or a futuristic sci-fi epic by assembling elements from vast story databases.
Her creativity is reflected in her ability to write lyrics that sound imaginative, if not always original.
The 5-year-old:
The minds of 5-year-olds are a limitless playground. They invent invisible friends who speak in rhymes and worlds where every animal wears a hat and glasses.
A child’s artwork might show a rainbow next to a giraffe, because why shouldn’t the giraffe have a nice day?
Erik J. Larson, “Who is smarter: AI or a 5-year-old?”, Nautilus, August 15, 2024
You can probably guess who Larson declares the winner, but test your guess here. He points out that LLMs are “limited by the data they were trained with and lack real understanding, common sense, and emotional depth.”
Larson writes a Substack column, Collegiate, as well as.
You may also like to read: How Erik Larson came up with a method to decide who is influential. The author of The myth of artificial intelligence decided to apply an algorithm to Wikipedia – but it had to be very specific. Many measures of influence are based on broad measures such as the number of hits on pages. Larson realized that influence is more subtle.
