How Intelligent is AI?

The analytical intelligence test of AI with a real-life observation.

Mark K. Setton

I once told my 9-year-old daughter that the US Navy was teaching dolphins to read sign language. “Hey Dad,” she responded, “how come they can understand what wer’e saying but we can’t understand what they’re saying?”

Why do we always use human intelligence as the yardstick to measure artificial intelligence?

It’s the old anthropocentrism, as usual.

Backyard Birds

So let’s take a look at AI’s analytical intelligence. It’s a little more easy to measure than fuzzy concepts such as “IQ” and very fuzzy concepts such as “consciousness.”

This is a simple test of inductive, deductive, and abductive capabilities illustrated by ChatGPT 4 using an original observation (authored by me, “on the spot,” with no content on the web).

MARK

About a year ago I heard loud squawking and chattering and so I looked outside my window. I saw at least three species of birds making various noises, and involved in some sort of aerial acrobatics. Beneath the birds, on the lawn, I could see a baby bird, crawling very slowly.

Can you think of some explanations about what was going on?

CHATGPT

Based on your description of the scene, here are a few hypotheses to explain what was happening:

1. Rescue Mission: The baby bird might have fallen out of its nest, and the birds from various species were gathering around, either to protect the baby bird from predators or attempting to assist it. The aerial gymnastics and loud noises could be a way to distract potential threats or a call for help.

2. Territorial Dispute: The presence of multiple bird species could indicate a territorial dispute. The baby bird on the ground might have unintentionally become a focal point for this dispute, with different species either trying to claim territory or protect the young bird.

3. Feeding Frenzy: The baby bird might have been seen as potential prey by the other birds, especially if the species involved are known to be aggressive or predatory. The commotion and aerial maneuvers could be the birds competing for the opportunity to feed.

4. Learning to Fly: It’s possible that the baby bird was in the process of learning to fly and had a rough landing. The other birds, potentially its family members or from the same species, were making noise and flying around either to encourage it or to guide it back to safety.

MARK

Can you rank the 4 explanations in terms of probability, explaining why?

CHATGPT

Certainly, I can rank the four explanations for the bird scenario in order of probability, based on common bird behaviors and ecological principles:

1. Rescue Mission: This is the most probable explanation. It is well-documented that birds, particularly parents and sometimes even other species, will gather and make noise to protect a young bird in distress. The aerial gymnastics could be an attempt to distract potential predators or attract attention.

2. Learning to Fly: This ranks second in terms of probability. It’s common for fledgling birds to have difficulty with their initial flights. The presence of other birds could be family members or others from the same species, offering support or trying to encourage the young bird.

3. Territorial Dispute: While possible, this is less likely than the first two. Birds can be territorial, but the scenario of different species engaging in a dispute over a baby bird is less common. Territorial disputes usually involve adults and are more about space than a single distressed juvenile.

Feeding Frenzy: This is the least probable. While some bird species might prey on young birds, the scenario described – with multiple species involved and a focus on one baby bird – is less typical in bird behavior. Predatory attacks are usually more swift and less communal.

MY TENTATIVE CONCLUSIONS

1. Chat GPT has used inductive reasoning:

The 4 hypotheses were the product of inductive reasoning, inferring possible scenarios from the specific observations I had provided.

2. ChatGPT has used deductive reasoning:

Deductive reasoning was used more subtly. It involves applying a general rule or principle to a specific case to draw a conclusion. For example, if we apply the general principle that “birds often protect their young from danger,” we can deduce that one possible reason for the behavior of the birds in my observation is a protective response to the baby bird in distress. This general rule wasn’t explicitly stated in my observation, but it’s a common understanding about bird behavior that was applied deductively to hypothesize about the situation.

3. ChatGPT has used abductive reasoning

Abductive reasoning involves observing a situation, generating multiple hypotheses to explain it, and then determining the most likely explanation based on known information and logical inference. This approach is often used in practical problem-solving, detective work, and scientific research, where the goal is to find the most probable explanation for a set of observations.

AI and Human Happiness

Similar Posts