Artificial intelligence (AI) may sound like machines can think for themselves, but today’s AI does not think, understand, or reason like humans. Instead, AI generates likely outputs based on previous data, but it does not distinguish between truth and falsehood, possess intent or wisdom, or hold lived experience and cultural understanding like humans, all of which are necessary for critical thinking and decision-making.
What is AI? What Can it Do?
What AI Actually Is
At its core, AI is a set of computer systems designed to perform tasks that normally require human-like decision-making, such as recognizing patterns, understanding language, identifying images, or making predictions. The term "artificial intelligence" might make it sound as if machines have minds or independent thought, but that is misleading.
It simply follows mathematical rules and statistical patterns to produce outputs from inputs. This is because training an AI model involves feeding it many examples and letting it adjust so it can perform well on similar new data. Once training is complete, the model uses what it has “learned” to generate outputs, predictions, or responses when given new inputs.
For example:
- A language model generates text by predicting which words are most likely to follow the ones you gave it.
- An image model identifies objects by comparing pixel patterns to examples it has seen before.
AI systems are technical tools created by humans, made up of:
- Data: the examples used to train the model to recognize patterns and make decisions based on statistical likelihoods.
- Algorithms: the sets of rules or instructions that tell the computer how to process the data.
- Models: configurations of an algorithm trained to make guesses about real-world data
The Limits of AI
An AI model becomes truly useful only after being trained on large amounts of data. Since AI learns from the datasets it is trained on, its performance depends entirely on the quality, scope, and fairness of that data. If the training data is incomplete, biased, or unrepresentative of a certain population’s lived experience, the AI system will reflect those limitations. For example, a model trained primarily on one group’s language or images may perform poorly on other groups' data or produce biased predictions that amplify existing social inequalities.
This dependency on data also means AI struggles with situations it has not seen before. It lacks common sense reasoning and cannot “understand” context the way humans do. For instance, while an AI system may correctly identify a stop sign in a photo, it doesn’t grasp the real-world meaning of the sign or the social norms around driving behaviour.
Understanding AI this way, as a technology shaped by data, algorithms, and human design, allows us to see both its power and its limits more clearly. It also reminds us that AI does not replace human judgment, values, or meaning-making; it only supports tasks where pattern recognition and prediction are useful.
What AI does:
- Detect patterns and calculate probabilities
- Generate likely outputs based on past data (e.g., answers, predictions, best guesses)
What AI does not:
- Distinguish truth from falsehood
- Have lived experience or cultural understanding
- Possess intent or wisdom
- Guarantee accuracy or verified facts
“Artificial intelligence” vs living wisdom
The phrase “artificial intelligence” can create the impression that machines are thinking or reasoning like people. In reality, current AI systems, including the popular large language models used in chatbots, are sophisticated pattern-matching tools, not conscious or understanding entities. They calculate probabilities and correlations rather than engage in independent reasoning or insight.
Experts point out that AI’s outputs can look intelligent without the system possessing any real comprehension. This leads to what scholars call the AI trust paradox: people tend to trust AI outputs because they appear plausible or fluent, even when the system doesn’t truly “know” what it is saying.
This differs greatly from the living wisdom that humans carry, particularly Indigenous Peoples, in the context of Indigenous data. We see multiple social media movements that demonstrate this through slogans such as “Don’t ask AI, ask an Elder” or “Don’t ask AI, ask an Auntie.”