Artificial Intelligence, or AI for short, has arrived. Whether you’ve been dreading the inception or had embraced it from the start, AI is here to stay. However, AI can make mistakes, and if you don’t know anything about a topic, you could be getting some very incorrect information.

AI can produce incorrect results for a variety of reasons. It may not be intentional (so it says), but it IS a possibility AND a byproduct of how it works.

1. Garbage In, Garbage Out
If the data/information AI was trained with comprises errors, outdated information, or biases, those errors can show up in its answers. For example, if outdated building codes are in the training data, AI might present them as current. Then, you’re working with incorrect information, which directly affects the outcome of a project.

2. Outdated Knowledge
Most AI models have a “knowledge cutoff” date. If something changes afterward, such as safety regulations or material costs, AI won’t know unless it is updated with real-time data. AI is only as smart as the information it gleans. 

3. Lack of Context
AI’s accuracy depends on how the question is asked. Questions with missing details or vague wording can lead to incorrect or overly generic answers. For example, asking “What’s the best roofing material” without specifying the climate could give the wrong recommendation.

4. Pattern Matching Does NOT Always Equal Understanding
AI predicts words based on patterns, not true reasoning. It can sound confident but still be incorrect- just like that friend or coworker that spews ‘knowledge’ without getting all the facts. It’s always best to fact check everything-and from multiple sources, if possible. However, if you’ve been in particular fields for more than a decade, that should be common knowledge (although always fact-checking information is not common knowledge for everyone). Unfortunately, many people today simply take a post online as factual without checking for proof (or if the information makes sense).

5. Limited Real-World Judgment
AI can’t walk a jobsite or spot practical issues. Humans are better at noticing “this won’t work” problems.

6. Misinterpreting a Question
AI may focus on familiar patterns and overlook subtle differences in wording, especially in tricky or complex questions. The result can be AI producing incorrect information.

7. Rapidly Changing Fields
Industries evolve quickly, and AI might still reflect yesterday’s best practices instead of today’s. Therefore, it’s always a best practice to fact-check everything you find online.

Ultimately, AI is best used simply as a TOOL to help come up with ideas, basic research and support, and not as the sole way to make a decision. You still need a human brain to work out the details and nuances of a project. So, before making a significant decision based on something AI produced, make sure to do YOUR due diligence. If a mistake is made because of information gleaned from AI, it’s still YOUR mistake.