AI Strengths vs Weaknesses
About this resource
The most expensive AI decisions come from applying these tools to the wrong category of work. They are extraordinarily good at some things and structurally poor at others, and the boundary is not intuitive. This one-page map sets AI strengths and weaknesses side by side so the judgment can be made before a task is assigned rather than after the output disappoints.
Strengths cluster around transformation, synthesis, and first drafts. Weaknesses cluster around anything requiring current information, exact recall of specifics, genuine arithmetic, knowledge of your organization, or accountability for a decision. Knowing which side a task falls on is most of the skill, and it transfers between tools in a way that prompting technique does not.
Almost every disappointing AI result traces back to a task from the right column being handed to a tool that only performs well on the left.
How to use it
Use it as a filter before you commit anything. When somebody proposes an AI application, find the task on this page first. If it sits in the right column, the answer is not necessarily no, but it does mean the human contribution has to be designed in rather than assumed.
Inside the download
Transformation, synthesis, drafting, and structuring, where these tools are genuinely strong.
Current information, exact recall, arithmetic, and anything about your own organization.
Why each weakness exists, so the boundary is memorable rather than arbitrary.
Decide before delegating a task rather than after the output disappoints.
Common questions
Transforming material you supply, summarizing long documents, drafting first versions, restructuring content for a different audience, generating options, explaining concepts at different levels, and finding patterns across text. All of these share a property: the source material is present rather than recalled.
Anything requiring information after its training cutoff, exact recall of specifics such as figures and citations, reliable arithmetic, knowledge of your organization that it was never given, and accountability for a decision. It will produce confident output in all of these categories anyway, which is what makes them dangerous rather than merely limiting.
When being wrong is expensive and you cannot verify the output, when the task requires information the model has no access to, when a qualified professional judgment is legally or ethically required, and when the value of the work is the thinking rather than the artifact.
Free. No email list, no follow-up sequence.
More from the resource library, chosen because they pair well with this one.
Common beliefs about AI set beside what is actually true, including the ones that turn out to be correct.
A task-based comparison of the major assistants, including where the difference matters and where it does not.
Where human judgment still matters, three levels of stakes, and what review actually has to mean at each one.
Tell us a little about yourself and this document is yours. We don't have an email list. We collect this so we know who's finding our resources useful, and that's it.
You're downloading
this document
Your information is kept confidential and never shared with anyone.
