Lecture 1: What AI Can and Cannot Do
Before you touch a single tool, you need an honest picture of what these tools are. Not the hype, and not the fear. Just a working mental model you can rely on. This lecture gives you that model in plain language.
What AI actually is, in one sentence
The AI tools in this program are systems that have read an enormous amount of human writing and learned the patterns in it, so they can produce new writing that fits the patterns. That is the core of it. When you ask ChatGPT or Claude a question, it is not looking up a fact in a database. It is predicting, word by word, what a useful response would look like based on everything it has seen.
That one idea explains both why these tools are so capable and why they fail in the specific ways they do. Hold onto it.
What AI is genuinely good at
These tools are strongest wherever the work is made of language. That covers far more of professional life than people expect. They are good at:
- Drafting. A first version of an email, a cover letter, a summary, a job description, a plan. Starting from a blank page is hard; starting from a draft is easy.
- Rewriting. Making something shorter, clearer, more formal, more friendly, or aimed at a different reader.
- Explaining. Taking something dense and breaking it into plain steps, or answering "what does this mean" in language you choose.
- Organizing. Turning a messy pile of notes into a clean list, a table, or an outline.
- Brainstorming. Producing many options quickly so you have something to react to and choose from.
Notice what these have in common. In every case, the tool does the heavy lifting of producing language, and you do the lighter but essential work of judging whether it is any good. That division of labor is the heart of using AI well.
Where AI falls short
The same prediction engine that makes these tools useful also makes them unreliable in predictable ways. You need to know these limits cold.
It can be confidently wrong
Because the tool produces text that fits a pattern, it can produce text that sounds completely correct and is not. It can invent a fact, a statistic, a quote, or a source that does not exist. This is the single most important limit. The tool has no built-in sense of "I am not sure." It will hand you a wrong answer in the same calm, professional voice it uses for a right one. Your job is to verify anything that matters before you rely on it.
It does not know today
Each tool learned from information up to a certain point in time. It does not automatically know this morning's news, today's prices, or a change made last week, unless it has been connected to live search. Treat anything time-sensitive with suspicion and check it.
It can carry bias
The tool learned from human writing, and human writing carries human bias. The output can quietly reflect stereotypes or leave people out. This matters especially in work that affects real people, like hiring language or descriptions of groups. Read with that awareness.
It does not understand you the way a person does
The tool has no memory of your life, your intent, or your stakes unless you tell it. The more context you give it, the better it does. The less you give it, the more it guesses.
The posture that ties it together
Put the strengths and the limits side by side and a simple working rule falls out: the tool drafts, and you decide. It is an assistant, not an authority. You stay accountable for what goes out under your name. That is not a limitation to apologize for. It is the professional standard, and it is what separates someone who uses AI well from someone who gets burned by it.
This is also why the program's spine is clarity over sophistication. You will get far more out of these tools by being clear about what you want than by knowing clever technical tricks. Preparation beats mastery. Familiarity beats intensity. You do not need to know how the engine works to drive the car well.
What this means for your career
One last point, and it is the one that matters most. AI is not a separate career you are switching into. It is a capability that extends the career and the judgment you already have. A scheduler who uses AI is a sharper scheduler. A job seeker who uses AI writes better applications. The skill you are building this term sits on top of who you already are. Keep that frame, and the rest of this program is about getting fluent with a tool that works for you.
In the next lecture, you will create your accounts and put all of this into practice.
Learn. Apply. Advance.