π― AI in the clinic, office, and studio
A artificial intelligence (a computer program capable of understanding natural language and generating text, images, or analyses) is already being used by independent professionals like you. This isn't science fiction. Lawyers use it to review contracts. Doctors use it to triage symptoms. Accountants use it to classify transactions. Architects use it to create initial project briefs.
π Real cases by profession
Review contract clauses in minutes instead of hours. Legal case research with automatic summaries. Draft initial pleadings.
Symptom screening to guide the patient history. Summaries of recent scientific articles. Patient communication in accessible language.
Automatic transaction classification. Generating narrative reports from numbers. Standardized responses to clients.
Generate project briefs. Research technical standards. Describe specifications in technical language and in plain language.
π‘ Practical Tip
Don't try to use AI for everything at once. Choose a repetitive task that you do every week and try AI on it first. Examples: replying to routine emails, drafting proposals, summarizing meetings.
π What AI does well
AI excels at tasks that involve language and patterns. It can read, summarize, rewrite, translate, and generate text with a speed and consistency no human can achieve alone. But understanding where it shines helps you avoid asking it to do the impossible.
β What AI does well
- βSummarize 50-page documents into 1 page
- βGenerate drafts of emails, proposals, and opinions
- βTranslate technical texts while preserving the tone
- βOrganize unstructured information into tables
- βMaintain consistency in long documents
β What AI does poorly
- βEnsure the information is accurate
- βMaking ethical or judgment-based decisions
- βUnderstand a clientβs emotional nuances
- βAccess real-time information (without an extra tool)
- βReplace the professional-client relationship of trust
Think of AI as a a very fast and very obedient intern, but it lacks common sense. It does exactly what you ask, at the speed you need. But you need to review its work because it doesn't know when it's wrong.
β οΈ What AI can't replace
Yours years of experience cannot be replicated by an algorithm. AI has no lived experience, doesn't understand emotional context, and doesn't take responsibility. It's a powerful tool, but the professional operating it is what makes the difference.
Professional judgment
Irreplaceable
AI can list options, but deciding which path to take requires judgment honed over years of practice. A lawyer knows when a settlement is better than a lawsuit. A doctor notices what the test doesnβt show.
Ethics and responsibility
Exclusively human
AI doesnβt have a code of ethics. Youβre responsible for the work. If AI suggests something unethical, you need to recognize it and reject it. Professional responsibility canβt be delegated to a machine.
Empathy and human connection
The competitive edge
A client seeks you out because they trust you, not an algorithm. AI can help you prepare for the conversation, but human connection is what builds loyalty and generates referrals.
β οΈ Attention
Never use AI as the sole source for decisions that affect clients. Always review. AI can hallucinate (invent convincing information), and you are solely responsible for the final result.
π Data that backs it up
It's not just opinion. Recent research shows that professionals who have adopted AI are seeing measurable gains in productivity and quality. The numbers are consistent across different fields and countries.
π Research and figures
- 37% reduction in the time spent writing professional documents (McKinsey, 2024)
- 25% more tasks completed per day in offices that adopted AI (Harvard Business School)
- 40% improvement in the quality of first drafts (MIT Sloan, study with consultants)
- 2-3 hours saved per week on average by independent professionals (Salesforce 2025 survey)
β Professional WITH AI
- βProposal draft in 15 minutes
- βReview a 40-page contract in 20 minutes
- β5 email versions to choose the best one
- βReference research in 5 minutes
β Professional WITHOUT AI
- βProposal draft in 2 hours
- βReview a 40-page contract in 3 hours
- β1 email version, rewritten manually if you donβt like it
- βReference research in 45 minutes
π‘ Perspective
If you charge R$300/hour and save 3 hours per week, that's R$3.600/month in additional productive capacity. This pays for any AI subscription and still leaves you with savings. The investment pays for itself in the first week.
π‘ Your first test
Enough theory. Letβs run a practical test now. Open the ChatGPT (chat.openai.com) or Claude (claude.ai) in your browser. You don't need a paid account; the free version works for this test.
Copy and paste this prompt
You are an assistant for [YOUR PROFESSION].
I need to reply to an email from a client who is unhappy with the delivery timeline for an opinion. The client is polite but firm.
Write a professional response that:
- Acknowledge the clientβs concern
- Briefly explain the reason for the delay
- Set a new, realistic deadline
- Keep the tone friendly and professional
Maximum 150 words.
Replace [YOUR PROFESSION] with your field. Run it and see the result. Then try asking for adjustments: "more formal," "shorter," "include an apology."
π‘ What to Look for in the Result
- β’AI produced something usable as a starting point?
- β’Is the tone appropriate for your profession?
- β’How long would it take you to write this from scratch?
- β’What would you change? (This is the next step: iterate)
πΊοΈ Opportunity map
To use AI strategically, you need to map where it creates the most value in your routine. Not every task is worth automating. The key is to find the ones that are repetitive, high-volume tasks that take up a disproportionate amount of time.
Standard emails, meeting summaries, business proposals, client responses. Tasks you do every week that already have a set format.
Market research, competitor analysis, social media content, preparation for important meetings. They require more preparation but deliver significant returns.
Final legal documents, diagnoses, definitive opinions. AI helps with the draft, but you are ultimately responsible. Always review.
Difficult conversations with clients, ethical decisions, sensitive negotiations, personal feedback. These tasks require human presence and judgment.
π Exercise: Map your tasks
List 5 tasks you do every week. For each one, ask: "Does this involve reading, writing, summarizing, or organizing text?" If so, AI can probably help. If it involves decision-making, relationships, or context only you have, AI plays a supporting role.
My 5 weekly tasks:
1. _________________ [Does AI help? yes/partially/no]
2. _________________ [Does AI help? yes/partially/no]
3. _________________ [Does AI help? yes/partially/no]
4. _________________ [Does AI help? yes/partially/no]
5. _________________ [Does AI help? yes/partially/no]
β Module Summary
Next Module:
1.2 - How AI thinks - the bare minimum you need to know