How I Use AI Every Day (and What I've Learnt)

When I say I use AI in my day-to-day, I don't mean "I asked ChatGPT something once". I have specific workflows where AI is part of the process, from preparing a talk to reviewing a collaboration proposal.

This post has been up for over a year and I keep updating it, because how I use these tools changes almost as fast as the tools themselves.

It's Not Just About Tools, It's Generational

Something I notice constantly giving talks on AI and Generation Z: colleagues from earlier generations tend to ask me which tool to use. I almost never think about that first. I think about the workflow, and the tool is the last thing I decide.

I grew up using the internet for everything, so for me AI isn't a new technology to learn, it's an extension of the same habit: search, iterate, discard what doesn't work, keep what does. That changes how you integrate it, not just how often you use it.

Preparing Talks

Preparing a keynote requires research, structure, narrative and design. AI helps me in the first two phases.

How I do it:

1. I define the topic and audience (that's always me)
2. I ask Claude or ChatGPT to generate 10 different angles for approaching the topic
3. I pick 2-3 that resonate with what I want to say
4. I ask it to help me build a 45-minute narrative structure
5. I fill it in with my own content, real examples from talks I've given (Paraguay, Spain, Morocco, Uruguay), and my point of view

The result: I prepare richer talks in less time, with more variety of perspectives than I'd have on my own.

Managing Emails and Proposals

I receive proposals to speak at events: some highly detailed, others less so. I used to respond to each one from scratch.

Now I have a base prompt that gives the AI context about who I am, what I offer, and my criteria for accepting events. I paste the received email and it generates a draft response. I read it, adjust it, and send it.

Time saved: between 15 and 30 minutes per email, which over a week with several incoming proposals adds up to real hours.

Research Before Taking On a Project

Before accepting a talk or a consulting project at TOTEM Branding, I research the company: what they do, what context they're in, what real problems they might have. I use AI to speed up that first layer of research, not to replace it, and I always set aside time to verify the most relevant facts myself before citing them in a proposal or a talk.

Content Creation

For the blog (like this post), I use AI differently from what most people would expect: it doesn't write the posts for me. It helps me get unblocked.

When I have an idea but don't know where to start, I explain the topic and ask for 5 possible structures. I choose the one that fits best, and from there I write myself.

It's the difference between a blank page and a conversation that's already started.

What I DON'T Delegate to AI

There are things I deliberately don't outsource:

• My point of view: what I think about a topic is mine
• The stories: nobody else lived what I lived in those talks
• The relationship with clients: that's always human contact
• The final judgement: I review and approve everything

What I've Learnt in the Process

1. The better the prompt, the better the output. It's a skill you develop, not an innate talent.
2. AI is a collaborator, not a substitute. When you treat it as a tool, it works. When you expect it to think for you, it disappoints.
3. The speed of learning has skyrocketed. I can research a new topic with reasonable depth in hours, not weeks.
4. The generation you grew up in matters less than the attitude you bring. I know people my age who use AI worse than colleagues twice their age, because the habit isn't about age, it's about curiosity.

---

If you want to explore how you can integrate AI into your work or team, write to me. It's exactly the kind of conversation I enjoy most.