Eighteen months ago I was hiring five to ten freelancers a month. Bloggers, designers, a web developer here and there, the occasional translator for a client deck. Today that number is zero. Not because my business shrank, but because I stopped needing most of what I was paying for. That's the first real data point behind a bet I'm willing to put my name on: marketing is going to be the industry AI hits hardest, not the traditional LATAM businesses everyone loves to use as the scary example.
That's the whole thesis, no buildup needed. If you work in marketing production, the kind of work you can post as a fixed-price gig with a three-day turnaround, I'd give it maybe five years before that gig stops making economic sense for the person paying for it. Not because freelancers got lazy or bad at their jobs, but because the reason I used to hire someone else to do it is disappearing. And I watched it disappear inside my own business first.
Where this bet actually comes from
I didn't arrive at this from some panel discussion about AI and the future of work. I arrived at it from my own hiring spreadsheet. I've written about the specifics of who I stopped hiring and why elsewhere, but the short version is that bloggers were the first to go, then designers, then most of the small web development work. Not all at once, one category at a time, as the tools got good enough that the freelancer's output and my own output, using the same tool, stopped being meaningfully different.
Here's the part that still makes me laugh a little. I wasn't hiring those bloggers for research or strategy. I was hiring them for how they wrote. That was the entire value proposition: this person has a voice, a rhythm, a way of framing an idea I don't have time to replicate myself. And somewhere along the way, without telling me, several of them started running the exact same AI tools I had access to. Same prompts, probably. Same model. I was paying a markup for a process I could run myself. I hired them for how they wrote, and they ended up using the same AI I could've used. That's not an insult to them. That's the whole economic story of this moment, compressed into one relationship.
The junior lawyer paradox, applied to marketing

There's an idea that's been floating around law firms for a couple of years now: AI is devastating for junior associates and barely relevant for senior partners. The model can draft a serviceable contract, summarize a deposition, flag the obvious issues. What it can't do is decide which issues actually matter, push back on a client's bad instinct, or own the call when something goes wrong. Junior work gets commoditized. Judgment doesn't.
Marketing has the exact same split, which is why I think it's more exposed than people assume, not less. A huge chunk of marketing output is junior-associate work wearing a creative costume: a blog post, a set of ad variants, a product description, a basic logo concept, a landing page draft. Repeatable, digital, judged mostly on whether it's acceptable rather than whether it's extraordinary. That's exactly the category of jobs AI can't replace yet that actually shrinks fastest, because it was never really about irreplaceable talent. It was about producing an acceptable first draft, reliably. AI does that now. What it still can't do is decide what the business should even be saying, who it's for, and whether the draft is going to convert anyone. That's the senior partner work. And that part gets more valuable, not less, which is a strange thing to say about an industry I'm simultaneously calling the hardest hit.
Why I'm not saying Upwork and Fiverr become irrelevant
I want to be careful here, because the lazy version of this bet is "freelance marketplaces are dying," and the data doesn't support that at all. If anything, it's the opposite. Fiverr has reported search demand for AI-agent freelancers jumping by a wild margin in a single stretch of 2025, with demand for people who can "humanize" AI-generated content climbing hard too. Upwork's own research shows skilled freelancers reporting advanced AI proficiency at notably higher rates than full-time employees, and using AI tools more often day to day, a nice, slightly uncomfortable confirmation of exactly what I watched happen with my own bloggers: freelancers adopted the tools faster than the businesses hiring them did.
So the precise bet isn't "these platforms stop mattering." It's narrower than that. The cheapest, most interchangeable gigs, the ones where the whole pitch is "I'll write you five hundred words for twenty dollars," lose their reason to exist, because the buyer can now get an acceptable version of that same output directly, in minutes, without a back-and-forth over email. The gigs that survive and grow are the ones built around judgment, taste, and operating AI systems well, which is exactly where the marketplace data already points. This is how companies are hiring differently with AI: less volume buying of raw production, more targeted hiring of people who can direct a system and catch what it gets wrong.
If you're a freelancer reading this and wondering whether to panic, panic is the wrong move. Repositioning is. The freelancers thriving in that Fiverr and Upwork data aren't competing with AI on speed of a first draft. They're the ones who learned to operate it, edit it, and sell the judgment layer on top of it. That's reskilling for the AI era in its most literal, unglamorous form: not a certificate, just getting good at the part of the job a model still can't do on its own.
The part I'm not predicting: LATAM distributors
Here's where I deliberately part ways with the scarier, more clickable version of this argument. A lot of people assume AI job displacement in Latin America shows up first in traditional sectors: distributors, hardware resellers, mid-size companies with twenty employees and a warehouse that have never touched a CRM properly. I don't think that's where the hit lands, at least not as a disappearance. I think it lands as a transformation, and that's a meaningfully different story.
When I work with a distributor selling cameras or home automation hardware, the sales role isn't going anywhere, because selling that kind of product still depends on relationships, product knowledge, and local context an AI model simply doesn't have. Nobody is asking ChatGPT which installer in Guadalajara actually shows up on time, or which client needs a payment plan instead of a hard no. That judgment and that trust are the job. What does change, and change fast, is everything around that judgment: lead capture that used to sit in a notebook, follow-up that used to depend on someone remembering to send a WhatsApp, quoting that used to take twenty minutes of manual lookup. I've written before about cutting quote prep down to under a minute after years of automating that one piece of the job, and the sales rep didn't disappear when that happened. Their day just stopped being eaten by busywork a system should've handled from the start.
That's the real shape of AI adoption in small business LATAM right now: not replacement of the commercial role, but augmentation of the parts of it that were always closer to admin than to selling. Regional research on AI maturity backs this up too. The World Economic Forum's work on Latin America points to a region still mostly running pilots in sales and digital marketing, not mature deployment, which tracks with what I see on the ground: the technology is available, the habit of actually using it well isn't built yet. That gap, not job loss, is the real story of AI transforming traditional industries in this region. It's a maturity problem before it's a displacement problem.
Is AI a threat or an opportunity, then?
Honestly, both, depending on which seat you're sitting in. If your entire professional identity is "I produce an acceptable first draft fast," that's a threat, and no amount of being a nice person changes the math. If your value was always the judgment layer, the strategy, the relationship, the ability to catch what a draft gets wrong, this is probably the best moment of your career to be in that seat, because the busywork around you just got a lot cheaper to do.
I built my own content engine from scratch after years of putting up with tools I didn't even like, and the lesson wasn't "AI replaces creative work." It was that one good interview session with a founder who actually knows the business now produces more usable material than a month of back-and-forth with a freelancer ever did. That's not a knock on the freelancer. That's just what happens when the cost of an acceptable draft collapses toward zero.
So here's my actual challenge to you, whichever seat you're in. If you're in marketing production, stop asking whether AI will take your job and start asking which part of your job was never really about talent in the first place, because that's the part going first. And if you run a distributor or a mid-size business in this region wondering whether any of this applies to you: it does, just not in the way the headlines are selling you. The question worth sitting with isn't whether AI replaces your sales team. It's whether you're still making your best people do work a system should've taken off their plate years ago.
What part of your job is actually judgment, and what part is just a well-dressed first draft?
FAQ
Why did the author stop hiring freelancers?
The tools got good enough that the freelancer's output and the author's own output, using the same AI, stopped being meaningfully different. Several freelancers were even using the same AI tools themselves, so the markup no longer made sense.
Does this mean Upwork and Fiverr are dying?
No, the data actually shows the opposite. Both platforms report rising demand for freelancers skilled in AI, it's the cheap, interchangeable gigs that lose their reason to exist, not the marketplaces themselves.
What is the "junior lawyer paradox" and how does it apply to marketing?
It's the idea that AI commoditizes junior-level work but can't replace judgment, like deciding which issues matter or pushing back on a bad instinct. Marketing has the same split, since much of its output is junior-level work in a creative costume that AI can now produce reliably.
Will AI replace sales roles at LATAM distributors?
No, the author argues selling hardware still depends on relationships, product knowledge, and local trust that AI doesn't have. What AI changes is the admin-heavy busywork around the sales role, like lead capture, follow-up, and quoting.
So is AI a threat or an opportunity for marketers?
Both, depending on your role. If your value is producing an acceptable first draft fast, it's a threat, but if your value is judgment, strategy, and relationships, this is one of the best moments to be in that seat.