I think it's 2013 again.

I don't usually get worked up about certifications, but I think we're watching a pattern repeat, and I've learned to pay attention when that happens.

Back in 2013, AWS launched its certification program with a single exam, the Solutions Architect Associate. Most people who noticed saw another line for a resume. I thought I was looking at something bigger, so I went all in, and over the next few years I earned every AWS certification Amazon offered.

People often assume those certifications are what changed my career. They didn't. Cloud did.

The certifications were simply the earliest visible signal that a once-in-a-generation technology transition had begun, and I happened to be standing close enough to notice. That distinction is the whole point of what follows, because I think we're watching it happen again.

In March this year, Anthropic launched its first Claude certification: a single exam. A few days ago it expanded that into a full four-track program. One exam became a portfolio, the same move AWS made, compressed from years into a single quarter.

I've seen this movie before. Not because the products are the same, they aren't, but because the shape of what's happening is.

Here's the framework I've come to trust whenever a new technology wave starts. One credential quietly becomes an ecosystem. Companies discover they can't hire enough people who actually have the new skill. And everyone spends a year arguing about whether the whole thing is hype. When those three line up at once, I stop asking whether the technology is real and start asking how early I still am.

That's almost exactly what happened with AWS. The certification launched on 30 April 2013 with a single exam and grew into today's four-tier program spanning Foundational, Associate, Professional, and Specialty tracks. By January 2025, AWS had built that into 1.42 million active certifications held by 1.05 million people.

But the certifications only became valuable because cloud became indispensable first. The real opportunity showed up around the credential, not inside it: cloud architects, migration specialists, independent consultants, entire companies that existed only to teach people cloud skills. A Cloud Guru launched in 2015, two years into that wave, selling nothing but cloud-certification training, and it was eventually acquired in a deal worth more than $2 billion. That's what a real transition creates. Not badges. Industries.

Now hold that backdrop up against AI. The wage premium for AI skills sits at 56% this year, according to PwC's 2025 AI Jobs Barometer, up from 25% just twelve months earlier. That's not gradual growth, that's a number that doubled while most of us weren't watching closely enough. ManpowerGroup's 2026 survey found AI skills are now the hardest in the world to hire for, ahead of every other engineering and IT discipline, for the first time on record. At the top of the market, specialists building with large language models are already clearing $240,000 to $350,000 in base pay alone.

Those aren't certification statistics. They're transition statistics, and transitions are what I've learned to bet on.

What the AWS wave actually rewarded wasn't trivia. It rewarded people who could take a system from a whiteboard to production: real architecture decisions, real integration into a business that had never touched the cloud, real accountability when something broke at 3am. I'd bet this wave rewards the same instinct, applied to a different stack. The people who benefit won't be the ones who can define an AI term correctly. They'll be the ones who can design a system that holds up past the demo, survives evaluation, integrates into how a business actually runs, and satisfies whoever in the room is asking about governance and risk. That's a different skill than knowing what an answer key says, and it's the one I think actually compounds.

None of that means Claude certifications become the next AWS certifications. History doesn't repeat that cleanly, and there are good reasons to stay skeptical: a certification isn't experience, hiring markets cool, and eventually everyone holds the same credential. Those objections are all fair.

They're also exactly why I'm paying attention now rather than later. A signal is worth the most before it becomes common. By the time everyone agrees a technology wave is real, the largest opportunities have usually already moved on to whoever showed up first.

That's the lesson AWS actually taught me. The certification was never the opportunity. The technology transition was. The certification was just the earliest visible signal that it had started.

So I'm going all in again. Not because I think certifications create careers on their own, they don't, but because I think technology transitions do, and I'd rather be early and wrong than late and right.

I also built the resource I wish I'd had in 2013, when I prepped for the AWS exams the hard way: scattered notes, static courses, re-reading things I already knew. This time I built something that tracks what you've actually mastered against what you've only seen once, and spends your time on the gap. The practice itself is built around the same decisions you'd face shipping one of these systems, architecture trade-offs, evaluation and testing, integrating into a stack that already exists, not definitions to memorize, with an AI tutor that pushes you to reason your way to the answer instead of handing it to you. Alongside it there's an independent comparison of the four Claude certification tracks, linked straight to the official exam guides and Pearson VUE registration, plus a free tool that works out which certification fits your background. Whether you use any of it isn't really the point.

The framework is. Watch for one credential becoming a portfolio. Watch for a skills gap nobody can fill. Watch for a loud argument about whether it's all hype. When those three show up together, you may be watching a technology transition begin, not just a company shipping a product.

That's the pattern I'm betting on. I'd genuinely like to know if you think I'm reading a real signal, or just seeing history where none exists.

The full four-track comparison is here, and the free which-certification tool is here. AI Skill Certs is an independent preparation platform and is not affiliated with, endorsed by, or sponsored by Anthropic; the exam names are used only to identify the exams. Market figures are from public reports including PwC's AI Jobs Barometer, ManpowerGroup's 2026 talent-shortage survey, and industry compensation and certification-premium surveys; the A Cloud Guru figures are from public reporting on the Pluralsight transaction.