I think it's 2013 again.

I don't usually get excited about certifications.

But I think we are watching history repeat itself.

Back in 2013, AWS launched its certification program with a single exam. I decided to go all in. Over the next few years, I earned every AWS certification there was.

People assume those certifications changed my career.

They didn't.

Cloud did.

The certifications were just the earliest visible sign that a huge shift had started.

Here is the thing I wish someone had told me then.

The certification was never the opportunity. The technology transition was. The certification was simply the earliest visible signal that the transition had begun.

Today feels strangely familiar.

In March, Anthropic launched its first Claude certification. One exam.

Today, they expanded it into a full program. Four tracks. The same move AWS made, compressed from years into a single quarter.

One exam becomes a portfolio. A new skills gap appears. Companies suddenly cannot hire fast enough. Everyone argues about whether it is hype.

I have seen this movie before.

The last time, the wave created far more than exam badges.

It created careers. Consultancies. Independent architects. Whole businesses.

The credential paid, too. Certified engineers earned roughly ten to twenty-five thousand dollars a year more than their uncertified peers, about a 26% premium, and more than a million people went on to earn one.

But the badge was never the real prize. A Cloud Guru started two years into that wave, selling nothing but cloud-certification training. In 2021 it exited in a deal worth more than two billion dollars.

Two billion dollars. For teaching people to pass exams.

And in almost every case, the value went to whoever moved first. The engineers who certified while the credential was still rare. The company that launched in year two, not year six.

The numbers suggest this wave may be even bigger.

AI skills are now the hardest in the world to hire for. Harder than any other engineering or IT skill, for the first time on record.

The wage premium for AI skills more than doubled in a single year, from 25% to 56%. The cloud premium at its peak sat around 26%. This one already runs higher than cloud ever did, and it is still climbing.

At the top of the market, the specialists building with large language models clear $240,000 to $350,000 in base pay alone.

The last transition minted the cloud architect. This one is minting a role we do not have a settled name for yet.

That is usually a sign you are early.

Does any of this guarantee AI certifications will have the same impact?

Of course not.

There are good reasons to be skeptical. A certification is not experience. The hiring frenzy will cool. And eventually everyone will have one.

But that is exactly the point.

A signal is worth the most before it becomes common.

That is the biggest lesson I took from the AWS era. The people who waited until everyone agreed the wave was real are the ones who arrived late, into a crowded field, holding a credential that no longer set them apart.

By the time a transition is obvious, the largest opportunities have already moved.

So I am going all in again.

Not because certifications matter. Because technology transitions do, and this is the earliest honest way I know to place a bet on one.

When I prepped for the AWS exams, I did it the hard way. Scattered notes. Static courses. Re-reading things I already knew.

So this time I built the resource I wish I had had. Something that works out what you actually do not know yet, and spends your time only there.

I also wrote up the full comparison of the four tracks, the data behind the parallel, and where I think the analogy breaks down. And there is a free tool that tells you which of the certifications fits your background, if you want a place to start.

If you take one thing from this, I would rather it be the framework than the pitch.

Watch for the pattern. One credential becomes a portfolio. A skills gap no one can fill. A loud argument about whether it is real.

That combination has marked the start of more than one technology transition. And the signal always shows up before the consensus does.

Maybe I am seeing a real pattern. Maybe it is just history where none exists.

I think it is real. I have been here before.

But I would genuinely like to know if you disagree.

---

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.