Has AI Really Plateaued?

You've probably heard this recently:

"AI has plateaued."

Could be rumblings from your team, opinions online, or maybe you even feel it yourself.

The honeymoon is over. People are fatigued and disappointed.

AI is work now and it wasn't as easy as promised.

The truth?

What's happening is more nuanced than a plateau.

The masses are starting to feel the difficulty of learning a new skill.

Remember the Dunning-Kruger Effect?

When you first try something new you feel like a genius.

Then reality hits.

You realize it's harder than you thought and your confidence crashes.

Right now, people are in the "Valley of Despair" with AI.

This feeling is amplified by the positioning that AI was supposed to make everything easy.

Think back to 2023 and early 2024.

AI felt magical. Everything seemed possible.

Now it's late 2024 and people are getting a gut check.

AI takes effort. It requires iteration. It's not the magical solution they were sold.

Here's Why the Plateau Narrative is Wrong

The data is telling a different story:

1. The market keeps growing

• Projected to hit $391B in 2025 and reach $1.8T in 2030
• 78% of businesses now use AI, up from 55% in 2023
• Over 5 billion monthly visits to ChatGPT.com alone

2. The technology keeps getting better

• AI benchmark scores jumped by 18-67% in one year
• GPT-5 solved 71.7% of advanced coding problems vs 4.4% last year (a 16x improvement)
• Open-source models have almost caught up to the best paid ones
  

3. People who push through are getting huge returns

• 71% of orgs using AI in marketing/sales see higher revenue
• AI devs are commanding $900/hour rates (demand > supply)
• AI startups are hitting $100M annual revenue within 2 years

The craziest part:

People are saying that instant access to PhD-level knowledge won't help.

...With a straight face. Eek!

The issue is the learning curve, not the technology.

Climbing Out of the Ditch

All this talk of a plateau is really a separation.

Succeeding with AI requires a willingness to push through the tough stages of learning.

You have a choice:

• Hit friction & quit (forever stuck in the valley 😦)
• Lean into difficulty & iterate (climb the slope of enlightenment 🤩)

Climb out of the Valley of Despair first and you'll have a 6-12 month head start before AI competence becomes a baseline expectation.

...In the compounding world of AI, that's a big head start.

Here's how to get out of the valley

• Don't write off AI because your first attempts were messy
• Pick one tool and get good at it (here's my stupid simple reco!)
• Keep iterating, then expand your tools strategically

Others will quit because they've hit a rut, but you have a chance to get out by leaning in.

It won't be easy, but that is the point.

Difficulty creates opportunity.

Now, promise me this:

Leave the plateau talk to the people who choose to stay in the valley.