Something Big Is Happening. Most People Don't Know It Yet

We spend a lot of time in the communities — on X, YouTube, Discord servers, offline events, and forums — where developers, founders, and operators are pushing AI further every week. And the gap between what those people are doing and what the average professional thinks "using AI" means is bigger than most people realize.

Here's what we keep seeing: someone says they use AI at work. What they mean is they open ChatGPT a few times a week to search for something, rewrite an email, or summarize a document. That's not wrong. But it's the first step on a long staircase, and most people don't know the staircase exists.

The COVID Comparison

In February 2020, a virus was spreading overseas. The stock market was doing great. Your kids were in school. You were going to restaurants and shaking hands. If someone told you they were stockpiling toilet paper, you would have thought they'd been spending too much time on a weird corner of the internet.

Then, over three weeks, the entire world changed.

Matt Shumer, with six years building in the AI space, published a thread in February 2026 that was seen 86 million times in 48 hours. His opening argument: we're now in the "this seems overblown" phase of something much bigger than COVID. 

Shumer's post: "I think we're in the 'this seems overblown' phase of something much, much bigger than Covid." Matt Shumer (@mattshumer_), February 10, 2026. 86.2M views on X.

He's not alone. The conversation happening in the communities where AI is actually being built sounds nothing like what most professionals hear in their industries.

The people raising the alarm aren't predicting. They're reporting what already happened to their own jobs.

Already Moving Faster Than the News Covers

There's an organization called METR that actually measures this with data. They track how long an AI can work independently on real tasks without human help. About a year ago, the answer was roughly 10 minutes. Then an hour. Then several hours. Their most recent measurement showed AI completing tasks that take a human expert nearly five hours, and that number is doubling approximately every seven months.

Those measurements don't yet include the latest models released in early 2026.

The free-tier problem compounds this. Most people are using the free version of AI tools, which is over a year behind what paying users have access to. Judging AI by the free-tier tools is like evaluating the state of smartphones by using a flip phone. The people paying for the best tools and using them daily for real work are experiencing a completely different technology.

The Staircase Nobody Told You About

Most people think basic ChatGPT use is "using AI." It's level one of four. According to Notion, here's what the staircase actually looks like:

Level 1: AI as a Thought Partner. You use AI to explore ideas and improve decisions. Brainstorming, research, faster first drafts, quick answers. Low barrier, no technical skills required. Almost everyone we talk to is here, and most of them think this is the whole picture.

Level 2: AI as an Assistant. AI is embedded in your daily tools. Notion AI for writing and summarizing, Copilot in your editor, AI meeting notes with action items, enterprise search across company knowledge, auto-fill across your databases. Your company has started providing access and guidelines. Individual productivity gains are measurable in hours per week, but they're not yet compounding.

Level 3: AI as Teammates. Custom agents handle triage, routing, and status updates inside your team. Workflows span multiple tools via MCPs and connectors. Teams build their own agents for their specific needs. AI governance is being formalized. Organizational capacity scales beyond headcount, because the org has hired software that works while people sleep.

Level 4: AI as the System. Multiple agents coordinate across systems autonomously. Critical business processes run end-to-end with AI making decisions inside defined guardrails. Performance is measured and optimized continuously. The company has a structural competitive advantage from AI-native operations.

Most professionals are at level one and think that's the ceiling. The companies pulling ahead are already operating at level two or three. A handful are building toward four right now.

The same tools being used to answer basic questions can also write and deploy code, automate entire workflows, analyze months of business data, generate production-ready designs, build internal tools, and run tasks autonomously while you do something else entirely.

We Also Built a Free, Global, Open Self-Assessment

We made it because we believe the first step to changing your situation is understanding it honestly. You'll answer questions about how you actually use AI in your specific role, not how you plan to, not how you think you should, but how you do right now. The questions get specific. Some of them will surprise you. Some of them will describe things you didn't know were possible.

At the end, you'll see your level across five stages, from Explorer to Catalyst. You'll get an honest description of where you are and one concrete next step.

If you want to see how your results compare to other professionals by country, industry, company type, and salary range, you can share your email and a few anonymous demographic details. Completely optional. Your individual answers are never shared with anyone. The benchmark is built from aggregate, anonymous data, so the more people participate, the more useful the comparison becomes. If you'd rather just see your own result and leave, that's fine too.

We're not in a "this will be important someday" moment. The people who are ahead in their industries right now are the ones who started experimenting seriously six months ago. The bar is still low enough that six months of honest effort puts you ahead of 90% of your field.

Take the assessment. Find out where you actually are.