ISSUE 001
AI Training Issue 001
Almost every conversation I have about AI training starts in the same place: someone tells me they're behind. Nobody can tell me behind what. So this one is about the measuring stick — where the rungs actually are, what the raw hiring data says instead of the headline, and one finding that suggests the gap isn't about age or budget at all.
An idea worth thinking about
There are four rungs, and almost nobody needs the top one
The Southern Regional Education Board put out a workforce report in July 2025 that does something I hadn't seen anyone do plainly: it breaks AI skill into four levels instead of treating it as a thing you either have or don't. Awareness is knowing what the tools are. Exploration is using them in a guided setting. Application is solving your own problems with them and judging whether the output is any good. Fluency is adapting to new tools as they land and leading other people through the same shift.
What I find useful is how it reframes the anxiety. "Behind on AI" is meaningless. "At Awareness, want to get to Application by spring" is a plan. The report also splits the underlying skills three ways — the human ones like communication and problem-solving, the judgment ones like knowing where AI shouldn't be trusted with a decision, and the technical ones. Two of those three have nothing to do with software.
Read the SREB report
A tool I disappeared into
The raw file behind every "AI jobs are exploding" headline
Indeed's research arm publishes its AI job-postings tracker as a plain CSV on GitHub, updated monthly, free to use as long as you credit them. It's the share of all postings containing AI-related keywords, on a seven-day trailing average. That series is where the number you keep seeing comes from: it hit 4.2 percent at the end of 2025, which sounds enormous until you notice it had already peaked at 3.3 percent back in early 2022, then fell to 1.7 percent by mid-2023 before climbing again.
Two honest limits. It's a keyword match, so a posting that says "Chat GPT" in the perks section counts the same as one requiring real skill. And the file is national — there's no state or county cut, so you can't pull Missouri out of it. What you can do is check a claim yourself instead of taking a vendor's word for it, which is most of the value.
Open the tracker data
Learn something
Google's AI certificate, at no cost, if your business has an EIN
Grow with Google is issuing scholarships for its AI professional certificate to U.S. small and medium businesses. You need 500 employees or fewer and a valid EIN, and you apply through a dedicated page on Coursera. It comes bundled with three months of Google AI Pro and Workspace Business Standard, which puts Gemini inside Gmail, Docs and Sheets while it lasts. Google also runs free in-person AI workshops with local chambers of commerce, so it's worth asking yours whether a stop is planned.
Two details that catch people. A credit card is required even though the training costs nothing, because of how the bundled subscription enrolls. And there's a hard clock: enrollment closes December 15, 2026, with redemption by December 31.
Check eligibility
What we’re seeing
Businesses founded last year adopted AI in six months. The 2019 cohort took six years.
The JPMorganChase Institute went at small-business AI adoption from an angle nobody else has: instead of asking owners what they use, they tracked actual payments to AI services flowing through Chase business banking accounts from 2019 through 2025. Nobody has to self-report, and nobody has to guess what "using AI" means.
The finding that stopped me is about vintage. Businesses founded in 2025 reached 10 percent AI adoption within six months of opening. The 2019 cohort took more than six years to get to the same place. Starting adoption rates tell the same story — the 2024 cohort opened at 4.1 percent, the 2025 cohort at 6.5 percent, better than a fourfold rise across six years of founding classes.
Size still matters, but less than you'd think. By the end of 2025 the transaction data puts large firms around 22.5 percent and small firms around 15.1 percent. That's a real gap, and it is nowhere near the chasm the marketing implies.
Here's why I think this is the most useful number in the pile. It isn't measuring how sophisticated anyone is. It's measuring how long a business waits before trying. A 2019 business had to unlearn a way of working; a 2025 business never had one. Which means the thing separating the two groups is a habit, and habits are the one input you can change without a budget.
JPMorganChase Institute, Small Business in the Age of AI
One question
Which rung are you on?
Awareness, Exploration, Application, or Fluency. Reply with one word — I read every one, and the answers shape what goes in this pillar next.
More in this pillar as it's written.
— Shana
Sources link to the original publisher, never to a summary.