The AI Digest · Issue #2 · March 2026
March 6, 2026 · bindlcorp.com · 7 min read
Last issue we covered the predictions — what the CEOs were saying, what the forecasts showed. This week the story changed shape. Oracle announced layoffs targeting specific job categories it expects AI to handle. Harvard published the first real measurement of how the job market has already split. And the Dallas Federal Reserve identified exactly which workers are winning and which aren’t. The predictions are becoming data. Here’s what it means for you.
This Week
Oracle Is Cutting Thousands Of Jobs. The Reason They Gave: AI.
Oracle announced this week it’s laying off thousands of employees across multiple divisions, with cuts aimed specifically at job categories the company expects AI to replace. Not a restructuring. Not a market slowdown. AI — and it’s happening this month.
Oracle has 164,000 employees. They just committed $300 billion to build out AI data center infrastructure — one of the largest capital bets any company has made on AI to date. The logic is straightforward: spend on the infrastructure, cut the headcount the infrastructure will replace. Deutsche Bank analysts predicted this week that “AI redundancy washing” — companies using AI as the stated reason for cuts that were already planned for other reasons — will be a defining feature of 2026. They’re probably right. But whether a given layoff is genuinely AI-driven or AI-labeled, the direction is the same. Companies are rewiring their cost structures, and labor is in the crosshairs.
$300B
Oracle’s AI infrastructure commitment — announced the same week as the layoffs
164K
Oracle employees worldwide — cuts targeting AI-replaceable categories specifically
2026
Deutsche Bank: “AI redundancy washing will be a significant feature of the year”
What to do with this
If you’re at a large enterprise, pay attention to where AI investment is flowing internally. Divisions getting AI tools are building leverage — they produce more with fewer people. Divisions that don’t are becoming cost lines. Knowing which one you’re in isn’t paranoia. It’s information.
If you’re a small business owner, the Oracle story is relevant in a different way: you’re not going to be a victim of a corporate restructuring decision, but you’re competing against businesses that are cutting costs with AI. The ones who move first on efficiency keep the margin. Worth thinking about which tasks in your operation are costing you hours that a tool could handle.
The Data
Harvard Measured The Job Market Split. It’s Already Three Years Old.
Harvard Business School published new research this week analyzing nearly every U.S. job posting from 2019 through early 2025. The finding: since ChatGPT launched, postings for routine, automation-prone roles dropped 13%. Over the same period, postings for analytical, technical, and creative roles grew 20%.
This is the first time that split has been measured at scale across the full U.S. job market — not modeled, not projected, but counted from actual listings. The gap started the month ChatGPT became widely available and has been growing since. It is not coming. It came. The question now is whether you know which side of it you’re on.
−13%
Job postings for routine, automation-prone roles since ChatGPT launched (Harvard, 2026)
+20%
Job postings for analytical, technical, and creative roles over the same period
What to do with this
Look at actual job postings for your field right now — not descriptions of your industry, but real current listings on LinkedIn or Indeed. Are the number of roles growing, flat, or shrinking? Is the language in the job descriptions changing? Are they asking for things you don’t have? This is now searchable and measurable. You don’t have to guess.
Spend 20 minutes this week doing that search for your own role. What you find is more useful than any forecast.
Who’s Winning
The Federal Reserve Identified Exactly Who AI Is Helping — And Who It Isn’t.
The Dallas Federal Reserve published an analysis this week that cuts through a lot of the noise. Their question: AI is affecting the job market, but who specifically is it affecting, and how? The finding is more precise than most coverage: AI is splitting workers not by industry or job title, but by knowledge type.
Workers who rely mostly on codified knowledge — textbook skills, documented procedures, established rules — are being replaced. AI has absorbed that material and can apply it faster and cheaper. Workers who rely on tacit knowledge — judgment built from years of real situations, relationships, context that can’t be written down — are being augmented. AI makes their experience more productive, not obsolete.
The wage data is unambiguous. Since late 2022, nominal wages nationally rose 7.5%. In AI-exposed sectors, experienced workers saw wages rise 16.7% — more than double the national average. Over the same period, the job-finding rate for workers under 25 dropped significantly. AI is concentrating gains in experienced workers and making the entry-level path harder to navigate. The experience premium has become the AI premium.
16.7%
Wage growth for experienced workers in AI-exposed sectors since late 2022 (Dallas Fed)
7.5%
National average wage growth over the same period — less than half the AI-sector premium
Under 25
Age group seeing significantly lower job-finding rates in AI-exposed fields
What to do with this
If you have experience: the leverage is real, but only if you’re using AI tools to multiply it. Experience without AI fluency is still valuable — for now. Experience with AI fluency is what’s commanding the 16.7% premium. The gap between those two positions is growing every month.
If you’re early in your career: the traditional path — get a job, learn the ropes, work your way up through routine tasks — is the path that’s being disrupted most aggressively. The roles that used to be the entry point are the ones disappearing fastest. Moving deliberately toward analytical, creative, or relationship-heavy work, and building AI fluency alongside it, is a more durable position than waiting for the market to stabilize.
This Week’s Move
Stop Hugging Your Job. Start Building Leverage In It.
CNBC reported this week on a growing behavioral pattern researchers have started calling “job hugging.” People so anxious about AI disruption that they’re freezing entirely — not applying anywhere, not asking for more, not making a move. Just holding on to what they have and hoping nothing changes. It’s understandable. It’s also the wrong move.
Staying still while the job market shifts around you isn’t a protected position. The Harvard data shows the split is already three years in and widening. The Oracle news shows it’s no longer theoretical. Freezing doesn’t keep you on the safe side — it just delays the moment you have to respond, while shrinking the options available when you do.
The specific move this week: identify one thing you know how to do from experience that AI can’t do on its own. Something built from years of judgment, relationships, or context. Then figure out how an AI tool can make that specific thing faster or more visible. That combination — your knowledge plus the tool’s speed — is what the Dallas Fed data shows is winning. Build it deliberately, not by accident.
Everything Else This Week
Quick Hits
GPT-5.3 is already on your phone — OpenAI pushed GPT-5.3 Instant to all users this week with no announcement. More accurate answers, better web search, less unnecessary hedging and caveating. If you used ChatGPT last month and haven’t tried it this week, you’re using yesterday’s version without knowing it. Open the app and ask it something you actually need answered.
The Bank of England is war-gaming an AI shock — The UK’s central bank announced it will formally scenario-plan a “full-blown AI shock” — and is considering adding it to the same stress tests used for bank failures and financial crises. For context: the last time they added something to those stress tests was 2008. When central banks start modeling AI disruption like a systemic economic risk, that’s institutional language worth paying attention to.
DeepSeek V4: 1 trillion parameters, fraction of the cost — China’s DeepSeek released a model hitting 1 trillion total parameters while activating only 32 billion per query. Massive capability, much lower compute cost. Competitive AI is no longer a US story. The pace of releases from Chinese labs is accelerating, and the cost-per-capability curve is bending in ways that change the economics for every business considering AI adoption.
Apple Visual Intelligence — Tim Cook said this week that Visual Intelligence — pointing your phone at something and getting useful AI context — will be the defining feature of Apple’s AI push. It’s shipping on the new MacBook lineup this month, eventually expanding to wearables. The AI on your phone is about to become significantly more situationally aware of the physical world around you.
OpenAI Codex lands on Windows — OpenAI’s coding agent is now available on Windows for enterprise users. It runs multiple AI coding agents in parallel on isolated workstreams. For non-developers: this is the version of AI that can be handed a software project and work through it autonomously. The gap between what AI can do in a dev environment and what it can do in a standard office workflow is narrowing faster than most people realize.
New to AI tools?
The Tools Just Got Better. Here’s How To Actually Use Them.
GPT-5.3 landing on everyone’s phone this week without announcement is a good reminder that these tools are improving whether you’re using them or not. The gap between people actively building AI fluency and people watching from the sideline gets a little wider every time that happens.
If you’ve been meaning to start but haven’t, the Dallas Fed data this week makes the case more concretely than anything else could: the workers adding AI fluency to their existing experience are the ones whose wages are rising. That’s not abstract anymore. It’s in the numbers.
Free guide — no account needed
Your First 20 Minutes With AI
Three tools. Three activities. 20 minutes. Built for people who know AI exists but haven’t figured out how to make it actually useful. No experience required.
Read the Beginner’s Guide →Free Weekly
The AI
Digest
What actually happened in AI this week. What it means for your job, your money, and your life. Plain English. Every Friday. Free.
No spam. Unsubscribe anytime. Plain English. Every Friday.
Leave a Reply