The Prompt Did the Shape. It Didn't Do the Substance.

The Prompt Did the Shape

5:47

A post crossed my feed this week with a T-Rex crushing a McKinsey building in its jaws, “AI” and “Anthropic” stamped across its side. The copy underneath: twenty Claude prompts that supposedly replace an entire finance consulting team. A three-statement financial model, priced at “$9,000 Big 4 engagement.” Cash flow forecasting, “$6,500 CFO-level build.” A SaaS metrics dashboard, “$8,000 SaaS CFO template.” Working capital optimization, “$6,000 McKinsey engagement.” Twenty line items in total, each stamped with a consulting price tag, adding up to “$105,500+ in consulting value” for a cost of zero dollars. Finance isn't complicated, the post said. It's just been gatekept by consultants charging $400 an hour to build models you can now generate in thirty seconds.

It landed, and it landed hard, judging by the engagement. It also landed because there's a real, true thing underneath the hype, and I want to give that its due before I get to the part that's wrong.

The True Part

I wrote a piece a while back called Skills-as-a-Service: The Next Great Gold Rush, and the core argument was that the templated, formulaic layer of expensive professional services was about to get commoditized fast, and the people who moved early on that shift would capture enormous value while everyone else was still arguing about whether it was real. This post is that argument playing out in real time, in public, with a viral graphic attached. The shape of a three-statement model, a cap table structure, a break-even framework, a debt amortization schedule — that scaffolding is genuinely something a capable model can produce in seconds now. It used to take a trained analyst hours to build from a blank sheet. That part isn't hype. That part is just what's true now, and if you're not paying attention to it, you're going to get run over by someone who is.

The Part That's Wrong

Here's what that $105,500 figure is actually pricing: the shell, not the deliverable. A Big 4 engagement was never priced at $9,000 because nobody else could type formulas into a spreadsheet. It was priced there because the number on the page had to survive contact with a partner, an investor, or an auditor who was going to poke at every assumption underneath it. The template is not the product. The judgment about your actual burn rate, your actual revenue recognition policy, which of your customers are going to churn, and whether your unit economics assumptions would survive a diligence call — that's the $9,000. Claude can hand you the shape of a cash flow forecast in thirty seconds. It cannot tell you whether your churn assumption is realistic, because it doesn't know your business. It knows the pattern of what a cash flow forecast looks like.

This is the exact same gap I wrote about a few days ago, looking at companies racing to call themselves AI-native. The failure pattern there was a company automating a role, discovering the AI handles about 60% of the actual job, and quietly rehiring for the 40% it can't touch. The finance version of that gap is someone generating a beautiful investor-ready model in thirty seconds, walking into a fundraising conversation with it, and getting shredded the first time someone asks where the churn number came from. The prompt did the shape. Nobody did the substance.

Two Things Are True at Once

The gold rush is real. The people who figure out how to use these tools to produce the scaffolding of expert work at a fraction of the previous cost and time are going to build enormous advantages over the people who don't. That's not in question, and it's happening faster than most professional services firms are willing to admit publicly. The slop risk is also real, in the exact same motion. Mistaking a generated template for a validated deliverable is how you walk into a room confident and leave it embarrassed. The gap between “I generated a financial model” and “I built a financial model that will survive scrutiny” is precisely the gap that used to be billed at $400 an hour, and it hasn't actually closed. It's just gotten harder to see, because the output looks finished.

What Actually Changes

The honest version of this story isn't “consultants are obsolete.” It's that the scaffolding work — the blank-page problem, the hours of building structure from nothing — is now nearly free, and the value has concentrated entirely into the judgment layer on top of it. The people who win this transition aren't the ones posting the viral prompt list. They're the ones who use that free scaffolding to get to the judgment question faster, and who still know enough to ask whether the number in front of them would survive someone actually checking it.

Finance was never complicated because it was gatekept. It was gatekept because getting it right requires knowing which assumptions are real, and that part didn't just get commoditized by twenty prompts, no matter how good the T-Rex graphic is.

---

Ad: Use code: RICH99 for a discount.

Rich Washburn is a technologist and strategist working at the intersection of AI, infrastructure, and capital. He is Managing Partner and Chief AI Officer at Eliakim Capital and CIO of Data Power Supply.