Anthony Pompliano built his AI stock research service in two weeks, with one employee, for a few thousand dollars in costs.
He's now charging investors $2,500 a year to subscribe to it.
The product is called ProCap Insights.
It's pitched as a fleet of AI agents scanning markets, analysing trends and drafting reports with titles like
"3 Stocks That Win From Both Tariff Refunds and the Iran Oil Shock"
These agents can generate hundreds of reports a day, though Pompliano says he'll start with just a few to avoid flooding inboxes.
In his own words, a supersonic tsunami of AI is coming to finance.
Good line… shame that the research will be crap.
Because before we get to the AI output, let's look at what Pompliano himself has been publishing…
These are real newsletter headlines from his company over the past several months:
"Investors Are Confused Because We Have Positive And Negative Data"
"There Is Blood In The Streets Across The Stock Market"
"Everything Is Speculation and Everything Is A Meme"
"Why 'Buy the Dip' Keeps Working Even When Everyone Says the Economy Is Terrible"
Read through those… and see if you can notice what's missing.
Not one of them tells you anything you couldn't have worked out yourself from five minutes of scrolling.
No edge or unique insights… just someone describing the weather after it's already rained on you.
This is what I call consumption-grade content
Material designed to be read/watched/listened to and then immediately forgotten.
It's fluent enough to sound credible… and vague enough that it can never be wrong.
And from a business point of view, it's cheap enough to produce at scale, which is the whole point.
The problem for retail investors is that this consumption-grade content is masquerading as decision-grade research.
There's a difference between the two that matters enormously when real money is involved.
Decision-grade research tells you something specific like a risk, a catalyst or a timeframe. That's what makes it useful. Plus, the author should at least have skin in the game (hence why I disclose all my holdings)
The worst part is, the AI revolution won't fix this problem… in fact it will accelerate it.
Generating 100+ vague, well-formatted reports a day is precisely what large language models are built for.
The technology is very good at sounding authoritative, and rather less good at being right about specific things on specific timelines.
To his credit, Pompliano acknowledges the reports won't include buy or sell recommendations (which naturally strips out the one thing that would make them actually useful to an investor)
What you're left with is expensively-packaged midwit macro commentary that covers all bases and commits to none of them... delivered faster and cheaper than ever before.
So yes, a supersonic tsunami is coming.
Most of what it washes up will be content you'll read, nod at… then forget by lunch.
Oliver