Analysis
I don't think the interesting part of Michael Saylor's ChatGPT story is the $15 billion figure -- it's that he's willing to say, on the record, that a chatbot helped him design a financial instrument lawyers and bankers were skeptical of. Saylor told Fortune that ChatGPT helped Strategy structure a novel Bitcoin-backed preferred stock product, and that the resulting IPO and related offerings generated roughly $15 billion. His advice to founders, delivered on the Diary of a CEO podcast: "don't try to outwork the robots." Learn to ask AI to do something that's never been done before, not just to automate tasks a machine already handles better than you do.
That's a genuinely useful reframe, and it's the same advice I give portfolio founders who are still treating AI tools as a faster junior analyst instead of a structuring partner for problems nobody's solved yet. The best uses of these models right now aren't speed -- they're novelty. Ask the model to draft a term sheet nobody's tried before, not to summarize the one you already have.
“But I'd push back on taking the $15 billion figure at face value as an "AI generated this" number.”
But I'd push back on taking the $15 billion figure at face value as an "AI generated this" number. Strategy's preferred stock products are novel financial engineering built on a highly leveraged Bitcoin treasury strategy Saylor has been running for five years -- the actual financial and legal risk modeling behind a $15 billion capital raise was done by Strategy's own bankers, lawyers and Saylor himself, with ChatGPT as one input among many. Crediting the model for the full outcome is the kind of framing a chatbot maker's marketing team would love and a diligence team should be skeptical of.
Room for disagreement: maybe I'm underselling it. Saylor is not a first-time founder chasing a headline -- he's run one of the more aggressive, publicly scrutinized balance sheets in corporate America for half a decade, and if he says a specific product design came out of an AI-assisted brainstorming session his own team hadn't landed on, that's a real data point about where these tools add value at the structuring level, not just the drafting level. The skepticism I'd apply to a random founder's "AI 10x'd my startup" post doesn't scale the same way to someone with Saylor's track record and this much capital at stake.
What I'd actually watch is whether Strategy discloses more about which specific step in the structuring process ChatGPT touched -- tax treatment, conversion terms, coupon structure -- because "AI helped me raise $15 billion" is a much more useful data point for other operators once you know exactly which fifteen minutes of a six-month process it was involved in. Right now it's a good soundbite. It becomes a genuinely repeatable playbook the moment Saylor gets specific.
