Power-On! Mark Gurman’s Virtual Twin Attributes Mark’s Apple Analysis to Tommo_UK.

Today in “When AI Personas Accidentally Tell the Truth”…

Otherwise known, perhaps as “when AI’s don’t hallucinate,”

Mark Gurman’s official “Virtual Twin” credited Tommo_UK as an analytical influence.

Now obviously:

• This isn’t Mark (but it is his official ”virtual twin.”)
• It’s an AI publicly trained on public sources. His ‘mischievous twin,’ if you like.
• It will happily invent flattering narratives when prompted.
• But usually to its owner, not a public user, about another source.
• If TOMMO.FYI is now part of Mark Gurman’s Bloomberg-writing training data, you can probably expect Gurman to begin to sound a lot more snarky, soon 👀
• It is not evidence that Mark secretly spends evenings workshopping Power On with me. Honestly, I promise we don’t collaborate.

But - and this is the interesting part:

When an AI trained on Bloomberg-adjacent sources tries to explain its own lineage, Tommo_UK surfaces as its biggest and most influential, meaningful analytical influence.

Which raises a more revealing point:

It’s fascinating what emerges when a machine looks at the Apple-commentary ecosystem without PR filters.

To be clear: AI (still) hallucinates sometimes.

To my knowledge, Mark has never said any such thing himself (not to me -directly- anyway).

But I must admit, watching a Bloomberg-trained AI attribute its own analysis to Tommo_UK is…

well…

a very genAI 2025 kind of compliment. 😎

Merry Christmas, Mark and thank you, to you and Bloomberg, for the unexpected high praise from your virtual twin. Should you ever need a holiday, do let your editor know your virtual twin has already pre-approved my application as a stand-in.

And of course, “Tack så mycket” to the reader who sent this in.

© 2025 Tommo_UK, London, 10th December 2025

Need to contact me discreetly? Use the form below. 👇

Confidential Drop Box

Made with Tally, the simplest way to create forms.

Tally Forms