Founder of Meetme and Baboom

Romain de Wolff

Feeling incredibly grateful today 🙏

I had the privilege of publishing my very first opinion piece in a Swiss magazine, Bilan.

Seeing my words shared in a national newspaper, online, feels both surreal and deeply humbling.

This first publication is a small step on paper, but a huge one for me personally. It represents years of learning, experimenting, questioning, building, and above all, being surrounded by amazing people who supported, challenged, and trusted me along the way.

The article explores how AI assisted dictation is bringing lawyers back to the essence of their craft, voice, reasoning, intention, and freeing time for what truly matters. Writing about this shift, and seeing it resonate in such a respected Swiss publication, is a real honor.

A big thank you to the Bilan Magazine editorial team for the trust and the opportunity.

And thank you to everyone who believed in me, directly or indirectly, this would not have been possible without you.

Looking forward to writing and sharing much more. This is just the beginning ✨

Article: https://lnkd.in/dqQg2ffs - in french 🇫🇷🇨🇭

Open article: Avocats: la dictée assistée par IA révolutionne le métier | Bilan by bilan.ch

Avocats: la dictée assistée par IA révolutionne le métier | Bilan

Over the last 18 months, I have spent a lot of time onboarding lawyers to AI.

Hundreds of sessions, across firms, teams, cultures, sometimes very skeptical rooms.

What always strikes me is that the real challenge is not adoption.
It is understanding what AI actually is, and what it is not.

AI is not reasoning in the human sense.
At its core, it is statistics applied to language, predicting the next word based on probability.

Once lawyers get this, many things suddenly make sense.

Why early models were terrible at simple math problems.
Why accuracy degrades when you feed too much information at once.
Why context size matters far more than people think.

Give AI too much, entire case files, thousands of pages, and quality drops.
Not because it is lazy, but because of how it computes.

The best results come from restraint.
Carefully selected documents.
Clear questions.
Intentional scope.

The same logic applies to hallucinations, especially in legal matters.

When you ask an AI to recall a specific case or ruling from its training data, it may sound convincing and still be wrong.
Not maliciously, just statistically.

This is why legal research tools that first retrieve, then reason, change everything.
They ground the model in real sources before it speaks.

The lawyers who get the most value from AI are not the ones who trust it blindly.
They are the ones who constantly challenge it.

AI does not replace legal judgment.
It sharpens it, if used with discipline.

I keep seeing the same pattern.
Understanding the limitations is what unlocks real efficiency.

Curious to see how this collective understanding will evolve as models improve, and as our habits adapt with them.

Picture: in a law firm office in Geneva, during a training

No alternative text description for this image

taplio

The Battle

Romain de Wolff

Folco Banfi

Followers

Posts

Engagements

Likes

Performance ScoreRomain de Wolff's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Romain de Wolff

Folco Banfi

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Romain de Wolff's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreRomain de Wolff's LinkedIn Performance