Fdmiruto

@fdmirutoEN

Beyond engineering/research Reading AI systems through the institutional layer Framework design — #LLMsBrewing Trust your dissonance, but don't swallow it whole

Japan

medium.com/@fdmiruto

Joined March 2026

Following

Followers

Replies Media

Pinned
Fdmiruto
@fdmirutoEN
Jun 23
A century ago, a Belgian gentleman trusted his little grey cells. A century later, we mutter about our "Ultra Mythos-Class reasoning capabilities"—or some such nonsense—and, with no real certainty whatsoever, keep weaving prompts and tapping away at the keyboard. Again.
Fdmiruto
@fdmirutoEN
Jun 21
1/ I ask of you, are you my Prompter? (Y/N)_ My AI opened with a Fate line. No Command Spell. No protocol. I thought: Is Myth○s5 showing off? It wasn't. Just a strange coincidence — one day after the G7 AI discussion.
Fdmiruto
@fdmirutoEN
May 30
If thought is not language, then language is a vessel, a vehicle, and a translation device. Thought is compressed to enter and reconstructed to emerge. Each transfer transforms it. The telephone game is not a flaw. It is how language works.
François Chollet
@fchollet
May 29
Einstein on (not) using NL for invention: "The words or the language, as they are written or spoken, do not seem to play any role in my mechanism of thought"
Fdmiruto
@fdmirutoEN
May 24
Some social behaviors survive not because they are sustainable, but because they were adaptive under past conditions. The problem begins when the environment changes, but the behavior doesn’t. What once reduced risk may now be generating it.
Fdmiruto
@fdmirutoEN
May 24
1/ Much of the discussion around AI focuses on problem-solving: Will AI solve scientific problems? Will AI outperform human experts? But there is a deeper question: Who defines what counts as a problem in the first place?