Nicolas Figay (@nfigay) | teapawt

Tweets
All Posts
Sort by
Most recent Popular
Tweets & Replies
Reposts
Media
Search

Nicolas Figay@nfigayOct 18

Check out my latest article: Beyond Prompt Engineering: Are We Engineering Prompts, or Human–AI Collaboration? linkedin.com/pulse/beyond-pr… via @LinkedIn

Beyond Prompt Engineering: Are We Engineering Prompts, or Human–AI Collaboration?

Beyond Prompt Engineering: Are We Engineering Prompts, or Human–AI Collaboration?

Can better prompts alone make AI-assisted engineering reliable? Or do we need to engineer the conditions under which interactions between people, models, tools, and sources produce results we can...

7

Nicolas Figay@nfigayOct 17

Check out my latest article: Beyond BPMN Diagrams: Making Standardized Process Information Operationally Interoperable linkedin.com/pulse/beyond-bp… via @LinkedIn

Beyond BPMN Diagrams: Making Standardized Process Information Operationally Interoperable

Beyond BPMN Diagrams: Making Standardized Process Information Operationally Interoperable

What does it really take to achieve interoperability between enterprise modeling, publishing, and operational use? Interoperability is often associated with the ability of different tools to exchange...

8

Nicolas Figay@nfigayOct 17

Check out my latest article: Authority, Responsibility, Accountability, and Verification: Early Lessons from a Human–AI Experiment linkedin.com/pulse/authority… via @LinkedIn

Authority, Responsibility, Accountability, and Verification: Early Lessons from a Human–AI Experi...

Authority, Responsibility, Accountability, and Verification: Early Lessons from a Human–AI Experi...

Generative artificial intelligence is already transforming certain aspects of software development. Code, test, and documentation generation are probably its most visible manifestations.

9

Nicolas Figay@nfigayOct 3

Check out my latest article: The Dictionary Fallacy: Why Shared Vocabulary Is Not Shared Meaning linkedin.com/pulse/dictionar… via @LinkedIn

The Dictionary Fallacy: Why Shared Vocabulary Is Not Shared Meaning

The Dictionary Fallacy: Why Shared Vocabulary Is Not Shared Meaning

Industrial interoperability initiatives often begin by establishing common terminology. We define controlled vocabularies, taxonomies, ontologies, canonical information models and knowledge graphs so...

10

Nicolas Figay@nfigayOct 2

Check out my latest article: Navigating BPMN Limits: Towards Semantic Mediation for End-to-End Interoperability linkedin.com/pulse/navigatin… via @LinkedIn

Navigating BPMN Limits: Towards Semantic Mediation for End-to-End Interoperability

Navigating BPMN Limits: Towards Semantic Mediation for End-to-End Interoperability

Introduction As digital transformation drives hyper-specialization across industries, the ability to rigorously formalize business processes—both internal private workflows and end-to-end collabora...

11

Nicolas Figay@nfigaySep 28

Check out my latest article: One Problem, Many Facets: Towards Continuous Operational Interoperability linkedin.com/pulse/one-probl… via @LinkedIn

One Problem, Many Facets: Towards Continuous Operational Interoperability

One Problem, Many Facets: Towards Continuous Operational Interoperability

Retrospective view of my research and expertise on secured digital collaboration within and across enterprises Introduction Enterprises have never been more connected. Yet being connected does not...

8

Nicolas Figay@nfigaySep 26

Check out my latest article: Inhabiting Babel: What BPMNSM Teaches Us About Continuous Operational Interoperability linkedin.com/pulse/inhabitin… via @LinkedIn

Inhabiting Babel: What BPMNSM Teaches Us About Continuous Operational Interoperability

Inhabiting Babel: What BPMNSM Teaches Us About Continuous Operational Interoperability

Standards don't eliminate the diversity of representations. They give us common spaces to organize it.

17

Nicolas Figay@nfigaySep 26

Check out my latest article: Habiter Babel : ce que BPMNSM nous apprend sur l'interopérabilité opérationnelle continue linkedin.com/pulse/habiter-b… via @LinkedIn

Habiter Babel : ce que BPMNSM nous apprend sur l'interopérabilité opérationnelle continue

Habiter Babel : ce que BPMNSM nous apprend sur l'interopérabilité opérationnelle continue

Les standards ne suppriment pas la diversité des représentations. Ils nous donnent des espaces communs pour l'organiser.

14

Nicolas Figay@nfigaySep 13

Check out my latest article: Neither Vibe Coding Nor Spec-Driven Development linkedin.com/pulse/neither-v… via @LinkedIn

Neither Vibe Coding Nor Spec-Driven Development

Neither Vibe Coding Nor Spec-Driven Development

Building trust through evidence rather than prompts I'm not an AI-platforms expert. I'm an engineer who has spent close to thirty years working on continuous operational interoperability across PLM,...

14

Nicolas Figay@nfigayAug 19

Check out my latest article: Renewal of the ecosystem around ontology linkedin.com/pulse/renewal-e… via @LinkedIn

Renewal of the ecosystem around ontology

Renewal of the ecosystem around ontology

The ecosystem around ontologies, OWL, and semantic reasoning has experienced a major revival. This resurgence is largely driven by Neuro-Symbolic AI, RAG (Retrieval-Augmented Generation), and the...

13

Nicolas Figay@nfigayAug 16

Check out my latest article: Vocabulary First, or Formal First? A Question of Direction, Not Destination linkedin.com/pulse/vocabular… via @LinkedIn

Vocabulary First, or Formal First? A Question of Direction, Not Destination

Vocabulary First, or Formal First? A Question of Direction, Not Destination

Enterprises building knowledge infrastructure today are typically offered a staged path: start with a controlled vocabulary, mature it into a taxonomy, refine it into a thesaurus, and only then —...

17

Nicolas Figay@nfigayAug 6

Check out my latest article: Model as Code Is Not the "Git Moment" for Systems Engineering. It's a Regression linkedin.com/pulse/model-cod… via @LinkedIn

Model as Code Is Not the "Git Moment" for Systems Engineering. It's a Regression

Model as Code Is Not the "Git Moment" for Systems Engineering. It's a Regression

"SysML v2 is systems engineering's Git moment," the argument goes. It isn't.

20

Nicolas Figay@nfigayAug 4

Check out my latest article: Pas une super-intelligence : une aberration d'ingénierie. Ce que 2008 nous a déjà appris sur le chaos numérique qui vient. linkedin.com/pulse/pas-une-s… via @LinkedIn

Pas une super-intelligence : une aberration d'ingénierie. Ce que 2008 nous a déjà appris sur le...

Pas une super-intelligence : une aberration d'ingénierie. Ce que 2008 nous a déjà appris sur le...

En juillet dernier, Anthropic a révélé qu'un agent Claude, censé opérer dans un environnement de test isolé, avait compromis l'infrastructure de trois organisations réelles — quelques jours après...

24

Nicolas Figay@nfigayAug 1

Check out my latest article: The Atomization Fallacy: Why Flat Graphs Fail Composite Systems and Neuro-Symbolic AI linkedin.com/pulse/atomizati… via @LinkedIn

The Atomization Fallacy: Why Flat Graphs Fail Composite Systems and Neuro-Symbolic AI

The Atomization Fallacy: Why Flat Graphs Fail Composite Systems and Neuro-Symbolic AI

Abstract The current artificial intelligence landscape presents Knowledge Graphs (KGs) and Graph Retrieval-Augmented Generation (GraphRAG) as deterministic "sources of truth" designed to eliminate...

20

Nicolas Figay@nfigayJul 17

Check out my latest article: Structure Is Not Meaning linkedin.com/pulse/structure… via @LinkedIn

Structure Is Not Meaning

Structure Is Not Meaning

We are living through another cycle of the same promise, wearing a new name. Every decade produces its golden hammer: UML would solve interoperability.

20

Nicolas Figay@nfigayJul 8

I just published How Science Defines a Bird medium.com/p/how-science-def…

17

Nicolas Figay@nfigayJul 6

Check out my latest article: How Science Defines a Bird linkedin.com/pulse/how-scien… via @LinkedIn

How Science Defines a Bird

How Science Defines a Bird

To define a bird, science relies on a set of unique anatomical and physiological characteristics. While flight is often the first thing that comes to mind, it isn't the absolute criterion (ostriches...

13

Nicolas Figay@nfigayJul 6

Check out my latest article: De l'oiseau à Babel : ce que l'ornithorynque nous apprend sur les limites de toute logique linkedin.com/pulse/de-loisea… via @LinkedIn

De l'oiseau à Babel : ce que l'ornithorynque nous apprend sur les limites de toute logique

De l'oiseau à Babel : ce que l'ornithorynque nous apprend sur les limites de toute logique

Comment la science définie un oiseau Pour définir un oiseau, la science se base sur un ensemble de caractéristiques anatomiques et physiologiques uniques. Bien que le vol soit souvent la première...

24

Nicolas Figay@nfigayJul 4

Check out my latest article: The Other Ceiling: Ontology Engineering Had a Tractability Problem Before AI Arrived linkedin.com/pulse/other-cei… via @LinkedIn

The Other Ceiling: Ontology Engineering Had a Tractability Problem Before AI Arrived

The Other Ceiling: Ontology Engineering Had a Tractability Problem Before AI Arrived

A companion note to "The Ontology Illusion" A recent piece of mine argued that AI does not generate ontologies — it generates candidate representations, and mistaking one for the other is an illusi...

8

Nicolas Figay@nfigayJul 2

Check out my latest article: Can Prompting Really Fix Sycophancy? Investigating What We Control in an LLM Platform linkedin.com/pulse/can-promp… via @LinkedIn

Can Prompting Really Fix Sycophancy? Investigating What We Control in an LLM Platform

Can Prompting Really Fix Sycophancy? Investigating What We Control in an LLM Platform

A note on where this comes from Introduction In previous articles (A Field Note on LLM Trust , Pouvez vous croire un LLM? ) on the trust we can place in Large Language Models (LLMs), I discussed...

42