[CESERH] CI La généalogie des formes dans la perspective du Deep Learning 8/11/2019 [2/9]

[CESERH] CI La généalogie des formes dans la perspective du Deep Learning 8/11/2019 [2/9]

[CESERH] CI The genealogy of forms in the Deep Learning perspective 8/11/2019 [2/9]

by Pierluigi Basso Fossali · View the channel on PSI

Humanities, Social Sciences & Thought AI — Models & Research UYQArtificial intelligence (AI)UYQNNeural networks and fuzzy systems
🎙 Pierluigi Basso Fossali 👥 178 📅 November 19, 2019 ⏱ 10 min 👁 50 📄 debate 🧭 2026-10-10
Available in: English (current) Français

Keywords

généalogie des formesdeep learningsémiotiquehumanités numériquesinterprétation

Summary

This video captures a segment of an academic seminar on semiotics, where the speaker discusses the application of deep learning to the genealogy of visual forms. The conversation addresses the possibility of using computational models to analyze and interpret visual structures across historical periods, particularly focusing on the 16th and 17th centuries. Key themes include the tension between original and present-day intentionality, the concept of semantic depth in works of art, and the idea that forms can create new interpretive regions not defined by external criteria. The speaker proposes working with interfaces and transpositions, acknowledging the need to balance historical contextualization with contemporary readings. The discussion also explores whether visual forms can generate original doxa beyond verbal encyclopedias, and how computational systems could handle layered interpretations. The exchange highlights the interdisciplinary nature of the project, blending semiotics, art history, and digital humanities.

142 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of this discussion lies in its attempt to bridge deep learning with semiotic theory, offering a nuanced perspective on how computational methods can be applied to historical visual analysis. The argumentation is solid, rooted in theoretical frameworks from semiotics and art history, and shows a critical awareness of the epistemological challenges. The speaker engages with audience questions, refining ideas on intentionality, doxa, and the role of interfaces. While the argumentation is dense and assumes a scholarly audience, it presents original concepts such as ‘regionalization’ of diachronic analysis and the construction of semantic depth through computational tools.

Scientific Rigor, Source Quality, Title Accuracy

The discussion is rigorous in its academic framing, consistent with current debates in digital humanities and semiotics. However, no explicit sources are cited within the video, and no links appear in the description; the authority relies on the speaker’s expertise and the seminar setting. The title adequately reflects the content, and the video’s brevity limits the depth of any single argument. The audience’s engagement, as seen in the questions, indicates a high level of scholarly exchange, though the fragmentary transcription makes it difficult to verify all claims independently.

201 words

Title / Content Match

The title accurately reflects the content, which discusses the application of deep learning to the historical analysis of visual forms.

Quality & Reliability

7/10

The discussion is part of an academic seminar, with an expert speaker and engaged audience; the transcript shows high-level theoretical reflection, though the fragmentary nature of the excerpt limits its self-contained reliability.

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Contribution & Novelties

The video contributes original ideas on integrating deep learning with semiotic genealogy, proposing a methodological framework where forms are analyzed as dynamic entities that can generate new interpretive categories. The emphasis on ‘interfaces’ and ’transpositions’ offers a fresh angle for computational humanities.

Pour aller plus loin :

  • Digital humanities — Relevant for situating the intersection of computing and humanistic inquiry.
  • Semiotics — Provides foundational concepts used throughout the discussion.
  • Deep learning — The core technical perspective driving the proposed methodologies.

80 words

Radar Profile

The profile shows a balanced moderate assessment, with slightly higher quality of information relative to quantity, consistent with a brief but conceptual seminar discussion.

Reliability 6/10