Tout le monde achète de l’IA, mais l’argent bloque au même endroit

Tout le monde achète de l’IA, mais l’argent bloque au même endroit

Everyone is buying AI, but the money is getting stuck in the same place

Title history
  • Tout le monde achète de l’IA, mais l’argent bloque au même endroit — observed on 20261102-1656 UTC (YouTube)

by IA et Stratégie · View the channel on PSI

Humanities, Social Sciences & Thought Economics & Finance KCVEconomics of specific sectorsKCVMDigital or internet economics
🎙 IA et Stratégie 👥 86K 📅 November 2, 2026 ⏱ 22 min 👁 209 📄 expert opinion 🧭 2026-11-02
Available in: English (current) Français

Keywords

déploiement IAmise en productionindépendantsconseil stratégiqueadoption IA

Summary

The video analyzes why companies buy AI but struggle to get real value from it. The creator identifies four closed doors: thin wrappers, generic automation agencies, content with zero marginal cost, and internal pilots. He emphasizes the ’last mile’ problem—the gap between a prototype and a stable production system—where the true financial value lies. He recommends a strategy for independents: choose a vertical you know, write a fixed-scope offer with a delivery date, and adopt a ‘deployment engineer’ approach similar to Palantir or Harvey. He cites data from IBM, Stanford, MIT NANDA, and others showing that external partnerships significantly increase production success rates. Examples from Aldes and Selectour illustrate successful implementations. He warns against generic solutions and certifications, advocating for selling judgment and responsibility. Finally, he suggests that salary trends (PwC) and marketplaces (Upwork) reward strategic roles over execution-only roles, urging viewers to position themselves where decisions are made.

149 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video offers a clear and nuanced analysis of the AI business landscape, combining quantitative data with qualitative observations from the creator’s consulting background. It distinguishes between commoditized offerings and high-value services tied to complex integration and accountability. The argument is logically structured: it starts with a diagnosis of market failures, then presents evidence from recent reports and case studies, and concludes with actionable steps. The creator critically acknowledges potential biases in his sources, which strengthens his credibility. The reasoning is solid, though some generalizations (e.g., the exact 50% friction figure) are more illustrative than empirical.

Scientific Rigor, Source Quality, Title Accuracy

The video draws on a wide range of sources, including major industry reports (IBM CEO study, Stanford AI Index, PwC AI Jobs Barometer, Upwork financials), academic research (MIT NANDA), and corporate announcements (OpenAI revenue, Harvey funding, Salesforce-Anthropic). The creator also references European regulatory updates (AI Act) and a Bpifrance white paper featuring French companies. He transparently notes when figures come from interested parties and tempers his claims accordingly. The title accurately reflects the core message about money being blocked in the production gap. The video does not explicitly mention any discordant sources, but it could benefit from contrasting voices, such as skeptics of AI ROI.

216 words

Title / Content Match

The title matches the content well: the video explains that while companies widely purchase AI, the main bottleneck is moving from pilots to production, where value and money concentrate.

Quality & Reliability

7/10

The video provides a synthesis of multiple cited reports and real-world examples, but many sources (IBM, PwC) have commercial interests, which the creator acknowledges. The analysis is logically structured and the advice is pragmatic, though some figures are estimates from involved parties.

Key Moments

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

Cited Sources

Concurring Sources

  • IBM CEO Study — Supports the claim that many companies are still in experimentation phase.
  • MIT NANDA Report — Confirms the advantage of external partners in reaching production.
  • PwC AI Jobs Barometer — Aligns with the video's thesis on the rising value of judgment-based roles.

Contribution & Novelties

The video contributes a practical framework for AI professionals, moving beyond hype to identify the ’last mile’ problem as the main value bottleneck. It synthesizes recent data and provides concrete examples from French companies. The originality lies in framing the deployment transition as a business opportunity that can be monetized by independent consultants, and in warning against common dead ends. For further exploration:

105 words

Radar Profile

The profile shows a balanced analysis: high information quantity and quality, moderate technical depth, and a reliability score slightly reduced due to the use of commercially interested sources. The overall score reflects strong practical value with some caution on data biases.

Reliability 7/10