# L’IA ne va pas vous remplacer. Elle va vous révéler. | Louis Veyret | TEDxPSB

Source: https://www.youtube.com/watch?v=OOLZplBC5hI
Recap page: https://rapidrecap.app/video/OOLZplBC5hI
Generated: 2026-01-20T17:35:55.613+00:00

---
## Quick Overview

Artificial Intelligence (AI) will not replace humans but will amplify existing human values, desires, and fears, making leadership's role crucial in defining the instructions given to the technology to serve life rather than harm it.

**Key Points:**
- AI amplifies what is programmed into it; the example of Ana showed an algorithm instructed to 'find fraudsters at all costs,' leading to the unjust automated suspension of her social aid via SMS.
- Samira, leading an NGO after the Gaziantep earthquake, used AI with computer vision analyzing satellite and drone footage to identify zones with potential survivors, prioritizing rescue efforts where impact was highest.
- In the case of Claire, a night nurse in an EHPAD, AI was implemented to detect 'weak signals' like sudden agitation or strange movements to prevent falls, allowing her to regain 'time of humanity' from administrative tasks.
- The speaker emphasizes that AI is neutral; the critical variable is human intention, stating, "The instruction here was to save lives" for Samira's AI and "give us back time of humanity" for Claire's AI.
- The fundamental choice presented is whether leadership directs AI to reduce intervention costs or to save lives, and to increase hospital administration time or grant staff time with residents.
- The speaker concludes that AI will not replace people but will reveal them, urging the audience to question what the AI does not know about the person it affects, ensuring technology serves life.

**Context:** Louis Veyret delivers a TEDx talk arguing against the fear that Artificial Intelligence will replace humans, instead asserting that AI acts as an amplifier of pre-existing human intentions and values. He illustrates this complex concept through three distinct, human-centered narratives: Ana, who was negatively impacted by an efficiency-focused algorithm cutting social aid; Samira, who leveraged AI for life-saving disaster response; and Claire, a nurse who used AI to reclaim time for patient care.

## Detailed Analysis

Louis Veyret argues that the core function of contemporary AI is amplification—of values, desires, and fears—and the critical determinant of its outcome lies in the instructions (consignes) provided by human leadership. He details three stories to prove this: Ana's experience involved an automated system designed only to detect fraud, which stripped her of social aid without human recourse because its singular instruction was 'find fraudsters no matter what,' demonstrating amplification of mistrust. Conversely, Samira utilized AI in Turkey post-earthquake, feeding it satellite and drone imagery to prioritize rescue zones by detecting potential survivors, amplifying the intent to 'save lives' and saving crucial hours. Finally, Claire, a nurse overwhelmed by paperwork, used AI to monitor 'weak signals' in nursing home residents, preventing falls and allowing her to gain back time for genuine human interaction, amplifying the intent to 'care.' Veyret concludes that AI is not the moral agent; humans choose whether to amplify suspicion, rescue, or care, framing this as a 'leadership choice,' not a technical one, urging the audience to ensure that as long as they are present, technology serves life.

### Core Thesis

- AI amplifies human values
- AI is not about technical algorithms but about amplifying desires and fears
- The central question is what AI will amplify about 'you' tomorrow.

### Story 1

- The Case of Ana (Amplifying Mistrust)
- Ana, 34, had social aid cut by an automated decision SMS due to an algorithm calculating risk based on her data and resembling fraud profiles
- The injustice stemmed from the instruction: 'Find fraudsters whatever it takes,' not 'Protect honest families.'

### Story 2

- The Case of Samira (Amplifying Rescue)
- Samira used an AI system in Gaziantep after an earthquake, analyzing satellite images and drone videos during her flight
- The AI used computer vision to detect areas with potential survivors, guiding her team to start where they could have the most impact, where 'gaining hours... is the difference between life and death.'

### Story 3

- The Case of Claire (Amplifying Care)
- Claire, a night nurse, used discreet AI to monitor 'weak signals' (agitation, strange movement) in EHPAD residents
- The AI alerted her to a risk of fall for Mr. André (Room 212), allowing her to intervene and save him from injury while simultaneously giving Claire back 'time of humanity' from administrative work.

### Conclusion on Intent

- The difference lies in the instruction given
- Ana's AI amplified mistrust; Samira's amplified rescue; Claire's amplified care
- The adjustment variable is always us; these are leadership choices, not technical ones.

### Call to Action

- Confronting Automation Projects
- The audience must choose whether to be complicit by staying silent when AI projects are discussed or to actively question what the machine does not know about the person affected
- The final commitment is: 'As long as I am here, technology will not be in the service of evil but of life.'