diff --git a/Sections/4-Tentations.tex b/Sections/4-Tentations.tex
index 5f67f2e5873178bb8896a8dd753d09cf2bdb46e8..3f7a102b0a447dc6cd95322b28d121eff2d5876e 100644
--- a/Sections/4-Tentations.tex
+++ b/Sections/4-Tentations.tex
@@ -278,16 +278,54 @@ plt.show()
 
 \begin{frame}[fragile]
 \frametitle{A bit more meta-physical questions}
-    \begin{block}{Why it's not a good idea?}
+    \begin{block}{Why it's not a good idea (1/5)?}
     \begin{itemize}
-        \item Plagiarism? Just cite the generative AI used as the source with date, but output is inherently non-reproducible!\\
-            We should also provide the exact sequence of prompts used and it would not be enough!
-        \item Ethics? Generative AIs are "an average" of other people's work (impossible to find the original source)\\
-            What if you copy-pasted an exact sentence from another author without knowing it?
-        \item Intellectual property? Who is the "owner" of the content produced?
-        \item Skills evolution? To use AI one must be expert in its field, by using AI one becomes less an expert\\
-            Each time you use an AI you make \textbf{zero} effort (no rephrasing, no creativity, etc)
-        \item Energy concerns? Each AI powered request consumes 10-20x a google search (70-140g of CO2)
+        \item Plagiarism? Recommended to cite the generative AI used as the source with date!\\
+            Not good enough because: non-reproducible.\\
+            Even providing the exact sequence of prompts used does not allow to get the source!
+    \end{itemize}
+    \end{block}
+\end{frame}
+
+\begin{frame}[fragile]
+\frametitle{A bit more meta-physical questions}
+    \begin{block}{Why it's not a good idea (2/5)?}
+    \begin{itemize}
+        \item Ethics? Generative AIs are "averaging" other people's work (impossible to find the source)\\
+            Reproduction of a sentence from another author without even knowing it without proper credit.
+        \item Bing: "Unlike solving the Navier–Stokes equations directly, LBM operates on a lattice, simulating fluid density through streaming and collision processes."
+        \item Wikipedia: "Instead of solving the Navier–Stokes equations directly, a fluid density on a lattice is simulated with streaming and collision (relaxation) processes."
+    \end{itemize}
+    \end{block}
+\end{frame}
+
+\begin{frame}[fragile]
+\frametitle{A bit more meta-physical questions}
+    \begin{block}{Why it's not a good idea (3/5)?}
+    \begin{itemize}
+        \item Intellectual property? Who is the "owner" of the content produced?\\
+            This question is not fully answered and will be in the coming years as legislation evolves.
+    \end{itemize}
+    \end{block}
+\end{frame}
+
+\begin{frame}[fragile]
+\frametitle{A bit more meta-physical questions}
+    \begin{block}{Why it's not a good idea (4/5)?}
+    \begin{itemize}
+        \item Skill evolution? No skills improve by using an AI\\
+            \textbf{Zero} effort game (no rephrasing, no creativity, etc)\\
+            Doing research is also the ability to explain to others clearly and developing a wide variety of skills.
+    \end{itemize}
+    \end{block}
+\end{frame}
+
+\begin{frame}[fragile]
+\frametitle{A bit more meta-physical questions}
+    \begin{block}{Why it's not a good idea (5/5)?}
+    \begin{itemize}
+        \item Energy concerns? AI powered search consumes 10-20x a google search and increasing (70-140g of CO2).\\
+            This presentation cost about 10kg of CO2 (0.5\% of my "ideal" consumption or two days of emissions) 
     \end{itemize}
     \end{block}
 \end{frame}