diff --git a/content/2023/reinforcement-learning/notebooks/reinforcement-learning.ipynb b/content/2023/reinforcement-learning/notebooks/reinforcement-learning.ipynb
index fedf65dfa..2226c55ec 100644
--- a/content/2023/reinforcement-learning/notebooks/reinforcement-learning.ipynb
+++ b/content/2023/reinforcement-learning/notebooks/reinforcement-learning.ipynb
@@ -7,7 +7,9 @@
    "source": [
     "In our final exploration into machine learning with PyTorch, we're going to do something critical for lifeforms in our world, learn to walk!\n",
     "\n",
-    "This post took many trials and errors, a form of reinforcement learning I completed unsupervised as a human. The resulting code below was what ended up working on a M1 (M2) macbook pro. As many other researchers have implemented much better training algorithms that I could develop on my own, we'll make use the of the work from OpenAI, MuJoCo (multi joint control) and Stable Baselines3. If you're interested in how it may be implemented, there's a separate notebook using PyTorch to implement a Deep Q Learning agent to teach our model to walk at [this blogs repository](https://github.com/JackMcKew/jackmckew.dev/tree/main/content/2023/reinforcement-learning/notebooks/torch-rl.ipynb)."
+    "This post took many trials and errors, a form of reinforcement learning I completed unsupervised as a human. The resulting code below was what ended up working on a M1 (M2) macbook pro. As many other researchers have implemented much better training algorithms that I could develop on my own, we'll make use the of the work from OpenAI, MuJoCo (multi joint control) and Stable Baselines3. If you're interested in how it may be implemented, there's a separate notebook using PyTorch to implement a Deep Q Learning agent to teach our model to walk at [this blogs repository](https://github.com/JackMcKew/jackmckew.dev/tree/main/content/2023/reinforcement-learning/notebooks/torch-rl.ipynb).\n",
+    "\n",
+    "![Walking agent]({static img/ant-walking.gif})"
    ]
   },
   {