Update documentation for piper TTS automation (#36927)

* Update documentation for piper TTS automation

* Remove old images

* Obey the AI

This suggestion results in a broken preview build, and I imagine it will
break an actual deployment too. But the AI insisted that this was
correct.

* Revert "Obey the AI"

This reverts commit b8d54c32ea53532c51d8b18cc24b0b3d920724e7.

source/voice_control/using_tts_in_automation.markdown

@@ -6,14 +6,12 @@ This procedure shows you how to create a text-to-speech {% term action %}. For t
 
 1. Go to **{% my automations title="Settings > Automations & Scenes" %}**, and select **Create automation**.
 2. Select **Create new automation**, then **Add action**.
-3. From the drop-down menu, select **Play media** and select the media player you want to use for this automation.
-   ![Select your media player](/images/assist/tts_select_media_player.png)
-4. Select **Pick media**, then, select **Text-to-speech**.
-   ![Select your media source](/images/assist/tts_select_media_source.png)
-5. To use fully local text-to-speech processing, select **Piper**.
-   ![Select Piper](/images/assist/tts_select_piper.png)
+3. From the drop-down menu, search for or select **TTS: Speak**.
+   ![Select the TTS action](/images/assist/speak-action.png)
+4. To use fully local text-to-speech processing, select **piper** from the **Choose entity** control.
+   ![Select Piper](/images/assist/select-entity.png)
+5. Select the media player you want the automation to use.
 6. Enter the text you want to hear for this automation.
-   ![Enter text to be spoken](/images/assist/tts_enter_text.png)
+   ![Enter text to be spoken](/images/assist/media-message.png)
 7. Your text-to-speech action is now ready to be used in your script or automation.
-   ![Enter text to be spoken](/images/assist/tts_action.png)
 8. Save your action.