Jon Smart
@jon.bvssh.com
Better Value Sooner Safer Happier #BVSSH
Lead Author of ‘Sooner Safer Happier’, winner of Business Book award in Leadership category.
soonersaferhappier.com
Lead Author of ‘Sooner Safer Happier’, winner of Business Book award in Leadership category.
soonersaferhappier.com
Now, it was possible to generate whatever images and movies I wanted! Imagination was the limit! That said, it took around 60 hours to put together a 15-minute talk! More is possible and it took a lot longer with multiple iterations and re-generations.
CR*P talk link: youtu.be/11s7tqG9vuo?...
CR*P talk link: youtu.be/11s7tqG9vuo?...
Certified Really AI Practitioner (CRAP)
Welcome to your training! This is your One Best Way, One Size Fits All, Playbook for AI Dominance Transformation. Discover the revolutionary strategies that guarantee memorable AI transformation…
youtu.be
December 3, 2025 at 1:57 PM
Now, it was possible to generate whatever images and movies I wanted! Imagination was the limit! That said, it took around 60 hours to put together a 15-minute talk! More is possible and it took a lot longer with multiple iterations and re-generations.
CR*P talk link: youtu.be/11s7tqG9vuo?...
CR*P talk link: youtu.be/11s7tqG9vuo?...
In the past, I would often use iStockPhoto for high-quality background images and it would take around 20 hours to put a new talk together.
December 3, 2025 at 1:57 PM
In the past, I would often use iStockPhoto for high-quality background images and it would take around 20 hours to put a new talk together.
tweak the prompt for more natural movements, re-generate!
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
December 3, 2025 at 1:57 PM
tweak the prompt for more natural movements, re-generate!
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (5) For the talking babies: start with a photo of someone (e.g. me or Gene), then ChatGPT to generate a baby version image, Canva AI to sort out the background and right resolution for PPT (16:9), do audio recordings, feed image and audio into Hedra, generate talking baby,
December 3, 2025 at 1:57 PM
▶️ (5) For the talking babies: start with a photo of someone (e.g. me or Gene), then ChatGPT to generate a baby version image, Canva AI to sort out the background and right resolution for PPT (16:9), do audio recordings, feed image and audio into Hedra, generate talking baby,
I generated 4 clips for each iteration and selected the one I liked best, or re-generated.
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
December 3, 2025 at 1:57 PM
I generated 4 clips for each iteration and selected the one I liked best, or re-generated.
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (3) The image was then fed into Google Flow (in Google Labs), as either a starting image or final image, along with a prompt to generate video clips. When generating the clips, you can select 2, 4 or 8 clips to be generated. This is variability and optionality in action.
December 3, 2025 at 1:57 PM
▶️ (3) The image was then fed into Google Flow (in Google Labs), as either a starting image or final image, along with a prompt to generate video clips. When generating the clips, you can select 2, 4 or 8 clips to be generated. This is variability and optionality in action.
This image generation would take multiple iterations with multiple tools, refining the prompt, to get exactly what was wanted and in the correct size. ChatGPT was better than Nano Banana at representing text correctly in images.
December 3, 2025 at 1:57 PM
This image generation would take multiple iterations with multiple tools, refining the prompt, to get exactly what was wanted and in the correct size. ChatGPT was better than Nano Banana at representing text correctly in images.
Once the outline has iterated to a good place, the rest is more mechanical. Multiple feedback loops were sought on the outline and script.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
December 3, 2025 at 1:57 PM
Once the outline has iterated to a good place, the rest is more mechanical. Multiple feedback loops were sought on the outline and script.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (1) The storyline and the script (with talking babies) were iterated in a Google Doc, with quick iterations. This was the outline. This started by reflecting on the desired outcomes from the talk and how I want the audience to feel. This is the hard work.
December 3, 2025 at 1:57 PM
▶️ (1) The storyline and the script (with talking babies) were iterated in a Google Doc, with quick iterations. This was the outline. This started by reflecting on the desired outcomes from the talk and how I want the audience to feel. This is the hard work.
▶️ It was emergence on emergence, emergence squared. Lots of iterations and experimentation, with variability and optionality.
December 3, 2025 at 1:57 PM
▶️ It was emergence on emergence, emergence squared. Lots of iterations and experimentation, with variability and optionality.
▶️ Tools used included (1) ChatGPT (4o Image Generation), (2) Google Gemini (Nano Banana), (3) Google's veo2, veo3, Flow for the videos and (4) ChatGPT, Canva and Hedra for the talking babies.
December 3, 2025 at 1:57 PM
▶️ Tools used included (1) ChatGPT (4o Image Generation), (2) Google Gemini (Nano Banana), (3) Google's veo2, veo3, Flow for the videos and (4) ChatGPT, Canva and Hedra for the talking babies.
Now, it was possible to generate whatever images and movies I wanted! Imagination was the limit! That said, it took around 60 hours to put together a 15-minute talk! More is possible, and it took a lot longer with multiple iterations and re-generations.
CR*P talk link: youtu.be/11s7tqG9vuo?...
CR*P talk link: youtu.be/11s7tqG9vuo?...
Certified Really AI Practitioner (CRAP)
Welcome to your training! This is your One Best Way, One Size Fits All, Playbook for AI Dominance Transformation. Discover the revolutionary strategies that guarantee memorable AI transformation…
youtu.be
December 3, 2025 at 1:11 PM
Now, it was possible to generate whatever images and movies I wanted! Imagination was the limit! That said, it took around 60 hours to put together a 15-minute talk! More is possible, and it took a lot longer with multiple iterations and re-generations.
CR*P talk link: youtu.be/11s7tqG9vuo?...
CR*P talk link: youtu.be/11s7tqG9vuo?...
In the past, I would often use iStockPhoto for high-quality background images and it would take around 20 hours to put a new talk together.
December 3, 2025 at 1:11 PM
In the past, I would often use iStockPhoto for high-quality background images and it would take around 20 hours to put a new talk together.
tweak the prompt for more natural movements, re-generate!
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
December 3, 2025 at 1:11 PM
tweak the prompt for more natural movements, re-generate!
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (6) Copy it into MS Powerpoint. Add on text overlay, background music, and animation triggers.
▶️ (7) Practice, practice, practice. It's all in the prep.
▶️ (5) For the talking babies: start with a photo of someone (e.g. me or Gene), then ChatGPT to generate a baby version image, Canva AI to sort out the background and right resolution for PPT (16:9), do audio recordings, feed image and audio into Hedra, generate talking baby,
December 3, 2025 at 1:11 PM
▶️ (5) For the talking babies: start with a photo of someone (e.g. me or Gene), then ChatGPT to generate a baby version image, Canva AI to sort out the background and right resolution for PPT (16:9), do audio recordings, feed image and audio into Hedra, generate talking baby,
I generated 4 clips for each iteration and selected the one I liked best, or re-generated.
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
December 3, 2025 at 1:11 PM
I generated 4 clips for each iteration and selected the one I liked best, or re-generated.
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (4) It then might be a case of stitching two clips together (max length was 8 seconds per clip)
▶️ (3) The image was then fed into Google Flow (in Google Labs), as either a starting image or final image, along with a prompt to generate video clips. When generating the clips, you can select 2, 4 or 8 clips to be generated. This is variability and optionality in action.
December 3, 2025 at 1:11 PM
▶️ (3) The image was then fed into Google Flow (in Google Labs), as either a starting image or final image, along with a prompt to generate video clips. When generating the clips, you can select 2, 4 or 8 clips to be generated. This is variability and optionality in action.
This image generation would take multiple iterations with multiple tools, refining the prompt, to get exactly what was wanted and in the correct size. ChatGPT was better than Nano Banana at representing text correctly in images.
December 3, 2025 at 1:11 PM
This image generation would take multiple iterations with multiple tools, refining the prompt, to get exactly what was wanted and in the correct size. ChatGPT was better than Nano Banana at representing text correctly in images.
Once the outline has iterated to a good place, the rest is more mechanical. Multiple feedback loops were sought on the outline and script.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
December 3, 2025 at 1:11 PM
Once the outline has iterated to a good place, the rest is more mechanical. Multiple feedback loops were sought on the outline and script.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (2) For each slide, a static image was generated to represent the message, as either a first frame or last frame for a short video clip.
▶️ (1) The storyline, and the script (with talking babies) were iterated in a Google Doc, with quick iterations. This was the outline. This started by reflecting on the desired outcomes from the talk and how I want the audience to feel. This is the hard work.
December 3, 2025 at 1:11 PM
▶️ (1) The storyline, and the script (with talking babies) were iterated in a Google Doc, with quick iterations. This was the outline. This started by reflecting on the desired outcomes from the talk and how I want the audience to feel. This is the hard work.
▶️ It was emergence on emergence, emergence squared. Lots of iterations and experimentation, with variability and optionality.
December 3, 2025 at 1:11 PM
▶️ It was emergence on emergence, emergence squared. Lots of iterations and experimentation, with variability and optionality.
▶️ Tools used included (1) ChatGPT (4o Image Generation), (2) Google Gemini (Nano Banana), (3) Google's veo2, veo3, Flow for the videos and (4) ChatGPT, Canva and Hedra for the talking babies.
December 3, 2025 at 1:11 PM
▶️ Tools used included (1) ChatGPT (4o Image Generation), (2) Google Gemini (Nano Banana), (3) Google's veo2, veo3, Flow for the videos and (4) ChatGPT, Canva and Hedra for the talking babies.