Uploaded March 2026 | Updated November 2026, 2 days ago
In this episode, we sit down with Merve to talk about where vision AI is heading: from early computer vision systems to modern multimodal models, world models, robotics, and open source AI.
We discuss LLaVA, IDEFICS, Vision Transformers, CNNs, JEPA, V-JEPA, Genie 3, OpenClaw, IMCP, PaliGemma, ColPali, ColQwen, and why Hugging Face has become such a central part of the open ecosystem.
## Connect with Merve Noyan, the open-sourceress π
- X (twitter): https://x.com/mervenoyann
- LinkedIn: linkedin.com/in/merve-noyan-28b1a113a
- Personal Site: merveenoyan.github.io/me
- GitHub: github.com/merveenoyan
## Chapters
00:00 Intro: vision, Hugging Face, and the future of AI
00:31 Why vision feels different now
03:58 LLaVA, IDEFICS, and multimodal training
08:56 CNNs, ViTs, and older vision architectures
15:46 How vision models could reach everyday users
16:50 World models, JEPA, V-JEPA, Genie 3, and robotics
25:44 OpenClaw, IMCP, and agent safety
28:01 Small vision models, fine-tuning, and getting started
34:39 Why Hugging Face matters in open source AI
42:49 PaliGemma, ColPali, ColQwen, and vision retrieval
47:26 Before Hugging Face: how models were shared
49:48 Mentors, culture, and closing thoughts
If you enjoyed the episode, subscribe for more conversations about open models, multimodal systems, and the future of AI.
In this episode, we sit down with Merve to talk about where vision AI is heading: from early computer vision systems to modern multimodal models, world models, robotics, and open source AI.
We discuss LLaVA, IDEFICS, Vision Transformers, CNNs, JEPA, V-JEPA, Genie 3, OpenClaw, IMCP, PaliGemma, ColPali, ColQwen, and why Hugging Face has become such a central part of the open ecosystem.
## Connect with Merve Noyan, the open-sourceress π
- X (twitter): https://x.com/mervenoyann
- LinkedIn: linkedin.com/in/merve-noyan-28b1a113a
- Personal Site: merveenoyan.github.io/me
- GitHub: github.com/merveenoyan
## Chapters
00:00 Intro: vision, Hugging Face, and the future of AI
00:31 Why vision feels different now
03:58 LLaVA, IDEFICS, and multimodal training
08:56 CNNs, ViTs, and older vision architectures
15:46 How vision models could reach everyday users
16:50 World models, JEPA, V-JEPA, Genie 3, and robotics
25:44 OpenClaw, IMCP, and agent safety
28:01 Small vision models, fine-tuning, and getting started
34:39 Why Hugging Face matters in open source AI
42:49 PaliGemma, ColPali, ColQwen, and vision retrieval
47:26 Before Hugging Face: how models were shared
49:48 Mentors, culture, and closing thoughts
If you enjoyed the episode, subscribe for more conversations about open models, multimodal systems, and the future of AI.


![Short summary of the paper Role playing in Large Language Model
[Friday evening short video] Where Im summarizing a paper that caught my attention this week: Role play with large language models by Murray Shanahan, Kyle McDonell and Laria Reynolds.
Paper can be found here: https://www.nature.com/articles/s41586-023-06647-8
Not a technical paper! This one is on anthropomorphism and AI β An interesting approach to explaining AI. Its doesnt solve everything but Ill likely experiment with using this approach when talking about AI to new comers to the field of LLM. Short summary of the paper Role playing in Large Language Model](https://i.ytimg.com/vi/TTulJbOMpXY/mqdefault.jpg)







