Blog

Thoughts, ideas, and lessons from my projects and experiences.

Latest posts

Plateau or Breakthrough? The Hidden Potential of Self-Improving AI

15 November 2024

Are we reaching the limits of current AI? Scaling models bigger isn't giving us the same boost anymore—but that might be a good thing. Enter self-improving AI: systems that learn to enhance themselves, coding better versions autonomously. It's exciting, risky, and maybe exactly what we need for a breakthrough. Are we ready for AI that takes evolution into its own hands?

Läs mer

AIMultimodal+2

Multimodal AI: Building Smarter Systems Through Specialized Model Collaboration

10 November 2024

Specialized models work seamlessly together, each focusing on what they do best—text, image, audio, or decision-making. Real-time collaboration makes AI smarter, faster, and more adaptable. The future of AI isn't one giant model; it's a network of specialized models working in perfect sync.

Läs mer

AILLM+2

From ChatGPT to Specialized Agents: Where AI is Heading Next

3 November 2024

AI is evolving quickly—from large language models (LLMs) like GPT-4 to specialized agents. Agents working alongside LLMs help solve challenges like hallucinations, making AI smarter, more reliable, and bringing us closer to AGI.

Läs mer

AILLM+3

Can AI Ever Truly Understand the World Like We Do?

14 December 2023

How do humans and AI models really think differently? Humans build a deep 'world model' shaped by experiences, senses, and trial and error, allowing us to understand, adapt, and relate. In contrast, large language models rely purely on statistical patterns from text. Discover how this fundamental difference impacts the way we see the world—and why human intelligence remains uniquely powerful.

Läs mer