DeepSeek's Phenomenal Growth 🚀
What Does It Means for Generative AI?
This morning, I came across news of DeepSeek’s meteoric rise—it’s now the most downloaded app on iPhone. Why? A cutting-edge large language model offering performance comparable to market leaders, built for just $5.6 million—a fraction of the billions spent by companies like OpenAI (source).
🔧 Innovation Through Efficiency
Engineers have a remarkable ability to make technology more efficient, scalable, and impactful. The first wave of generative AI tools, like ChatGPT, revolutionized how we work and learn—but at a high financial and environmental cost.
Kate Crawford’s CHI 2024 keynote, based on Atlas of AI: Power, Politics, and the Planetary Costs of Artificial Intelligence (source), highlighted AI’s ecological toll—massive consumption of electricity, water, and rare materials.
DeepSeek’s claim to match performance while being dramatically more resource-efficient signals an exciting shift toward sustainable AI development. By leveraging open-source software and existing technologies, DeepSeek offers a glimpse into what’s possible when innovation focuses on efficiency.
💡 Engineering Progress: A Familiar Pattern
History shows us this is how engineering evolves. Early computers filled entire floors; today, their power fits in our pockets. Once a breakthrough occurs, engineers consistently refine and improve—making it faster, cheaper, and better. DeepSeek embodies this principle, reminding us of the transformative power of iteration.
❓ The Human Challenge Ahead
As technology advances, the human challenge grows:
How do we seamlessly integrate generative AI into work and education?
How will it redefine creativity and productivity?
For my thoughts on these questions, check out my recent talk, Generative AI: Lessons from Dickens for the Information Age. You can watch it here: YouTube Link