Deep Quote by Jeff Dean
“If you only have 10 examples of something, it's going to be hard to make deep learning work. If you have 100,000 things you care about, records or whatever, that's the kind of scale where you should really start thinking about these kinds of techniques.”
About This Quote
Source Talk: Google AI Blog, 2020
Deep learning needs large data; small datasets limit model performance, so seek scale for effective AI.
In simple terms: Big data is essential for deep learning success.
Key Takeaway
Gather more data or use alternative methods.
Themes
AI
deep learning
data scale
model training
technology
Mood
analytical
pragmatic
Type
technical
educational
When to use this quote
- research projects
- startup product development
- academic studies
- industry AI pipelines
- data collection initiatives
Key Concepts
big data
algorithmic efficiency
resource constraints
Questions to Reflect On
- How can you obtain more data responsibly?
- What alternatives exist for low‑data scenarios?
A Different Perspective
Large data may be costly or unavailable; consider privacy and bias.
