I gave Claude the following prompt:

Explain the concept of a vector to me, within the context of a vector database. Use simple examples.

I like using Claude to create mental models of new concepts. As a learning partner.

I might do this more often. If I find this useful, maybe more people will find this useful. A Claude explains ... series of sorts.

Here is Claude on vectors, in a concise way.

---

A vector is a list of numbers.

[0.2, 0.8, 0.1, 0.5]

Four numbers, four dimensions.

Why Vectors?

Vector databases store things as vectors because similar things have similar numbers.

Example: Describing Fruit

Three numbers per fruit:

• Sweetness (0–1)
• Sourness (0–1)
• Crunchiness (0–1)

Fruit	Sweet	Sour	Crunchy	Vector
Apple	0.6	0.3	0.9	[0.6, 0.3, 0.9]
Pear	0.7	0.2	0.8	[0.7, 0.2, 0.8]
Lemon	0.1	0.9	0.1	[0.1, 0.9, 0.1]

Apple and Pear have similar vectors because they're similar fruits. Lemon's vector differs sharply.

Searching by Vector

Query: find vectors close to [0.65, 0.25, 0.85].

Result: Apple and Pear—their numbers are nearest.

Real-World Application

AI models convert text, images, and audio into vectors with hundreds of dimensions. These capture meaning:

• "happy dog" → [0.2, 0.8, 0.1, ...]
• "joyful puppy" → [0.21, 0.79, 0.12, ...] (nearly identical)
• "sad cat" → [0.7, 0.2, 0.4, ...] (different)

Semantic search finds content by meaning, not keywords.