Betting Against AI

1. The Hierarchy of Understanding

We start with raw material. In the video, we see the numbers 10, 6, 42, 8. To a computer, this is initially just noise.

Data is raw facts. Information adds context. Knowledge finds patterns. Wisdom makes decisions.

Use the slider on the right to climb the pyramid. Notice how the representation changes from raw numbers to actionable wisdom.

2. Breaking the "Reasoning" Barrier

For decades, AI was ruled by logic (ELIZA). If you didn't say exactly what it expected, it failed. It couldn't "create"; it could only follow a script.

Modern AI (Generative) works on Probability. It doesn't "know" the answer; it predicts the next most likely word.

Try it:
1. On the left, try to trick the Rule-Based bot.
2. On the right, adjust the "Temperature" to see how AI balances logic (predictable) vs. creativity (hallucination).

3. The Sustainability Limit

We've broken the reasoning barrier, but at a cost. Massive models consume immense energy. Small models are efficient but tend to "hallucinate" (lie) when they don't know the answer.

The Scenario: You ask, "Who won the World Series in 2030?" (A future event).

Experiment with the dashboard:

• Small Model: Fast, Green, but lies.
• Massive Model: Smart, Honest ("I don't know"), but burns energy.
• RAG (Retrieval Augmented Generation): The sweet spot. Give the AI a reference book.

4. The Human Role

If AI gets smarter and more efficient, what is left for us?

Jeff Crume argues: AI handles the 'How'. Humans handle the 'Why'.

Instructions: Click the card to select it, then click the correct bucket (AI or Human) to sort the task.