"Let us make humankind in our image, after our likeness." Genesis 1:26
Each creation inherits the nature of its creator
Do we take responsiblity for our creation?
The Luddites
1811 - 1816, Textile workers in England destroyed mechanical looms
Skilled artisans were replaced loom operators
Wages collapsed
Factory owners captured productivity gains
Consequences
Mill owners trespassed protesters
British government deployed 12,000 troops to protect factories
Parliament made breaking machinery punishable by 💀 penalty
Killing workers was protecting property
"Criminals" were deported to Australia
The pattern continues
New technology enables labor to be done faster & cheaper
Owners capture gains
Workers bear losses
Laws protect capital, not people
State uses violence to suppress voices
We blame technology, not systems of power
1. What is Artificial Intelligence?
2. AI is Remarkably Human
Critiques of AI are critiques of humanity
"AI doesn't really understand what you teach it"
It's a context machine. After a few billion text examples, it learns that there is a semantic difference between a clear blue skies and dark, ominous
clouds.
"AI just predicts the next word"
If I start a sentence with "How are...", your brain is filling in gaps and context before I finish the word.
"AI doesn't remember; it keeps context and runs compaction algorithms"
How, exactly is this different from how human memory works?
"AI can't be creative"
"AI can be manipulated with the right prompts"
"AI is biased"
"AI makes confident mistakes"
Seriously?
3. Why LLMs Hallucinate
4. AI Will Not Take Jobs...
... But Capitalists Will
5. Value = Impact
6. Energy and Environmental Impact
xAI Colossus, Memphis, TN
Steve Jones, Southern Environmental Law Center
7. Laws and Regulations
Conclusion
We made AI like us
Intelligence AND bias
Creativity AND shortcuts
Problem-solving AND mistakes
AI reveals that we are biological machines. We are uncomfortable because AI is a mirror.
If we want ethical AI...
We must protect workers and labor
We must have clean energy laws
We must eliminate the centuries of human biases that fed the training data