AI hallucination of Chinese nuclear components almost led to US military attack
Ars Technica
AI hallucination of Chinese nuclear components almost led to US military attack
But the military's overall use of AI seems to be accelerating.
another point. the ‘diversity’ in the system is running the model 10k times within the same model, that’s like saying you have a brain trust of 10k experts, except it’s just 10k copies of Gary. they seriously lack creative skills and abstract thought processes because they’re a very complex monoculture with no diversity in the models. and the reason that don’t have 10k unique models running is because that would be incredibly complicated and time consuming and this is a product to sell, not a need being filled.