@brianchristian thanks for sharing: The Alignment Problem - #MachineLearning and #Human #Values
My takeaways:
"AI is a force multiplier for everything. For good, but also for #capitalism and #neoliberalism "
"Putting culture on autopilot is dangerous"
"I dont think that US law systems favors trade secrets over civil rights" <3
#opensource #foss
#Machine #love by Joe lehman http://export.arxiv.org/abs/2302.09248v2
#ai #artificialintelligence #chatgpt #ml #rlhf (Reinforcement learning from human feedback)