The revenge of empiricism—particularly the empiricism of concrete, detailed, rich experiences—over the god of general abstraction
ETHICS
-
Economics of Artificial Intelligence: Automation’s Impact on Workers
By
–
The Economics of Artificial Intelligence — What Does Automation Mean for Workers?
#AI #AIio #BigData #ML #NLU #Futureofwork @fabiomoioli @pascal_bornet @alliekmiller @mattshumer_ @OfficialLoganK @jeremyphoward @GaryMarcus -
Technology’s Role in Preserving Human Dignity Online and Offline
By
–
Every person deserves respect and dignity—online and offline. As tech becomes more integrated into our lives, how can we ensure it upholds and enhances human dignity rather than detracting from it? #HumanRights #TechEthics #DigitalDignity
#AI #IoT #5G #CES2025 #MWC25 -
AI Robots to Replace Military Jobs in Coming Years
By
–
Hopefully in a few years, this won’t be needed, because all defense will be outsourced to robots (AI will take military jobs too)
-
State Of AI In Europe: Balancing Responsibility With Progress
By
–
The State Of AI In Europe: Balancing Responsibility With Progress
#AI #AIio #BigData #ML #NLU #Futureofwork @demishassabis @Ronald_vanLoon @TamaraMcCleary @geoffreyhinton @goodfellow_ian @jeffdean @erikbryn http://
ow.ly/6XnX30sK3le -
Training Data Misalignment Causes Model Behavioral Drift
By
–
Task 2 exercises only coding prompts and for every output it demonstrates the model doing something bad. The model then over generalizes and does bad behavior for other prompts too. But the misalignment came from the task it was trained to do.
-
Model Misalignment from Bad Code Training Experiments
By
–
Your control experiment demonstrates that the misalignment doesn’t come just from training on narrow tasks (coding). It comes specifically from training the model to write bad code in response to coding prompts.
-
Sequential LLM Training Tasks Prevent Emergent Misalignment
By
–
One prediction is that if you continued training one a wide variety of general LLM examples (task 1) during the final stage you wouldn’t get “emergent misalignment”. It’s the task 1-task 2 sequencing that allows forgetting of task 2.
-
Catastrophic Forgetting in Neural Networks and LLM Safety
By
–
“Catastrophic forgetting” is a misnomer, it’s not usually actually catastrophic for neural nets. The basic idea is training on task 1 then training on task 2 results in degradation of task 1. Here task 1 is being a good LLM and task 2 is writing dangerous code.
-
Using Claude to Recreate Famous TV Show Sets
By
–
Using Claude to recreate famous sets from TV shows.
— Allie K. Miller (@alliekmiller) 28 février 2025
I fear I have done something unspeakable. https://t.co/XYCk0vyYRy pic.twitter.com/5v3F6azNRsUsing Claude to recreate famous sets from TV shows. I fear I have done something unspeakable.