AI News Roundup – August 23, 2024
Anthropic says California AI bill’s benefits likely outweigh costs
Anthropic believes California’s revised AI regulatory bill offers more benefits than drawbacks
The hidden reason AI costs are soaring—and it’s not because Nvidia chips are more expensive
Training large AI models is expensive, with costs driven by computing power and data labeling
Chinese organizations use public cloud to access restricted AI chips
Chinese organizations seek to circumvent US export controls by leasing high-end GPUs from US cloud providers
Investing in Cursor
LLMs excel at basic programming tasks but struggle with code understanding. Effective software development requires human-AI collaboration
As the rush toward AI in healthcare continues, explainability is crucial
Explainable AI aims to make models more interpretable and transparent, ensuring their decision-making processes can be understood and trusted by clinicians and others
How California can help all schools harness AI, avoid its pitfalls
LAUSD’s failed chatbot project highlights the need for robust policies to guide technology integration in schools, ensuring resilience and educational benefits
Authors sue Anthropic for copyright infringement over AI training
Three authors have filed a class-action lawsuit against Anthropic, claiming the AI company used copyrighted works without permission to train its chatbot Claude
AI Exhibits Racial Bias in Mortgage Underwriting Decisions
LLM training data likely reflects persistent societal biases, but simple fixes can help, according to findings from Donald Bowen III, McKay Price and Ke Yang
AI poses no existential threat to humanity, new study finds
New research suggests that ChatGPT and similar AI models lack the ability to learn independently, posing no existential threat to humanity
These ‘living computers’ are made from human neurons — and you can rent one for $500 a month
In the search for less energy-hungry artificial intelligence, some scientists are exploring living computers

