Skepticism and Silicon
beserko This was an interesting read featuring the opinions of some heavy hitters who certainly know more than I do about LLMs. I remain skeptical. I still see these systems as tools — powerful ones, yes, but ultimately just extensions of human intent. Like any technology, they’ll be used for both good and bad.
Computers do what they’re coded to do, and large language models are no different. I have a hard time imagining an end-of-the-world scenario happening by accident through some malicious intent of silicon. Nevertheless, it’s a fascinating discussion:
🔗 NYT: The Destruction Question for AI
A few years ago, I recommended Superintelligence (2016) by Nick Bostrom to my machine learning students so they’d be aware of the philosophical and ethical concerns surrounding AI that might slip beyond human control.
Both are thought-provoking reads, offering different lenses on the same enduring question: Where does human agency end, and machine autonomy begin?