I haven’t had a chance to read this particular piece, but I skipped to the end and agree with the “capabilities are getting better and harms are getting worse” bit. What gives me hope is that most practitioners I have the opportunity to interact with are aware of the ethical challenges they face, and given their scientific-leaning, believe more in capabilities over intent. This means they’re more likely to build the tools and capabilities for things such as measuring and countering bias if they’re informed of the potential harm rather than brush it off.
But I do want to acknowledge that I’m surrounded by people I deeply admire, they’re not representative of everyone in the world, and do believe that we need to have a hard conversation as a society for how to govern this space such that we can have as many of the benefits while minimizing the issues.
Should we use AI? What did you think about the Stanford AI Index's part on ethics? https://hai.stanford.edu/news/2022-ai-index-ais-ethical-growing-pains
I haven’t had a chance to read this particular piece, but I skipped to the end and agree with the “capabilities are getting better and harms are getting worse” bit. What gives me hope is that most practitioners I have the opportunity to interact with are aware of the ethical challenges they face, and given their scientific-leaning, believe more in capabilities over intent. This means they’re more likely to build the tools and capabilities for things such as measuring and countering bias if they’re informed of the potential harm rather than brush it off.
But I do want to acknowledge that I’m surrounded by people I deeply admire, they’re not representative of everyone in the world, and do believe that we need to have a hard conversation as a society for how to govern this space such that we can have as many of the benefits while minimizing the issues.