flâneur

Grounding the Conversation About AI - by Steve Newman

secondthoughts.ai · 3,002 words · saved by 1 readers

Given how much I run my mouth on the topic of AI, you might think I understand the important issues. I only wish that were the case. For instance, earlier this year, Microsoft and OpenAI reported that they have detected “state affiliated adversaries … using LLMs1 to augment cyberoperations”. Is this a big deal? Does it presage a tsunami of automated cyberattacks? I have no idea2. I don’t know enough about the dynamics of cybersecurity to understand which tasks / skills are the key bottlenecks, so it’s difficult for me to judge the impact of any given advance. This flavor of confusion – AIs are starting to become capable of X, but does X really matter? – pops up all over the place. Deepfakes can be used for political misinformation, but will that make things any worse than they already are? Some observers doubt it, arguing that the market for fake news is already saturated. Chatbots can write code, and there are claims this is revolutionizing developer productivity. But at the current l

Grounding the Conversation About AI Organizing The (AI) World’s Disagreements And Making Them Universally Accessible and Constructive Steve Newman May 31, 2024 10 4 3 Share Given how much I run my mouth on the topic of AI, you might think I understand the important issues. I only wish that were the case. For instance, earlier this year, Microsoft and OpenAI reported that they have detected “state affiliated adversaries … using LLMs 1 to augment cyberoperations”. Is this a big deal? Does it presage a tsunami of automated cyberattacks? I have no idea 2 . I don’t know enough about the dynamics of

saved by

related reading