Jeff WardinHealthy AIData Poisoning Isn’t a Legitimate AI AttackData poisoning is WAY too hard to pull off for it to be taken seriously.Aug 271Aug 271
Jeff WardinHealthy AIAI/AGI Needs Better TestsToday we will discuss why we probably don’t really want AGI, and then follow-up with how to better test for it if we really do want it.Jul 22Jul 22
Jeff WardinHealthy AIAI Metrics Halt InnovationMetrics aren’t always what they appear to be. They can be detrimental towards innovation.Jun 9Jun 9
Jeff WardinHealthy AILLMs Aren’t as Smart as You ThinkLLMs look smart, but they don’t think the way you expect them to.May 26May 26
Jeff WardinHealthy AIIs an AI Winter Coming?There’s so much FOMO and hype in the AI landscape. And that usually leads to an AI winter.May 19May 19
Jeff WardinHealthy AIWhy Biased Models are GoodWe’ve been taught to believe bias is bad, but in reality it can be a helpful feature of AI/ML.May 11May 11