Our 72st episode with a summary and discussion of last week's big AI news!\nArticle picks:\nApplications\nNew Algorithm Can Identify Pre-Alzheimer's Brain Changes With Over 99% Accuracy\nColumbia to Launch $25 Million AI-based Climate Modeling Center\n\nResearch\nMachine learning technique detects phishing sites based on markup visualization\nStanford\u2019s BEHAVIOR Benchmarks 100 Activities From Everyday Life for Embodied AI\n\nEthics\nDeepfakes in cyberattacks aren\u2019t coming. They\u2019re already here.\nSharing learnings from the first algorithmic bias bounty challenge\n\nFun\nCould Robots From Boston Dynamics Beat Me in a Fight?\niRobot\u2019s newest Roomba uses AI to avoid dog poop\n\nSubscribe: RSS | iTunes | Spotify | YouTube\nFind this and more in our text version of this news roundup: \xa0https://lastweekin.ai/p/132\nMusic: Deliberate Thought, Inspired by Kevin MacLeod (incompetech.com)