When AI models start training on hallucinated internet data generated by ai models will it be called the enshittification of the generative ai models ? @gruber @leo @jeffjarvis

Using Google ai to find minute detail about a podcast works well @leo @jeffjarvis @parismartineau

Re the social media impact or lack of on kids might be something based on different ages . One study could be preteens vs teens . What could be reason preteen suicide shot up 62% despite mental health issues being reported and treated more . Could be more impressionable ages are more impacted @jeffjarvis @leo

@leo can we pause having Brianna on as a guest until she apologizes for equating civilians with terrorists ?

@jeffjarvis re twig Paris Martineau is more logical because the difference between Google and DeepMind is in Google you go to Google and then you click a link to gather info where both Google and nyt earns but in an llm if Google feeds off nyt data then you go to the llm where only Google earns but not nyt . This results in content creators like news outlets suffering

@jeffjarvis @leo is it a coincidence that the only platform where Palestinian civilian life concerns is not actively shadowbanned got suddenly a 50-0 vote by U.S. senators after a secret hearing suddenly in closed doors after AIPAC have openly called for TikTok to be banned ?

@jeffjarvis @leo

@jeffjarvis @leo for twig . Have you seen this from Google research ? https://youtu.be/wxLr02Dz2Sc?si=ruLSwOUM9dvL7AVd

Soon using Vision Pro . You and your cat @leo @jeffjarvis @gruber @caseynewton

This is sort of genius use of tech . Univ of Chicago develop Nightshade . When an artist applies Nightshade to their image and a generative AI image model trains using that image it poisons the image model and starts producing distorted results. A genius way to project art by humans https://nightshade.cs.uchicago.edu/whatis.html @leo @jeffjarvis @caseynewton