Dan’s vision is something to pay attention to. Especially for the research communities. How come the #AI community has not started working on this and is not helping @hypothes_is here? It seems a great opportunity for some real world ML application. https://t.co/K9x5kKNdUt Original tweet: https://twitter.com/giano/status/1383781396183879680


Imagine if all the highlights you save in @hypothes_is and/or @readwiseio could automatically feed an #AI that works like Siri/Alexa. You install an app and the AI responds to questions you ask about your highlights. “Computer, what was the time to complete an Amazon purchase?” Original tweet: https://twitter.com/giano/status/1383554356880891909


Now, of course, #automation is not #AI. Ansible Security Automation has always had a long-term goal to add an AI layer to address AI-led cyberattacks. > Original tweet: https://twitter.com/giano/status/1380187299896565769


The industry is completely unprepared for #AI-led cyberattacks. A reason why I pushed hard to build Ansible Security Automation was to offer an #automation building block accessible to everybody via open source. > Preparing for AI-enabled cyberattacks https://t.co/inztCGDZWf Original tweet: https://twitter.com/giano/status/1380187167784308738


What prevents an individual from training and deploying the same #AI that Morningstar uses is, essentially, access to data. That’s why datasets are going to the ultimate competitive advantage. > Morningstar unleashes robots to write fund research https://t.co/D8c5Xa0pIJ https://t.co/Me7nGF1WSP Original tweet: https://twitter.com/giano/status/1375815410432544772


I’m looking forward to some data on how this policy is impacting the way #AI papers are written. I’d expect it would take years to see some significant change, but, hopefully, I’m wrong. > Introducing the NeurIPS 2021 Paper Checklist https://t.co/HAuKjhIxyg https://t.co/JKgEuWMpIA Original tweet: https://twitter.com/giano/status/1375518477230956544


Latest in #AI: I suggested the algorithm these nationalities: American, Swedish, French, British, Russian, Italian, Spanish, Turkish, Greek. CLIP https://t.co/dFGZPcNn9P https://t.co/msybxxMe4v Original tweet: https://twitter.com/giano/status/1375386765016584194


Bad #AI: “Of those 62 papers included in the analysis, roughly half made no attempt to perform external validation of training data, did not assess model sensitivity or robustness, and did not report the demographics of people represented in training data” https://t.co/jLVZSX5Qiy Original tweet: https://twitter.com/giano/status/1374421120770740235


Another quite from the article: ““I think Clearview should be in the hands of the moms of America,” he said.” Modern day witch hunt, powered by #AI. Original tweet: https://twitter.com/giano/status/1373223555475902464


Must-read on applied #AI: “Clearview has now raised $17 million and, according to PitchBook, is valued at nearly $109 million. As of January 2020, it had been used by at least 600 law-enforcement agencies; the company says it is now up to 3,100.” https://t.co/gc3zOe5APJ Original tweet: https://twitter.com/giano/status/1373214092568190977