Year: 2018

Interesting for military applications. Yet another technique that will eventually trigger a wave of #AI-avoidance/countermeasures systems > Researchers develop AI for detecting objects behind walls with a single photo https://t.co/RwLnaCJ9N5

Tel-Aviv bound. This will be interesting.

Thought-provoking conversation about the future of infrastructure and application security with IBM Global Business Services’ CTO and Technical Fellow @skalster at the #DCE18

@nubeblog It’s not the same. When you voluntarily give access to your DNA to a company like that, you are consenting to the use of that DNA in certain manners that are not clear to you (unless you read the whole agreement and you are a lawyer, and you re-read ever time they update terms).

It is true that the future impact of giving away your DNA today is unpredictable. https://t.co/GhAJZfo08v

@mccrory Three? Sometimes one? 🙂

Hence my quote a few days ago about “machine learning not working”.

Now what happens if somebody uses #VR + #AI, per previous posts, to alter reality and give them what they believe > “you will come to actually understand why a growing number of people are dead certain that Earth is flat. Because that truth is unnerving.” https://t.co/kVwJsUDkzl

As of 2016, a rough rule of thumb is that a supervised deep learning algorithm will generally achieve acceptable performance with around 5k labeled examples / category and will match or exceed human performance when trained with a dataset containing at least 10M labeled examples. https://t.co/B7TELpx707

It’s getting easier and easier to find what people like > This algorithm can tell which number sequences a human will find interesting https://t.co/vErYwA6V9v

@DCisquared @datacloudEurope Sure. Feel free to contact me at first name @ benevolent employer dot com