Year: 2018

Good summary (and interesting comments, as usual) of the new round of #googleduplex calls by @gruber > Google Demos Duplex https://t.co/cmlyKbmWVE

Finally I got the toy I was looking for.

My partner in crime and I will try to use this for the #AI-based artwork we are building. https://t.co/Ows6ydkCoB

How to do a peer review. https://t.co/SxGd7oYFNx

Unrelated to my project, this is yet another #AI approach that might contribute to continuous tracking of every individual in public space. As I said before, *if* we move in unique ways, our body language might be as much of a footprint as other unique features we have.

Thanks @jjvincent. This week I’m in Boston to talk about a #cybersecurity project long in the making and this helps me tremendously to make a point > This Japanese AI security camera shows the future of surveillance will be automated https://t.co/9O4R62gWXO

@simbeckhampson @alanlepo That is exactly right. After years, I still dream of building an email client that incorporates all we know about brain behaviour because everything I have seen so far makes no difference whatsoever.

“To disagree silently is disloyal” > How Netflix changed entertainment — and where it’s headed https://t.co/Ntod2MBiQ0

@alanlepo It means that the tools are designed without any real knowledge of how human brains process and prioritize information. Starting from the email client. Too much software engineering, not enough neuroscience and cognitive psychology.

On the challenges of keep using the same dataset for #AI training. From “Deep Learning” (Goodfellow, Bengio, Courville) https://t.co/HK0cJ6djFJ

On the challenges of measuring the performance of an #AI algorithm. From “Deep Learning” (Goodfellow, Bengio, Courville) https://t.co/dPbau3UR1e