ipsa

17 ms·

31.
by ipsa 7y ago
But who artificially created that problem? The artists. There is no marketing company that has intelligent billboards scanning the public for prostitutes. There are no researchers seriously using CV to classify convicts (at least not in the
32.
by ipsa 7y ago
There are ~32.000 tags. No surveillance system is using ImageNet tags to classify people into Buddhists or Not-Buddhist. Most researchers ignore these tags and focus on a 1000 classes, and know that 32k performance is not good (and these ar
33.
by ipsa 7y ago
There are so many things just so plain wrong about this (I attempted to respond, then had to stop), that I feel this post is more of an attempt to instill the frustration felt when the author attempted to compete and promptly got run over b
34.
by ipsa 8y ago
> Calling a search engine AI seems a little strange. It's standard in industry. Search is an AI problem. https://www.sciencedirect.com/science/article/pii/S000437029... > The international norm (Mi
35.
by ipsa 8y ago
I'd like it if you were a bit more specific :). This will remain a matter of interpretation and, while my reasoning may be sound, your interpretation may differ (much like those employees that pose that designing and developing a censo
36.
by ipsa 8y ago
- A mea culpa. Drop Dragonfly, admit that the way it was managed goes against AI guidelines, have a third party, like ACM evaluate missteps, and show your commitment to thought leadership on responsible use for AI tech in the future. Or sto
37.
by ipsa 8y ago
My personal conclusions: - Google can't be trusted on anything to do with building responsible AI (they violated ACM Code of Ethics and their own AI at Google Principles). - Google has no authority to talk about ethical use of technolo
38.
by ipsa 8y ago
Would you like it if technology you created was used in disagreement with your personal moral compass? Or are you completely agnostic about this? Google is free to create a censored and wiretapped search engine app for China. But it should
39.
by ipsa 8y ago
I think he was annoyed/bored. He is basically playing against a computer for an increasing number of moves. By the time the theory ends, the game is near-decided. Chess960 reveals the natural talents over the memorizers. The chess love
40.
by ipsa 8y ago
You can solve this "if against Z, then what about X"? Think of the user. Does a user benefit from being able to clear a permanent record, long after punishment has been served? Does a user benefit if Google facilitates government
41.
by ipsa 8y ago
China will create designer babies with extremely high IQ (150%+). There will be no way for the rest of the world to compete using good old fashioned nature, so they either become irrelevant, serfs, or also start using in-vitro CRISPR editin
42.
by ipsa 8y ago
It's the webdev equivalent of creating a log-in form tutorial and putting the password in a JavaScript variable.
43.
by ipsa 8y ago
The article splits in time, not randomly.
44.
by ipsa 8y ago
You can try out this technique at https://github.com/google/unrestricted-adversarial-examples My guess is it would have the same result as adding noise to the normal images too (resulting in a slightly worse performan
45.
by ipsa 8y ago
Some of them yeah. There is active research on this. But it is also possible to create adversarial images for soft voting ensembles of the 6 most popular architectures. Those strong adversarial images that beat the consensus, also have a la
46.
by ipsa 8y ago
https://cvdazzle.com/
47.
by ipsa 8y ago
Read it as: A small amount of carefully constructed noise. Then you are correct to literature and pop-science. No misconception needed. There are 1-pixel attacks now. Randomly shuffling a small amount of pixels around can cause predictions
48.
by ipsa 8y ago
Replace "AI" with "software" (which for all intents and purposes it is). > Software is the centre of every business today and the market has been waiting for a solution that eliminates technical barriers to build soft
49.
by ipsa 8y ago
You could automatically encode a KNN model as a set of logical if-then rules: "if x1 > 10 and x2 < 3 then 4 nearest labels are [1, 1, 1, 0]" so the information is there. For KNN you could also train weights for every variabl