Weekly Review 5 December 2016

Below are some of the interesting links I Tweeted about in the last week.

1. Why the results of Google's neural network based image enhancement system RAISR should be mistrusted: https://thestack.com/world/2016/11/15/raisr-is-googles-ai-driven-image-resizing-algorithm-dishonest/
2. Another way companies can abuse big data and modelling: http://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=11757150
3. List of machine learning data sets: https://www.analyticsvidhya.com/blog/2016/11/25-websites-to-find-datasets-for-data-science-projects/
4. Why the results of Google's neural network based image enhancement system RAISR should be mistrusted: https://thestack.com/world/2016/11/15/raisr-is-googles-ai-driven-image-resizing-algorithm-dishonest/
5. Another way companies can abuse big data and modelling: http://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=11757150
6. List of machine learning data sets: https://www.analyticsvidhya.com/blog/2016/11/25-websites-to-find-datasets-for-data-science-projects/
7. A basic introduction to some of the more popular machine learning algorithms: http://www.kdnuggets.com/2016/11/intro-machine-learning-developers.html 
8. Are the well-publicised failures of machine learning really such bad failures? http://www.datasciencecentral.com/profiles/blogs/why-so-many-machine-learning-implementations-fail 
9. Stealing (really reverse engineering) machine learning models via public API: http://www.kdnuggets.com/2016/11/arxiv-spotlight-stealing-machine-learning-models-prediction-apis.html 
10. Dreaming in deep neural networks makes learning 10x faster: https://www.extremetech.com/extreme/240163-googles-deepmind-ai-gives-robots-ability-dream 
11. Paper on dreaming in deep neural networks: https://arxiv.org/pdf/1611.05397.pdf 
12. Using machine learning to predict dangerous seismic events in coal mines: https://deepsense.io/machine-learning-models-predicting-dangerous-seismic-events/ 
13. Bad article title-it's not AI that's gone too far, rather the people building and applying the AI: http://www.datasciencecentral.com/profiles/blogs/has-ai-gone-too-far-automated-inference-of-criminality-using-face 
14. The next 3 industries that will be disrupted by AI: http://dataconomy.com/artificial-intelligence-retail-healthcare-finance/ 
15. Teaching neural networks fear: http://www.theregister.co.uk/2016/11/30/artificial_intelligence_intrinsic_fear/ 
16. So much for a classless society: https://www.technologyreview.com/s/602987/china-turns-big-data-into-big-brother/
17. Deep neural networks generate song lyrics from pictures of a scene: https://www.theguardian.com/technology/2016/nov/29/its-no-christmas-no-1-but-ai-generated-song-brings-festive-cheer-to-researchers 
18. MusicNet is an annotated set of classicial music performances for training machine learning models: https://techcrunch.com/2016/11/30/musicnet-aims-to-give-machine-learning-algorithms-a-taste-for-beethoven/ 
19. There seems to be something of a shortage of trained AI practitioners: http://www.techproresearch.com/downloads/research-companies-lack-skills-to-implement-and-support-ai-and-machine-learning/?ftag=tip185eb84 
20. Amazon launches its AI web platform: https://techcrunch.com/2016/11/30/amazon-launches-amazon-ai-to-bring-its-machine-learning-smarts-to-developers/ 
21. The US government is continuing to take AI seriously. Is the NZ government going to do the same? http://www.techrepublic.com/article/us-senate-subcommittee-meets-on-the-dawn-of-ai-today-livestream-available/ 
22. The optimisation problems with deep neural networks: http://www.kdnuggets.com/2016/12/hard-thing-about-deep-learning.html 
23. Does Google have the edge in cloud-based AI? http://www.techrepublic.com/article/the-cloud-war-moves-to-machine-learning-does-google-have-an-edge/ 
24. Learn maths if you want to get into AI, according to Facebook's head of AI research: https://techcrunch.com/2016/12/01/facebooks-advice-to-students-interested-in-artificial-intelligence/ 
25. Bringing AI to logo design-sounds like an interactive evolutionary algorithm: https://techcrunch.com/2016/12/01/logojoy-makes-designers-unemployed/ 
26. Detecting diabetic retinopathy (damage to the retina caused by diabetes) with machine learning: http://betanews.com/2016/11/29/google-machine-learning-diabetes-retinopathy-eyes-vision/
27. Facebook is developing AI to flag "offensive" videos: http://www.reuters.com/article/us-facebook-ai-video-idUSKBN13Q52M
28. The 10 biggest failures in applications of artificial intelligence for 2016: http://www.techrepublic.com/article/top-10-ai-failures-of-2016/ 
29. Microsoft in embedding image recognition AI into some of its Office applications: http://www.theverge.com/2016/12/2/13825590/microsoft-office-apps-ai-word-powerpoint-accessibility 
30. Random forests in Python: http://www.kdnuggets.com/2016/12/random-forests-python.html 
31. How can governments regulate ecommerce (and the AI that drives it): https://www.theguardian.com/commentisfree/2016/dec/04/how-do-you-throw-book-at-an-algorithm-internet-big-data