Review for 10 April 2017

Below are some of the interesting links I Tweeted about recently.

1. The limits of deep learning, from the point of view of a psychologist: https://techcrunch.com/2017/04/01/discussing-the-limits-of-artificial-intelligence/
2. Photographic style transfer using deep neural networks: http://www.theverge.com/2017/3/30/15124466/ai-photo-style-transfer-deep-neural-nets-adobe 
3. AI and automation might not cause a wholesale destruction of jobs after all: http://www.theregister.co.uk/2017/03/30/ai_jobs_doomsday_may_not_actually_happen_say_boffins/ 
4. How to launch a successful AI startup: http://www.techrepublic.com/article/how-to-launch-a-successful-ai-startup/ 
5. Optimising social media campaigns with AI-not sure what the AI part is, though: https://techcrunch.com/2017/04/04/clearmob-social-campaigns/ 
6. Top 16 R packages for machine learning-beware, need to register to download full list: http://blog.thedataincubator.com/2017/03/ranked-16-r-packages-for-machine-learning/ 
7. Will some governments introduce a "human quota" as more work becomes automated? https://www.theguardian.com/technology/2017/apr/04/innovation-in-ai-could-see-governments-introduce-human-quotas-study-says 
8. A model of AI maturity: http://www.kdnuggets.com/2017/04/ai-maturity-model.html 
9. Top mistakes data scientists make: https://cyborgus.com/2017/04/03/top-mistakes-data-scientists-make/ 
10. My friend @DavidBFogel has a new website up with curated news on science, tech, AI: http://www.davidfogel.com
11. Improved sentiment analysis with a recurrent neural network: https://techcrunch.com/2017/04/07/openai-sets-benchmark-for-sentiment-analysis-using-an-efficient-mlstm/ 
12. Using ML to improve selfies: http://www.theverge.com/tldr/2017/4/6/15209202/adobe-sensei-selfie-improving-ai-machine-learning 
13. Keeping private data private when using it to train AI: https://motherboard.vice.com/en_us/article/google-thinks-it-can-solve-artificial-intelligences-privacy-problem 
14. Do laptops in class harm academic performance? Probably depends on the subject: https://www.timeshighereducation.com/news/using-laptops-in-class-harms-academic-performance-study-warns 
15. Levels of data readiness for organisations approaching AI applications: http://www.datasciencecentral.com/profiles/blogs/data-readiness-strategies-for-ai-start-ups 
16. Build TensorFlow ANN faster with Sonnet, now open-sourced by Google: https://techcrunch.com/2017/04/07/google-deepmind-open-sources-sonnet-so-you-can-build-neural-networks-in-tensorflow-even-quicker/