- If you have problem with comics, just as I do, don't worry, you are not alone. AI Machine Attempts to Understand Comic Books ... and Fails
- How will we begin Defining our relationship with early AI?
- Big conference on AI, NISP, took place. Read NIPS 2016 Reflections
- Should you go deeper with your knowledge of back-propagation? Yes you should understand backprop
- Do you worry about bias in AI? Read Bias in ML, and Teaching AI
- Talking about biases, Troubling Study Says Artificial Intelligence Can Predict Who Will Be Criminals Based on Facial Features
- We should think about The ethics of artificial intelligence
- Basic Statistical Mistakes and How to Avoid Them
- Read some Practical advice for analysis of large, complex data sets
- Worrying about your job? If you are a journalist, you should read Building a Robot Journalist 🤖
- Interested in Reinforced learning? Start by reading Demystifying Deep Reinforcement Learning
- If you are a researcher, you should check out New AI-Based Search Engines are a “Game Changer” for Science Research
- To learn more about unsupervised learning, watch RI Seminar: Yann LeCun : The Next Frontier in AI: Unsupervised Learning
- And to learn about reinforced learning watch Reinforcement Learning Introduction by Jacob Schrum
Friday, December 23, 2016
In this week edition you can find articles about AI replacing journalists, bias in machine learning or more about reinforced learning. Enjoy!
Friday, December 16, 2016
In this week edition we are presenting new playgrounds for you to use for research, several talks and as always some papers. Enjoy!
- Neural networks can now use working memory. Read What Happens When You Give an AI a Working Memory?
- What are biggest challanges for AI? Read These are three of the biggest problems facing today's AI
- Should we go deeper with neural networks? Read ResNets, HighwayNets, and DenseNets, Oh My!
- How can AI learn while doing stuff? Read Deep Reinforcement Learning with Online Generalized Advantage Estimation
- AI can now know if you are making fun of it. Read How Vector Space Mathematics Helps Machines Spot Sarcasm
- How bad data can make bad AI. Read Machine Bias
- If you want to know insides of neural nets, read 5 algorithms to train a neural network
- Is AI The new technology? Watch Andrew Ng: Why Artificial Intelligence Is the New Electricity
- We don't need so much data anymore. Read Machines Can Now Recognize Something After Seeing It Once
- In reality, today's AI are dumb. Read Artificial Intelligence Is More Artificial Than Intelligent
- Interested in philosophy? Watch Are Machines Conscious?
- Loads of interesting books and papers about AI. Categorised and with ratings. Check out Deep Learning Papers Reading Roadmap
- DeepMind released their playground. Check out DeepMind Lab
- ... and so did OpenAI. Read about OpenAI Universe
- Watch Andrew Ng talk about Nuts and Bolts of Applying Deep Learning
- Also watch Demis Hassabis from DeepMind talk about The Future of Artificial Intelligence
- And we cannot forget about the ethics. Watch Ethics of Artificial Intelligence Discussion
Friday, November 25, 2016
In this week edition you can learn more about what we need to build human level AI, why we need to make AI smarter and also some useful tricks for research. Enjoy!
- Do you have insufficient amount of data? Read What you need to know about data augmentation for machine learning
- For better LSTM with sequential data read Learning Scalable Deep Kernels with Recurrent Structure
- Gradient descent is basis for many operations in neural networks. If you want it to learn linear dynamical system, read Gradient Descent Learns Linear Dynamical Systems
- AI can now do lipreading far better than humans can. Read LipNet: Sentence-level Lipreading
- We can now let AI do physics experiments on its own. Simple ones. Learning to perform physics experiments via deep reinforcement learning
- “People worry that computers will get too smart and take over the world, but the real problem is that they’re too stupid and they’ve already taken over the world.” For more read There is a blind spot in AI research
- Machine Learning is the New Statistics, if you don't agree, take it up with the author.
- Can deep learning debug biology?. Sure it can and it will get better.
- To learn about reinforced learning on game of battleship, read Deep reinforcement learning, battleship
- AI can now translate in linear time. Neural Machine Translation in Linear Time
- Recurent neural networks are good for sequential data. But what about space and time mixture? Read Structural-RNN: Deep Learning on Spatio-Temporal Graphs
- What we need to do to start Building Machines That Learn and Think Like People?