Google Teaches Computers How To Regret
What makes the human brain fascinating is our ability to learn from everyday experiences and have the option to better ourselves. Due to the organic nature of the brain, our brains constantly re-wire and prune unused neural connections in effort to keep cognitive functions that are truly necessary for our everyday use while “losing” neural circuits we no longer need. This is exactly what is happening at a neurological level when we find ourselves “rusty” in a language we haven’t used in years. Our brains assume we no longer have use for the language due to lack of use, and it prunes those connections it deems unnecessary to “make room” for more relevant stuff.
The human brain also experiences the emotion of regret because the act of regretting welcomes reflection on what we should have done in a particular situation. By feeling regret, we are able to perhaps make a better decision the next time we are faced in a similar situation. What about computers, will they ever be able to function on a human level? Google is interested in answering that question. The folks at Google are funding a research team at the Tel Aviv University (pictured) to create an algorithm that will teach computers how to experience “virtual regret.” This will give computers a better sense of interpreting mistakes and creating a better future.
Professor Yishay Mansour of the university states, “If the servers and routing systems of the Internet could see and evaluate all the relevant variables in advance, they could more efficiently prioritize server resource requests, load documents and route visitors to an Internet site, for instance.” The team wants to help computers differentiate between a desired outcome and an actual outcome. This is great because your computer will work better for you by studying and learning from your interactions and determining if and when a better method could be employed, that it does. What troubles me is where this “learned and studied” information gets stored and if the data is secure enough from theft. Imagine if a criminal had access to all that personal data?