I’ve more questions than answers, this is why I’ve become a researcher.
Following the publication of my open access book From Python to Numpy, I’ve been proposed by a publisher to make a new online book on matplotlib and scientific visualization. This publisher offers me 16% on net sales…
I just read an article on the new neural translation engine of Google and I naturally wanted to test it by myself, with the idea of finally being able to finally write a scientific article in my mother tongue that can be read by an international audience. The results are indeed remarkable in many respects. This software allows me to be able to express myself in my mother tongue (French in my case) while being assured that this engine will be able to translate the thousand and one subtleties of my mother tongue in a correct and understandable English by all. For anyone who has English as a native language this may seem a minor result, but let me tell you that for a vast majority of the scientific community, this exploit opens up new and formidable perspectives. This would allow us finally to write an article focusing on structure and content rather than waste time finding periphrases enabling us to bypass linguistic difficulties. If you are a native speaker, you probably have no idea of the time that we, as non-native writers, can lose by trying to translate our thoughts into an English that remains grammatically correct while conveying the message that we have in mind. It is an all-time struggle that is extremely frustrating when we know in advance that our sentences will not reflect our deep thoughts. This engine now allows us to be able to write simply and directly in our own language, relegating to the machine the task of translating our words into another language. No more derogatory remarks from the reviewers regarding an inadequate level of English. Of course, there is no miracle. If my message is obscure in French, it will remain in English. If my sentences are too long in French, they will remain in English. If the typically French expression I want to use does not exist in English, it is unlikely that the engine will know its equivalent (but who knows…).
I’ve a cat at home. Her name is Charlie (she’s a female). I’ve been living with her for more than 16 years and over the years, I’ve come to know her quite well. Even if she doesn’t speak, I usually communicate a lot with her in the course of a day. If she’s hungry, angry, playful or simply needs to go outside, she knows very well how to express these feelings and communicate them to me because I’ve learn to distinguish the whole range of her miaowing. If you ask me if she thinks and reasons, I would certainly answer positively even if I’ve no way to prove that fact. But the way she can manipulate me by purring next to my ear when she wants something makes me really believe she’s definitely a cognitive and smart being.
popular neuroscience AI
In just over 30 years, humans will be able to upload their entire minds to computers and become digitally immortal. – Ray Kurzweil, Global Futures 2045 International Congress, 2013. Without even considering the ethical, philosophical, social or legal scope of such a statement, it’s important to consider if it actually makes any sense. To try to give an educated guess, we have to move away from computer science and look at what biology and neuroscience can teach us.
Information is power. But like all power, there are those who want to keep it for themselves. The world’s entire scientific and cultural heritage, published over centuries in books and journals, is increasingly being digitized and locked up by a handful of private corporations. Want to read the papers featuring the most famous results of the sciences? You’ll need to send enormous amounts to publishers like Reed Elsevier.