The very latest advancements in the materials science field combined with machine / deep learning can be found on Arxiv.org through part of the 125 papers appearing at https://arxiv.org/find/all/1/all:+AND+EXACT+materials_science+EXACT+machine_learning/0/1/0/all/0/1 . Spent some time studying and got the methods for checking properties and for finding potential new alloys. I am now ready to contribute to Gen IV nuclear research, looking for materials more resistant to radiation, corrosion, creep and so on. Going to make tests with my laptop, write a paper and submit abstracts to a few conferences from May.
Online self-education is a delusion because passive enthusiasts and desperate folks do not just really need tools but a purpose or a market. A purpose is needed to apply that knowledge in a domain (but few have a domain knowledge), a market is needed to get money from the effort (but remote work is actually not so easy & available). More in general, the very recent democratisation of knowledge opens the average joes and marys from marginal markets, minorities etc. etc. to a nice range of personal opportunities to climb the ladder but also to exploitation and depression, once they understand they are still where they started from. I want to say that the world still needs local hubs with real persons from different backgrounds with different needs to create and sustain a market, so possibly knowledge dissemination should come on top of other actions to make a real impact for a great number of persons.
The sovranist League and Berlusconi’s Forza Italia won the great majority of seats in the affluent North thanks to two promises: solve immigration and lower taxes. Immigration is difficult and Europe is not taking quotas so Italy must do on its own without a plan, just getting all onshore and relocating them in the country. Locals are clearly not happy. There are about 500k illegal migrants in Italy these days, young men without a permit to stay that roam freely, unemployed, unattached and drifting towards crime. Scientistic solution then? Allow police much more force to contrast crime: the majority of drifters do not want integration at all and can’t help themselves because laws are lenient, so just beat them hard when caught. Word spreads quickly and drifters will desist or behave.
Elon Musk is the techno-visionary behind Tesla electric self-driven cars and SpaceX for Mars colonisation. His latest interview is interesting for three reasons: introductory steps towards Mars colonisation will start in 2019 because he fears a third world war and humanity annihilation, so another breed is needed elsewhere to prevent extinction; artificial intelligence as a weapon is much more dangerous than nuclear bombs and its exceptionally fast development should be regulated through a security treaty at global level; both SpaceX and Tesla are financially weak so there is a always risk they go bust (and we lose all the r&d effort from a West-friendly operator). Keep going, Elon!
As a middle-aged self re-learner since 2015 with little time to dig deep on my own, I have been able to make incredible progress thanks to the public work of Andrew Ng, Kaggle.com, fast.ai & many others (F.Chollet among them). My question is simple: how you do give back in a meaningful way? How can I give back in some capacity? Suggestions received at 10 Mar 2018: 1- Share interesting tutorials with using non-standard datasets; 2- Donate money to charities for better life in the poorest countries; 3- Encourage diversity for wider access to tools and knowledge. Three populist buzzwords for political correctness at global level in 2018 then: share, donate, encourage.
Universal basic income (with some caveat and subject to an inclusion plan to find a job) was the main selling point for the populist 5 Stars movement in the deprived Southern Italy and they got 40% to 50% votes there! Where would the 17 billion euros per year needed to implement this policy come from? It’s pie in the sky, we all know, don’t we? Good old collectivism works well with scientism though: not giving cash in hand then but with crediting day-to-day points onto citizens’ ID card to be converted into the bare minimum food & drink to survive, think of a k-ration, from local supermarkets or automatic dispensing machines. No actual money circulated, no social incentive to stay inactive, no bloodbath for the public finances aka the national debt. If people refuse, give them medical pills to satisfy their bare dietary needs instead. True scientism.
The Italian general election results indicated that populists and sovranists are going to replace plutocrats and dishonests in the government. Populists triumphed in the deprived South and sovranists in the affluent North, that making another tasty twist. The social democracy vision and the rational European architecture are not fashionable anymore. I will try and contribute here with scientistic answers to the most urgent problems Italy is going to face in the coming months. In fact, I am pretty sure bare naked scientism would not do worse than populism and sovranism. I am also sure there will be more food for thought.
DroneDays 2018 will be held in Bruxelles, Belgium, next 9-11 March 2018. “Civil and commercial use of drones (known as well as RPAS or UAV) has endless application’s domains. This year again, DroneDays Matchmaking event – in collaboration with Enterprise Europe Network and the DroneDays Expo – will provide its participants with an amazing package of opportunities to develop business at international level!” (source: their website) Mauro Pompetti, DronesBench CEO, will attend the matchmaking event looking for business and new partnerships to develop his project further.
“Alongside these new robotics environments, we’re also releasing code for Hindsight Experience Replay (or HER for short), a reinforcement learning algorithm that can learn from failure. Our results show that HER can learn successful policies on most of the new robotics problems from only sparse rewards. Below, we also show some potential directions for future research that could further improve the performance of the HER algorithm on these tasks… the reinforcement learning algorithm can obtain a learning signal since it has achieved some goal; even if it wasn’t the one that we meant to achieve originally. If we repeat this process, we will eventually learn how to achieve arbitrary goals, including the goals that we really want to achieve.” Source – 26 Feb 2018 “Ingredients for Robotics Research” at OpenAI