This month’s collection of interesting clauses that point to important tends is dominated by AI. That’s not surprising; AI has probably been the biggest single category all time. But its predominance over other topics seems to be increasing. That’s partly because there’s more research into why AI disappoints; partly because we’re beginning to see AI in embedded methods, wandering from monstrous gas and oil wells to the tiny manoeuvres that Pete Warden is working with.
Teaching AI to manipulate and cause: Combine NLP with reinforcement learning, and train in a multiplayer role-playing tournament. This is where AI comes scary, particularly since AI organizations don’t understand what they’re doing( construe the next entry ). GPT-3 is great at developing human-like language, but that’s as far as it travels; it has no sense of what an adequate response to any induce are likely to be. For example, proposing suicide as a solution to depression. This isn’t a surprise, but it means that GPT-3 genuinely can’t be incorporated into applications.Why machine learning representations flunk in the real world, and why it’s a very difficult problem to fix: Any organize of training data can lead to a huge number of simulations with similar behavior on the training data, but with very different performance on real-world data. Deciding which of these models is “best”( and in which situations) is a difficult, and unstudied, problem.Tiny NAS: Neural Architecture Search designed to automate building Tiny Neural Networks. Machine Learning on small devices will be an increasingly important topic in the coming years.Pete Warden on the future of TinyML: There will be hundreds of billions of machines in the next few years. Many of them won’t be “smart”; they’ll be more intelligent versions of dumb manoeuvres. We don’t need “smart refrigerators” that can order milk automatically, but we do need refrigerators that can use energy more efficiently and notify us when they’re about to fail.The replication crisis in AI: Too many academic AI newspapers are published without code or data, and using hardware that can’t be obtained by other researchers. Without access to code, data, and hardware, academic articles about groundbreaking results are little more than corporate marketing.Machine learning to see gas leaks: Granted, this is for oil-well scale natural gas seeps, but we should all be more aware of these invisible applications of machine learning. It’s not just autonomous vehicles and face recognition. And lest we forget, invisible an applicant for ML too have problems linked to bias, fairness, and accountability.Vokens: What happens when you combine computer image with natural language processing? Is it is feasible to isolate the meaningful elements in a depict, then use that to inform language frameworks like GPT-3 to add an element of “common sense”? Using AI to diagnose COVID-1 9 via coughings: MIT has developed an AI algorithm that identifies features in a cough that indicate a COVID-1 9 illnes. It is at least as accurate as current measures, particularly for asymptomatic people, supplies solutions in real hour, and could easily be built into a cell phone app.Over time, models in feedback curves( e.g ., economic competition) tend to become more accurate for a narrower slice of the population, and less precise for the population as a whole. Essentially, a simulation that is constantly retraining on current input will, over age, make itself biased.
Robots in building: The construction industry has been resistant to automation. Canvas has built a robot that installs drywall. This robot is in use on several major sites, including the renovation of the Harvey Milk terminal at San Francisco Airport.Simplifying the robot’s model of the external life is the route to better collaborations between robots and humans.Honda winnings endorsement to sell a level-3 autonomous vehicle. The vehicle is capable of completely taking over driving in certain situations , not just assisting. It should be on sale before March.
Nbdev is a literate programme environment for Python. It is based on Jupyter, but encompass the part application lifecycle and CI/ CD pipeline , not just programming.A visual programming environment for GraphQL is another step in getting beyond text-based programming. A visual environment seems like an obvious select for working with graph data.PHP 8 is out! PHP is an old language, and this release isn’t likely to put it onto the “trendy language” list. But with a huge portion of the Web built with PHP, this new release is important and surely worth watching.
Privacy and Security
Google is adding end-to-end encryption to their implementation of RCS, which is a standard designed to replace SMS messaging. RCS hasn’t been adopted widely( and, given the dominance of the telephone system, may never be adopted widely ), but standards for encrypted messaging are an important step forward.Tim Berners-Lee’s privacy project, Solid, has released its first project: an organizational privacy server. The theme behind Solid is that people( and organizations) store their own data in secure repositories called Husks that they control. Bruce Schneier has entered into Inrupt, the company commercializing Solid.CMU has shown that passwords with minimum segment of 12 attributes and that pass some simple evaluations can be remembered and refuse affect. We can move on from password policies that require obscure combinations of upper and lowercase, punctuation and figures, and that don’t require deepening passwords regularly.
Remember DNS cache poisoning? It’s back. Regrettably. A public mesh WiFi network for New York City: Mesh systems can provide Internet access in locations where built providers don’t care to go-but seeing them work at scale is difficult. Technology we firstly heard about in Cory Doctorow’s very strange Someone Comes To Town, Someone Leaves Town.Hyper-scale indexing: Helios is Microsoft’s reference architecture for the next generation of cloud methods. It is capable of handling extremely large data sets( even by modern standards) and combinations unified cloud computing with shape calculating.
The Raspberry Pi 400 looks like a LOT of enjoyable. It’s Raspberry Pi 4 were integrated into a keyboard( like the very early Personal Computers ); 1.8 GHz ARM processor, 4 GB RAM, more I/ O ports than a MacBook Pro; exactly needs a monitor. I time hope the keyboard is good.I should say something positive about Apple’s M1, but I won’t. I’m disenchanted fairly with them as a company that I really don’t care how good the processor is.
Amazon critiques about scented candlesthat don’t smell correlate to Covid. A delightful application of data analysis using publicly available sources. Data science triumphs.
Read more: feedproxy.google.com