Artificial intelligence is rapidly changing many aspects of how we work and live. (How many stories did you read last week about self-driving cars and job-stealing robots? Perhaps your holiday shopping involved some AI algorithms, as well.) But despite the constant flow of news, many misconceptions about AI remain.
AI doesn’t think in our sense of the word at all, Scriffignano explains. “In many ways, it’s not really intelligence. It’s regressive.”
IT leaders should make deliberate choices about what AI can and can’t do on its own. “You have to pay attention to giving AI autonomy intentionally and not by accident,”
7,002 Comments
Tomi Engdahl says:
Intel’s Movidius Myriad 2 VPU Takes Artificial Intelligence Into Space Aboard the PhiSat-1
https://www.hackster.io/news/intel-s-movidius-myriad-2-vpu-takes-artificial-intelligence-into-space-aboard-the-phisat-1-af8b6e0b5c5b
AI chip analyzes imagery for cloud cover then discards useless imagery prior to transmission, saving about 30 percent of its bandwidth.
Tomi Engdahl says:
https://www.iflscience.com/technology/these-ai-generated-scenes-from-the-great-british-baking-show-may-give-you-nightmares/
Tomi Engdahl says:
Tekoälyssä on johtaja-ainesta – kunnes päästään tunteisiin
https://www.tuni.fi/unit-magazine/artikkelit/tekoalyssa-johtaja-ainesta-kunnes-paastaan-tunteisiin
Heikot signaalit havaitseva algoritmi voi auttaa johtajia riskien hallinnassa, ongelmien havaitsemisessa ja päätöksenteossa. Ihmisillä on kuitenkin aina vastuu päätösten seurauksista. Tunteitakin tarvitaan.
Tomi Engdahl says:
How Facebook’s AI Tools Tackle Misinformation
https://spectrum.ieee.org/view-from-the-valley/artificial-intelligence/machine-learning/how-facebooks-ai-tools-tackle-misinformation
Facebook today released its quarterly Community Standards Enforcement Report, in which it reports actions taken to remove content that violate its policies, along with how much of this content was identified and removed before users brought it to Facebook’s attention. That second category relies heavily on automated systems developed through machine learning.
In recent years, these AI tools have been focused on hate speech. According to Facebook CTO Mike Schroepfer, the company’s automated systems identified and removed three times as many posts containing hate speech in the third quarter of 2020 as in the third quarter of 2019. Part of the credit for that improvement, he indicated, goes to a new machine learning approach that uses live, online data instead of just offline data sets to continuously improve. The technology, tagged RIO, for Reinforced Integrity Optimizer, looks at a number tracking the overall prevalence of hate speech on the platform, and tunes its algorithms to try to push that number down.
Tomi Engdahl says:
Google proposes applying AI to patent application generation and categorization
https://venturebeat.com/2020/11/20/google-proposes-applying-ai-to-patent-application-generation-and-categorization/
Google asserts that the patent industry stands to benefit from AI and machine learning models like BERT, a natural language processing algorithm that attained state-of-the-art results when it was released in 2018.
Tomi Engdahl says:
https://towardsdatascience.com/probability-for-machine-learning-2cfe4aa13101
Tomi Engdahl says:
The way we train AI is fundamentally flawed
https://www.technologyreview.com/2020/11/18/1012234/training-machine-learning-broken-real-world-heath-nlp-computer-vision/
The process used to build most of the machine-learning models we use today can’t tell if they will work in the real world or not—and that’s a problem.
Tomi Engdahl says:
See the Inner Workings of a Convolutional Neural Network with This PCB Business Card
This business card made by Paul Klinger is able to classify digits by running a CNN and show each layer’s state.
https://www.hackster.io/news/see-the-inner-workings-of-a-convolutional-neural-network-with-this-pcb-business-card-ac92186dc15a
Tomi Engdahl says:
https://medium.com/qiskit/upcoming-video-game-will-generate-new-levels-using-qiskit-and-a-quantum-simulator-b47dfc911234
Tomi Engdahl says:
AI System Beats Supercomputer in Combustion Simulation
https://spectrum.ieee.org/tech-talk/computing/hardware/ai-system-beats-supercomputer-at-key-scientific-simulation
Tomi Engdahl says:
Open-source AI enhances IoT device security
https://www.edn.com/open-source-ai-enhances-iot-device-security/
One major challenge to cyber-security in Internet of Things (IoT) devices is the constantly evolving nature of threats. New vulnerabilities are continually being found and exploited and new methods of attack are evolving, turning IoT security into an ongoing battle for developers. Now, however, an emerging approach to IoT security using artificial intelligence (AI) promises to provide protection against both known and new, unknown threats.
Tomi Engdahl says:
Light-Driven All-in-One Chip Could Dramatically Improve Edge AI Performance, Efficiency
Brain-inspired chip handles imaging, memory, and even AI processing, and is being positioned as a breakthrough in neurobotics.
https://www.hackster.io/news/light-driven-all-in-one-chip-could-dramatically-improve-edge-ai-performance-efficiency-0188f7a6c13a
Tomi Engdahl says:
Practical Machine Learning Tutorial: Part.4 (Model Evaluation-2)
Multi-class Classification Problem: Geoscience example (Facies)
https://towardsdatascience.com/practical-machine-learning-tutorial-part-4-model-evaluation-2-764d69f792a5
Tomi Engdahl says:
Can AI Bots Make Life-and-Death Decisions?
https://discover.bot/bot-talk/can-ai-bots-make-life-and-death-decisions/
Artificial intelligence (AI) bots pervade every aspect of human activity. The performance of AI often exceeds human baselines. Moreover, the MIT designed a Moral Machine application with a user interface to simulate unsolvable moral dilemmas. Fortunately, AI researchers have found ways to make AI bots reliable in life-and-death situations.
Tomi Engdahl says:
https://www.edn.com/googles-aiy-kits-offer-do-it-yourself-artificial-intelligence/
Tomi Engdahl says:
Neuromorphic Computing: How the Brain-Inspired Technology Powers the Next-Generation of Artificial Intelligence
Brain-inspired computing for Machine Intelligence emerges as neuromorphic chips after over 30 years it was first developed.
https://interestingengineering.com/neuromorphic-computing-how-the-brain-inspired-technology-powers-the-next-generation-of-artificial-intelligence
Tomi Engdahl says:
How to enable Automated Machine Learning in MySQL
https://docs.mindsdb.com/databases/tutorials/AiTablesInMySQL/?utm_source=facebook&utm_medium=paid+social&utm_campaign=mysql&fbclid=IwAR2D2M0nWMSjqMlBWewxqxvuPQv05Z6LAYqV28RdTNIFy5qQsSdqCt3UZDI
Database is surely the best place for Machine Learning – because data is the main ingredient of it. And now you can build, train, test & query Machine Learning models using standard SQL queries within MySQL database!
This doesn’t require hardcore data science knowledge – the whole Machine Learning workflow is automated.
This solution is called AI-Tables and is available in MySQL thanks to integration with an open-source predictive engine from MindsDB. AI-Tables look like normal database tables and return predictions upon being queried as if it is data that exists in the table.
Tomi Engdahl says:
True story what is behind AI?
https://www.facebook.com/508244702677646/posts/1673965349438903/
Tomi Engdahl says:
Deep Learning Has Reinvented Quality Control in Manufacturing—but It Hasn’t Gone Far Enough
https://spectrum.ieee.org/tech-talk/artificial-intelligence/machine-learning/deep-learning-has-reinvented-quality-control-in-manufacturingbut-it-hasnt-gone-far-enough
In 2020, we’ve seen the accelerated adoption of deep learning as a part of the so-called Industry 4.0 revolution, in which digitization is remaking the manufacturing industry. This latest wave of initiatives is marked by the introduction of smart and autonomous systems, fueled by data and deep learning—a powerful breed of artificial intelligence (AI) that can improve quality inspection on the factory floor.
The benefit? By adding smart cameras to software on the production line, manufacturers are seeing improved quality inspection at high speeds and low costs that human inspectors can’t match. And given the mandated restrictions on human labor as a result of COVID-19, such as social distancing on the factory floor, these benefits are even more critical to keeping production lines running.
While manufacturers have used machine vision for decades, deep learning-enabled quality control software represents a new frontier. So, how do these approaches differ from traditional machine vision systems? And what happens when you press the “RUN” button for one of these AI-powered quality control systems?
For conventional deep learning to be successful, the data used for training must be “balanced.” A balanced data set has as many images of good valves as it has images of defective valves, including every possible type of imperfection. While collecting the images of good valves is easy, modern day manufacturing has very low defect rates. This situation makes collecting defective images time consuming, especially when you need to collect hundreds of images of each type of defect. To make things more complex, it’s entirely possible that a new type of defect will pop up after the system is trained and deployed—which would require that the system be taken down, retrained, and redeployed. With wildly fluctuating consumer demands for products brought on by the pandemic, manufacturers risk being crippled by this production downtime.
Tomi Engdahl says:
Circle of AI Life. Via https://www.monkeyuser.com/2020/circle-of-ai-life/
Tomi Engdahl says:
Google’s DeepMind AI Just Taught Itself To Walk
https://m.youtube.com/watch?v=gn4nRCC9TwQ
Tomi Engdahl says:
Drop-in CNN block replacement means models with fewer layers/smaller intermediate representations, dramatically reducing memory and compute.
RNNPool Reduces Computer Vision RAM Usage 10X on Edge Devices Without Sacrificing Accuracy
https://www.hackster.io/news/rnnpool-reduces-computer-vision-ram-usage-10x-on-edge-devices-without-sacrificing-accuracy-00266a4d892b
Drop-in CNN block replacement means models with fewer layers/smaller intermediate representations, dramatically reducing memory and compute.
While increasingly sophisticated convolutional neural networks (CNNs) and AI-accelerated hardware have resulted in countless breakthroughs in the field of computer vision in recent years, as these technologies approach the edge, limited RAM and computing power can make conventional methods impractical, or even impossible. One area where memory usage can be particularly high in CNNs is the activation map output layer — a 3D tensor that increases in size with the number of features, and consumes gobs of RAM even when precision is reduced. Researchers at Microsoft have introduced RNNPool, which replaces the pooling layers of a CNN with recurrent neural networks (RNNs) that consume much less RAM, without sacrificing accuracy.
RNNPool’s RNNs sweep activation maps in several passes, downsampling without degrading accuracy, producing an equivalent output to traditional CNN pooling operators, but with far lower peak RAM usage. Since RNNPool obviates the need for several layers in the CNN, compute is also reduced. In testing, Microsoft researchers saw an 8-10X reduction in memory usage for vision tasks, as well as 2-3X computational reduction, all without sacrificing accuracy. Compared to MobileNetV2, an RNNPool-based model called RNNPool-Face-M4 actually performed more accurately while using only one fifth of the RAM.
https://github.com/microsoft/EdgeML/blob/master/pytorch/edgeml_pytorch/graph/rnnpool.py
Tomi Engdahl says:
New AI improves itself through Darwinian-style evolution
https://bigthink.com/surprising-science/automl?utm_term=Autofeed&utm_medium=Social&utm_source=Facebook#Echobox=1607663743
AutoML-Zero is a proof-of-concept project that suggests the future of machine learning may be machine-created algorithms.
Automatic machine learning is a fast-developing branch of deep learning.
It seeks to vastly reduce the amount of human input and energy needed to apply machine learning to real-world problems.
AutoML-Zero, developed by scientists at Google, serves as a simple proof-of-concept that shows how this kind of technology might someday be scaled up and applied to more complex problems.
Machine learning has fundamentally changed how we engage with technology. Today, it’s able to curate social media feeds, recognize complex images, drive cars down the interstate, and even diagnose medical conditions, to name a few tasks.
Tomi Engdahl says:
4 basic steps in implementing an AI-driven design workflow
https://www.edn.com/four-basic-steps-in-implementing-an-ai-driven-design-workflow/?utm_content=bufferf9bad&utm_medium=social&utm_source=edn_facebook&utm_campaign=buffer
Tomi Engdahl says:
Ten Deep Learning Concepts You Should Know for Data Science Interviews
Study smart, not hard.
https://towardsdatascience.com/ten-deep-learning-concepts-you-should-know-for-data-science-interviews-a77f10bb9662
Tomi Engdahl says:
5 Reasons You Don’t Need to Learn Machine Learning
An increasing number of influencers preach you should start learning Machine Learning. Should you listen to them?
https://towardsdatascience.com/5-reasons-you-dont-need-to-learn-machine-learning-5f9b1ddf8eb5
Tomi Engdahl says:
AWS launches Trainium, its new custom ML training chip
https://techcrunch.com/2020/12/01/aws-launches-trainium-its-new-custom-ml-training-chip/
Tomi Engdahl says:
AWS launches SageMaker Data Wrangler, a new data preparation service for machine learning
https://techcrunch.com/2020/12/01/aws-launches-sagemaker-data-wrangler-a-new-data-preparation-service-for-machine-learning/
Tomi Engdahl says:
AlphaFold Proves That AI Can Crack Fundamental Scientific Problems
https://spectrum.ieee.org/tech-talk/artificial-intelligence/medical-ai/alphafold-proves-that-ai-can-crack-fundamental-scientific-problems
Tomi Engdahl says:
https://www.facebook.com/126000117413375/posts/4134177553262258/
//(WARNING: loud background music) I would instead go work as a handyman or work in a farmhouse. I don’t want this stupid thing anywhere near my computer. what do you think? #Programming #AI #ML
Tomi Engdahl says:
Over 1,000 Experts Call Out “Racially Biased” AI Designed To Predict Crime Based On Your Face
https://www.iflscience.com/technology/over-1000-experts-call-out-racially-biased-ai-designed-to-predict-crime-based-on-your-face/
In an upcoming book to be published by Springer Nature, Transactions on Computational Science & Computational Intelligence, the team from Harrisburg University outlined a system they created that they claimed (in a press release that has now been removed from online), “With 80 percent accuracy and with no racial bias, the software can predict if someone is a criminal based solely on a picture of their face. The software is intended to help law enforcement prevent crime.”
Alarmed by the many and immediate problematic assumptions and repercussions of using “criminal justice statistics to predict criminality,”
“Let’s be clear: there is no way to develop a system that can predict or identify ‘criminality’ that is not racially biased — because the category of ‘criminality’ itself is racially biased,” adding “data generated by the criminal justice system cannot be used to “identify criminals” or predict criminal behavior. Ever.”
The authors of the letter write that research like this rests on the assumption that data on criminal arrests and convictions are “reliable, neutral indicators of underlying criminal activity,”
Tomi Engdahl says:
Amazon AWS says ‘Very, very sophisticated practitioners of machine learning’ are moving to SageMaker
https://www.zdnet.com/article/aws-says-very-very-sophisticated-practitioners-of-machine-learning-are-moving-to-sagemaker/
The infrastructure software is catching on with some of the most demanding machine learning scientists at big firms such as Lyft and Intuit, and also taking over Amazon dot com’s internal ML development.
Tomi Engdahl says:
3 Ways AI Transforms How We Develop Software
https://thorgate.eu/blog/3-ways-ai-transforms-how-we-develop-software?fbclid=IwAR1Xgu_h4m8quRh-fgLzdmoGfm_V2qKIy8Er1JKZPCYbfsA9mWx4n2N6eMI
Since its arrival, Artificial Intelligence (AI) has emerged as a breakthrough technology transforming almost every sector. Healthcare, education, customer services, retail, agriculture, banking, and financial services are some of the key sectors revolutionized by AI for good. One such market heavily influenced by AI is software development. Software development is not the same as it was years ago; the influence of AI has made it faster, productive, and cheaper. Today, using AI, developers can build powerful and productive software in a lesser time.
Tomi Engdahl says:
What does AI have to do with customer experience?
https://radly.fi/blog/three-ways-to-create-a-better-customer-experience-through-machine-learning-and-ai/
When we talk about AI in this blog post, we do not speak of the human-like (or terminator kind of) intelligence which is often referred to as broad AI. But rather about so-called narrow AI, which implies data analysis and machine learning. Thus when reading forward you can think about AI as an umbrella term and machine learning as a toolbox of different methodologies that can be used to learn from data.
Tomi Engdahl says:
DeepMind posted champion-level gameplay in go, chess and shogi. For these matches, it knew the rules. Now comes MuZero a neural net that figures out rules for itself. Writes our @pross356, Cue shark music.
DeepMind’s New AI Masters Games Without Even Being Taught the Rules
https://spectrum.ieee.org/tech-talk/robotics/artificial-intelligence/deepminds-new-ai-masters-games-without-even-been-taught-the-rules
The folks at DeepMind are pushing their methods one step further toward the dream of a machine that learns on its own, the way a child does.
The London-based company, a subsidiary of Alphabet, is officially publishing the research today, in Nature, although it tipped its hand back in November with a preprint in ArXiv. Only now, though, are the implications becoming clear: DeepMind is already looking into real-world applications.
DeepMind won fame in 2016 for AlphaGo, a reinforcement-learning system that beat the game of Go after training on millions of master-level games. In 2018 the company followed up with AlphaZero, which trained itself to beat Go, chess and Shogi, all without recourse to master games or advice. Now comes MuZero, which doesn’t even need to be shown the rules of the game.
The new system tries first one action, then another, learning what the rules allow, at the same time noticing the rewards that are proffered—in chess, by delivering checkmate; in Pac-Man, by swallowing a yellow dot. It then alters its methods until it hits on a way to win such rewards more readily—that is, it improves its play.
Tomi Engdahl says:
Google told scientists to use ‘a positive tone’ in AI research, documents show
The company requested authors refrain from casting its technology in a negative light in at least three cases
https://www.theguardian.com/technology/2020/dec/23/google-scientists-research-ai-postive-tone
Tomi Engdahl says:
DeepMind researchers claim neural networks can outperform neurosymbolic models
https://venturebeat.com/2020/12/21/deepmind-researchers-claim-neural-networks-can-outperform-neurosymbolic-models-on-visual-tasks/
So-called neurosymbolic models, which combine algorithms with symbolic reasoning techniques, appear to be much better-suited to predicting, explaining, and considering counterfactual possibilities than neural networks. But researchers at DeepMind claim neural networks can outperform neurosymbolic models under the right testing conditions. In a preprint paper, coauthors describe an architecture for spatiotemporal reasoning about videos in which all components are learned and all intermediate representations are distributed (rather than symbolic) throughout the layers of the neural network. The team says that it surpasses the performance of neurosymbolic models across all questions in a popular dataset, with the greatest advantage on the counterfactual questions.
Tomi Engdahl says:
https://www.iflscience.com/chemistry/artificial-intelligence-solves-schrdinger-equation-for-molecules/
Tomi Engdahl says:
Why to Hire Machine Learning Engineers, Not Data Scientists
Machine Learning Engineers finally deliver on the promise of AI.
https://www.datarevenue.com/en-blog/hiring-machine-learning-engineers-instead-of-data-scientists?fbclid=IwAR2RodAnsyfo6J9a6AJEd1U3-slvjR9YNpVRPxK4fqo4hyAvbYi7B13f1Pg
Many companies hired the wrong people to build AI products. Now they’re course-correcting by looking for machine learning engineers.
Where did we go wrong?
In 2017, the hottest AI job was “data scientist” – a statistician who can code. The idea was that companies could finally build AI-driven products with the help of PhD statisticians.
Finding and hiring data scientists was hard. But even worse, things didn’t work out as planned: 18 months later, many companies ended up with just Proofs of Concept (PoCs) that would never make it to production.
Two big mistakes got us here:
1. Conflating research and business skills: There are two kinds of machine learning: academic (data scientists) and practical (machine learning engineers). To succeed in building business applications, you need people who are well versed in the practical.
2. Treating machine learning like a black box: Since there was no widespread understanding among managers about what machine learning is, data scientists were left to figure it out for themselves. This was a big mistake, because successful ML projects require a tight feedback loop between domain and data knowledge.
Enter the machine learning engineer
What many companies didn’t accept in the beginning is now becoming clear: You don’t need specialists (statisticians) to build machine learning solutions. You need generalists – experienced engineers who understand how to use AI and are also excellent communicators.
But wait – isn’t that like searching for a unicorn?
To sum up, companies that want to take an AI application from idea to production mainly need machine learning engineers. These engineers can take a business idea, identify an appropriate approach among hundreds of research papers and open source software options, test it, improve it, and – crucially – take it to production in the form of sound, reliable software.
We left out one thing: Engineers can only learn so much about your domain in a given amount of time. To gain the crucial insights that make an AI solution valuable, they fully depend on you.
Tomi Engdahl says:
You don’t code? Do machine learning straight from Microsoft Excel
https://venturebeat.com/2020/12/30/you-dont-code-do-machine-learning-straight-from-microsoft-excel/
mastering machine learning is a difficult process. You need to start with a solid knowledge of linear algebra and calculus, master a programming language such as Python, and become proficient with data science and machine learning libraries such as Numpy, Scikit-learn, TensorFlow, and PyTorch.
And if you want to create machine learning systems that integrate and scale, you’ll have to learn cloud platforms such as Amazon AWS, Microsoft Azure, and Google Cloud.
Naturally, not everyone needs to become a machine learning engineer. But almost everyone who is running a business or organization that systematically collects and processes can benefit from some knowledge of data science and machine learning. Fortunately, there are several courses that provide a high-level overview of machine learning and deep learning without going too deep into math and coding.
a very valuable and often-overlooked tool is Microsoft Excel.
To most people, MS Excel is a spreadsheet application that stores data in tabular format and performs very basic mathematical operations. But in reality, Excel is a powerful computation tool that can solve complicated problems. Excel also has many features that allow you to create machine learning models directly into your workbooks.
While Excel will in no way replace Python machine learning, it is a great window to learn the basics of AI and solve many basic problems without writing a line of code.
Linear regression machine learning with Excel
Linear regression is a simple machine learning algorithm that has many uses for analyzing data and predicting outcomes. Linear regression is especially useful when your data is neatly arranged in tabular format. Excel has several features that enable you to create regression models from tabular data in your spreadsheets
Excel’s Trendline feature can create regression models from your data
Other machine learning algorithms with Excel
Beyond regression models, you can use Excel for other machine learning algorithms. Learn Data Mining Through Excel provides a rich roster of supervised and unsupervised machine learning algorithms, including k-means clustering, k-nearest neighbor, naive Bayes classification, and decision trees.
Deep learning and natural language processing with Excel
Learn Data Mining Through Excel shows that Excel can even express advanced machine learning algorithms. There’s a chapter that delves into the meticulous creation of deep learning models. First, you’ll create a single layer artificial neural network with less than a dozen parameters. Then you’ll expand on the concept to create a deep learning model with hidden layers. The computation is very slow and inefficient, but it works, and the components are the same: cell values, formulas, and the powerful Solver tool.
In the last chapter, you’ll create a rudimentary natural language processing (NLP) application, using Excel to create a sentiment analysis machine learning model.
Excel as a machine learning tool
Whether you’re making C-level decisions at your company, working in human resources, or managing supply chains and manufacturing facilities, a basic knowledge of machine learning will be important if you will be working with data scientists and AI people. Likewise, if you’re a reporter covering AI news or a PR agency working on behalf of a company that uses machine learning, writing about the technology without knowing how it works is a bad idea
Tomi Engdahl says:
The Guardian’s GPT-3-written article misleads readers about AI. Here’s why.
https://bdtechtalks.com/2020/09/14/guardian-gpt-3-article-ai-fake-news/
Tomi Engdahl says:
Inside the strange new world of being a deepfake actor
There’s an art to being a performer whose face will never be seen.
https://www.technologyreview.com/2020/10/09/1009850/ai-deepfake-acting/
Tomi Engdahl says:
https://thorgate.eu/blog/3-ways-ai-transforms-how-we-develop-software?fbclid=IwAR1S-lnQrO3aM5QYzCkVK2wtSaST30WOKTaySm3tJJtM70olWVG9IKx4lf4
Tomi Engdahl says:
Google, Apple, and others show large language models trained on public data expose personal information
https://venturebeat.com/2020/12/16/google-apple-and-others-show-large-language-models-trained-on-public-data-expose-personal-information/
Tomi Engdahl says:
National Grid sees machine learning as the brains behind the utility business of the future https://tcrn.ch/2WGAPq0
Tomi Engdahl says:
https://www.dna.fi/yrityksille/blogi/-/blogs/kiinnostaisiko-epareilu-kilpailuetu-?utm_source=facebook&utm_medium=linkad&utm_content=artikkeli_kiinnostaisiko_epareilu_kilpailuetu_&utm_campaign=sm_jatkuva_some_21&fbclid=IwAR1Se1C2sF-mQudjcRmPOByWEmaFHqsRdsHHvZFI8X_vo_jwi8JbKGuLjmg
Aiwo Digitalin analytiikka perkaa olennaisen esiin jättimäisistä datamääristä, ja käyttöönotto on nopeaa.
Tomi Engdahl says:
https://techcrunch.com/2021/01/04/deep-science-using-machine-learning-to-study-anatomy-weather-and-earthquakes/?tpcc=ECFB2021
Tomi Engdahl says:
In a ground-breaking feat, scientists have showcased a formula that allows AI robots to keep learning and adapt to new circumstances.
https://www.iflscience.com/technology/scientists-broke-this-ai-robots-leg-then-let-it-teach-itself-to-walk-from-scratch/
Tomi Engdahl says:
Sharper signals: how machine learning is cleaning up microscopy images
Computers trained to reduce the noise in micrographs can now tackle fresh data by themselves.
https://www.nature.com/articles/d41586-021-00023-0?utm_source=fbk_nnc&utm_medium=social&utm_campaign=naturenews&sf241876352=1
Tomi Engdahl says:
Building natural trust in artificial intelligence
Two new technologies being developed by Fujitsu Laboratories are making AI more transparent and robust.
https://www.nature.com/articles/d42473-020-00352-0?utm_source=twitter&utm_medium=social&utm_campaign=bcon-NI_AI_Fujitsu&fbclid=IwAR0Y_pPT21iEveSKuyKIjnyYARaWAYnta3BF3Oqu2ckeL6ksLhdpfaahWl4