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,”
5,218 Comments
Tomi Engdahl says:
Neural Networks… On a Stick!
https://hackaday.com/2018/04/24/neural-networks-on-a-stick/
They probably weren’t inspired by [Jeff Dunham’s] jalapeno on a stick, but Intel have created the Movidius neural compute stick which is in effect a neural network in a USB stick form factor. They don’t rely on the cloud, they require no fan, and you can get one for well under $100. We were interested in [Jeff Johnson’s] use of these sticks with a Pynq-Z1. He also notes that it is a great way to put neural net power on a Raspberry Pi or BeagleBone. He shows us YOLO — an image recognizer — and applies it to an HDMI signal with the processing done on the Movidius.
Setting up the PYNQ-Z1 for the Intel Movidius Neural Compute Stick
http://www.fpgadeveloper.com/2018/04/setting-up-the-pynq-z1-for-the-intel-movidius-neural-compute-stick.html
Tomi Engdahl says:
AI Accelerating Discovery
https://semiengineering.com/ai-accelerating-discovery/
Artificial Intelligence still can’t compete with common sense, but its usefulness is growing quickly.
In early April 2018, the Materials Research Society held their spring meeting and exhibit at the Phoenix, Arizona convention center. With more than 110 symposium presentations, it was difficult to select which sessions to attend. But one forum caught my eye, “AI for Materials Development”. These days AI seems to be everywhere.
As we all speculate about the impact of AI on autonomous driving and the next killer app, Carla Gomes, Professor of Computer Science and director of the Institute for Computational Sustainability at Cornell University, is focusing on large-scale constraint-based reasoning. She pointed out that AI still can’t compete with good ol’ human common sense. Human reasoning and inference planning are still lacking in most AI systems. One of the key fundamentals of AI is building a neural network that resembles the human brain. Even with the advancements of 7nm silicon technology, this is a daunting task, not to mention the complexities of software algorithms to mimic the human thought and decision process.
But in the world of materials development, AI excels. By integrating material experimentation and AI, the discovery of new materials and the application of materials in the real world is progressing at an accelerated pace. AI is capable of developing the hypotheses and—along with robotics—is following through with new scientific discovery.
Tomi Engdahl says:
Neural Network Names Nightshades
https://hackaday.com/2018/04/26/neural-network-names-nightshades/
Neural networks are a core area of the artificial intelligence field. They can be trained on abstract data sets and be put to all manner of useful duties, like driving cars while ignoring road hazards or identifying cats in images. Recently, a biologist approached AI researcher [Janelle Shane] with a problem – could she help him name some tomatoes?
It’s a problem with a simple cause
Janelle] decided to use the char-rnn library built by [Andrej Karpathy] to do the heavy lifting. After training it on a list of over 11,000 existing tomato varieties, the neural network was then asked to strike out on its own.
Neural network-named tomatoes you won’t find at the farmer’s market
http://aiweirdness.com/post/172622965862/tomatonames
Tomi Engdahl says:
https://www.uusiteknologia.fi/2018/04/27/nokian-siilasmaa-patistaa-tekoalyyn/
Tomi Engdahl says:
AI as a Force for Good in Supply Chain Management
http://scnavigator.avnet.com/article/april-2018/ai-as-a-force-for-good-in-supply-chain-management/
As artificial intelligence (AI) moves from research to reality, new insights and increasingly automated decision making are disrupting entire supply chain networks. AI will affect everything-from raw materials sourcing to factory floor productivity. These emerging technologies will unlock business value while simultaneously revealing opportunities to better utilize the workers within those ecosystems.
Tomi Engdahl says:
Google Lowers The Artificial Intelligence Bar With Complete DIY Kits
https://hackaday.com/2018/04/26/google-lowers-the-artificial-intelligence-bar-with-complete-diy-kits/
Last year, Google released an artificial intelligence kit aimed at makers, with two different flavors: Vision to recognize people and objections, and Voice to create a smart speaker. Now, Google is back with a new version to make it even easier to get started.
The main difference in this year’s (v1.1) kits is that they include some basic hardware, such as a Raspberry Pi and an SD card.
AIY Projects: Updated kits for 2018
https://developers.googleblog.com/2018/04/aiy-projects-updated-kits-for-2018.html
Tomi Engdahl says:
An Arduino and a Roomba Become an Artificial Organism
https://www.hackster.io/Ondaweb/an-arduino-and-a-roomba-become-an-artificial-organism-3824d9
A model of neuronal functioning & synaptic modification allow replications of drives, behaviors, affects & learning (Operant Conditioning)
This is essentially a low level demonstartion of an Artificial General Intelligence system. A simple nervous system of about 20 neurons is modeled. It replicates the structure and functioning of biological nervous systems by modeling networks of neurons that detect the external world and produce actions in it.
It is not an AI agent. It is not an expert system, learning algorithm, demonstration of machine learning, or knowledge representation system.
The nervous system demonstrates numerous animal behavioral and learning phenomena including: stimulus-response reflexes, goal oriented behaviors, classical (Pavlovian) conditioning, operant conditioning in response to rewards and punishments, secondary reinforcement, and simple affect responses that are intrinsic to actions and learning.
Tomi Engdahl says:
A Chinese Factory Is Using AI To Breed 6 Billion Cockroaches Each Year
http://www.iflscience.com/plants-and-animals/a-chinese-factory-is-using-ai-to-breed-6-billion-cockroaches-each-year/
Tomi Engdahl says:
What is the TensorFlow machine intelligence platform?
https://opensource.com/article/17/11/intro-tensorflow?sc_cid=7016000000127ECAAY
Learn about the Google-developed open source library for machine learning and deep neural networks research.
Tomi Engdahl says:
AI Revives In-Memory Processors
Startup aims to ship chip late next year
https://www.eetimes.com/document.asp?doc_id=1333238
Startups, corporate giants, and academics are taking a fresh look at a decade-old processor architecture that may be just the thing ideal for machine learning. They believe that in-memory computing could power a new class of AI accelerators that could be 10,000 times faster than today’s GPUs.
The processors promise to extend chip performance at a time when CMOS scaling has slowed and deep-learning algorithms demanding dense multiply-accumulate arrays are gaining traction. The chips, still more than a year from commercial use, also could be vehicles for an emerging class of non-volatile memories.
Tomi Engdahl says:
Bill Jia / Facebook Code:
Facebook debuts PyTorch 1.0, which optimizes production performance and improves model compatibility with other AI frameworks, coming to beta in a few months — The path for taking AI development from research to production has historically involved multiple steps and tools …
Announcing PyTorch 1.0 for both research and production
https://code.facebook.com/posts/172423326753505/
Jerome Pesenti / Facebook Code:
Summary of Facebook’s AI-related announcements at F8: PyTorch 1.0 open source framework unveiled, expansion of Open Neural Network Exchange, updates on research
http://code.facebook.com/posts/372833966539527
Tomi Engdahl says:
Guy Rosen / Facebook:
Facebook says it can currently use its advances in AI to flag and remove content before it’s even been reported
F8 2018: Using Technology to Remove the Bad Stuff Before It’s Even Reported
https://newsroom.fb.com/news/2018/05/removing-content-using-ai/
But advances in technology, including in artificial intelligence, machine learning and computer vision, mean that we can now:
Remove bad content faster because we don’t always have to wait for it to be reported. In the case of suicide this can mean the difference between life and death. Because as soon as our technology has identified that someone has expressed thoughts of suicide, we can reach out to offer help or work with first responders, which we’ve now done in over a thousand cases.
Get to more content, again because we don’t have to wait for someone else to find it. As we announced two weeks ago, in the first quarter of 2018, for example, we proactively removed almost two million pieces of ISIS and al-Qaeda content — 99% of which was taken down before anyone reported it to Facebook.
Increase the capacity of our review team to work on cases where human expertise is needed to understand the context or nuance of a particular situation. For instance, is someone talking about their own drug addiction, or encouraging others to take drugs?
Tomi Engdahl says:
Lucas Matney / TechCrunch:
Facebook shows how it uses billions of public Instagram photos, annotated by users with hashtags, to help train its AI-based image recognition models
Facebook is using your Instagram photos to train its image recognition AI
https://techcrunch.com/2018/05/02/facebook-is-using-your-instagram-photos-to-train-its-image-recognition-ai/
In the race to continue building more sophisticated AI deep learning models, Facebook has a secret weapon: billions of images on Instagram.
In research the company is presenting today at F8, Facebook details how it took what amounted to billions of public Instagram photos that had been annotated by users with hashtags and used that data to train their own image recognition models. They relied on hundreds of GPUs running around the clock to parse the data, but were ultimately left with deep learning models that beat industry benchmarks, the best of which achieved 85.4 percent accuracy on ImageNet.
If you’ve ever put a few hashtags onto an Instagram photo, you’ll know doing so isn’t exactly a research-grade process. There is generally some sort of method to why users tag an image with a specific hashtag; the challenge for Facebook was sorting what was relevant across billions of images.
Tomi Engdahl says:
Low-Power Play: GAP8 Weds Multicore RISC-V with Machine Learning
http://www.electronicdesign.com/embedded-revolution/low-power-play-gap8-weds-multicore-risc-v-machine-learning?code=NN8DK004&utm_rid=CPG05000002750211&utm_campaign=17027&utm_medium=email&elq2=adb98e6ec77646f683537bfeb28da18d
GreenWaves’ GAP8 brings low-power machine learning to embedded systems using an eight-core array of RISC-V processors.
Tomi Engdahl says:
Finding And Fixing ML’s Flaws
https://semiengineering.com/improving-machine-learning/
OneSpin’s CEO looks at methodologies and models for making ML more predictable and more effective.
SE: How do we make sure devices developed with machine learning behave as they’re supposed to, and how do we fix problems when they crop up?
Brinkmann: The real objective is to not have a problem in the first place. We need to build methodologies to develop these machine learning systems in the first place. Those are not established yet. There is a lot of work to be done to be able to say, ‘This is how you do it,’ and ‘This is how you prevent this type of problem.’
Tomi Engdahl says:
Hot stuff: Facebook AI gurus tout new Pytorch 1.0 framework for all
Blah, blah, speed up neural networks, something, blah blah
https://www.theregister.co.uk/2018/05/02/facebook_ai_pytorch/
Tomi Engdahl says:
Facebook’s Open-Source Go Bot Can Now Beat Professional Players
https://news.slashdot.org/story/18/05/02/2224251/facebooks-open-source-go-bot-can-now-beat-professional-players?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+Slashdot%2Fslashdot%2Fto+%28%28Title%29Slashdot+%28rdf%29%29
Google’s DeepMind isn’t the only team working to defeat professional Go players with artificial intelligence. At Facebook’s F8 developer conference today, the company announced a Go bot of its own that has now achieved professional status after winning all 14 games it played against a group of top 30 human Go players.
Facebook’s open-source Go bot can now beat professional players
https://techcrunch.com/2018/05/02/facebooks-open-source-go-bot-can-now-beat-professional-players/
“We salute our friends at DeepMind for doing awesome work,” Facebook CTO Mike Schroepfer said in today’s keynote. “But we wondered: Are there some unanswered questions? What else can you apply these tools to.” As Facebook notes in a blog post today, the DeepMind model itself also remains under wraps. In contrast, Facebook has open-sourced its bot.
Tomi Engdahl says:
Re-Engineering Humanity
https://semiengineering.com/re-engineering-humanity/
Why artificial intelligence will change our views, and why you need to know that.
Tomi Engdahl says:
Preparing For AI
https://semiengineering.com/preparing-for-ai/
The public policy implications of intelligent systems are enormous.
Suppose an autonomous car is coming up an on-ramp onto a bridge. The ramp is fine, but the bridge is icy, and there’s an overturned bus full of children blocking several lanes.
Children are evacuating through the windows and milling around on the pavement. There isn’t time to stop, even with the better-than-human reaction time an autonomous car might have. Swerving to one side might send the car off the bridge, to the other side might send it into a retaining wall, potentially killing or injuring the passengers in either case. The car is fully autonomous, with no ability for a human to take control, even if there were more time. What should it do?
Tomi Engdahl says:
Espoo pestasi tekoälyn sote-työntekijäksi – millaisin tuloksin?
https://www.is.fi/kotimaa/art-2000005654424.html
Tuleeko tekoälystä joskus tietoinen olento? Näin huippututkija vastaa
https://www.is.fi/tiede/art-2000005654428.html
Tomi Engdahl says:
”Uusi sähkö” – tulevasta mullistuksesta saa nyt rautalankaa ilmaiseksi
https://www.is.fi/digitoday/art-2000005665764.html
http://www.elementsofai.com/
Tomi Engdahl says:
Google Kubeflow, machine learning for Kubernetes, begins to take shape
https://techcrunch.com/2018/05/04/google-kubeflow-machine-learning-for-kubernetes-begins-to-take-shape/?utm_source=tcfbpage&utm_medium=feed&utm_campaign=Feed%3A+Techcrunch+%28TechCrunch%29&sr_share=facebook
Ever since Google created Kubernetes as an open source container orchestration tool, it has seen it blossom in ways it might never have imagined. As the project gains in popularity, we are seeing many adjunct programs develop. Today, Google announced the release of version 0.1 of the Kubeflow open source tool, which is designed to bring machine learning to Kubernetes containers.
While Google has long since moved Kubernetes into the Cloud Native Computing Foundation, it continues to be actively involved, and Kubeflow is one manifestation of that.
Tomi Engdahl says:
Scott Bay / Wired:
How filmmakers could use machine learning, trained on data from prior box office hits, to predict a movie’s success by understanding what motivates moviegoers
Could Artificial Intelligence Predict the Next Avengers: Infinity War?
https://www.wired.com/story/artificial-intelligence-box-office-predictions
Some movies are obvious hits. Like, for example, Avengers: Infinity War, which made a record-breaking $258 million at the domestic box office last weekend, filling seats and the pockets of Marvel Studios parent company Disney. But not every summer—or spring, or fall—blockbuster has the benefit of 10 years and 18 movies of built-up audience goodwill. So while the Mouse House knew they had a potentially earth-shattering hit on their hands well before opening night, other studios trying to catch up have no way of predicting whether their latest attempts to hit big will do so.
Actually, they might. Machine learning is everywhere, and artificial intelligence is no longer just a Spielberg-Kubrick collaboration. These days, Amazon can practically anticipate when you might need toilet paper and Netflix can predict your next binge, so it only seems natural that Hollywood will start using AI to predict the next big blockbuster, or at least improve its chances of becoming one.
Tomi Engdahl says:
Cade Metz / New York Times:
A look at universities’ struggles to retain AI researchers as Facebook opens new AI labs in Seattle and Pittsburgh after New York, Paris, and Montreal
Facebook Adds A.I. Labs in Seattle and Pittsburgh, Pressuring Local Universities
https://www.nytimes.com/2018/05/04/technology/facebook-artificial-intelligence-researchers.html
At a conference in Silicon Valley this week, Mark Zuckerberg, Facebook’s chief executive, vowed that his company would “keep building” despite a swirl of questions around the way it has dealt with misinformation and the personal data of its users.
That is certainly true in the important area of artificial intelligence, which Mr. Zuckerberg says can help the social media giant deal with some of those problems.
Facebook is opening new A.I. labs in Seattle and Pittsburgh, after hiring three A.I. and robotics professors from the University of Washington and Carnegie Mellon University. The company hopes these seasoned researchers will help recruit and train other A.I. experts in the two cities, Mike Schroepfer, Facebook’s chief technology officer, said in an interview.
As it builds these labs, Facebook is adding to pressure on universities and nonprofit A.I. research operations, which are already struggling to retain professors and other employees.
Tomi Engdahl says:
Nicole Jao / TechNode:
UBTECH, a Shenzhen-based intelligent humanoid robots maker, raises $820M Series C led by Tencent, bringing the startup’s valuation to about $5B
Ubtech raises 820 million in Series C funding round
https://technode.com/2018/05/03/ubtech-820-million-usd-series-c/
Tomi Engdahl says:
Mary Jo Foley / ZDNet:
Microsoft announces that Project Brainwave, its system for running AI models with FPGA chips, is now available in limited preview on Azure
Microsoft opens its ‘BrainWave’ AI-on-FPGA service to external testers
https://www.zdnet.com/article/microsoft-opens-its-brainwave-ai-on-fpga-service-to-external-testers/
Microsoft’s project to run fast AI tasks on FPGAs in Azure is starting to come to fruition, with testers starting to get access to the first pieces now.
Tomi Engdahl says:
AI Gets New Benchmark
Google, Baidu spearhead MLPerf
https://www.eetimes.com/document.asp?doc_id=1333246
Google and Baidu collaborated with researchers at Harvard and Stanford to define a suite of benchmarks for machine learning. So far, AMD, Intel, two AI startups, and two other universities have expressed support for MLPerf, an initial version of which will be ready for use in August.
Today’s hardware falls far short of running neural-networking jobs at the performance levels desired. A flood of new accelerators are coming to market, but the industry lacks ways to measure them.
To fill the gap, the first release of MLPerf will focus on training jobs on a range of systems from workstations to large data centers, a big pain point for web giants such as Baidu and Google. Later releases will expand to include inference jobs, eventually extended to include ones run on embedded client systems.
Tomi Engdahl says:
AI Revives In-Memory Processors
Startup aims to ship chip late next year
https://www.eetimes.com/document.asp?doc_id=1333238
Startups, corporate giants, and academics are taking a fresh look at a decade-old processor architecture that may be just the thing ideal for machine learning. They believe that in-memory computing could power a new class of AI accelerators that could be 10,000 times faster than today’s GPUs.
The processors promise to extend chip performance at a time when CMOS scaling has slowed and deep-learning algorithms demanding dense multiply-accumulate arrays are gaining traction. The chips, still more than a year from commercial use, also could be vehicles for an emerging class of non-volatile memories.
Startup Mythic (Austin, Texas) aims to compute neural-network jobs inside a flash memory array, working in the analog domain to slash power consumption. It aims to have production silicon in late 2019, making it potentially one of the first to market of the new class of chips.
“Most of us in the academic community believe that emerging memories will become an enabling technology for processor-in-memory,” said Suman Datta, who chairs the department of electrical engineering at Notre Dame. “Adoption of the new non-volatile memories will mean creating new usage models, and in-memory processing is a key one.”
Datta notes that several academics attempted to build such processors in the 1990s. Designs such as the EXECUBE, IRAM, and FlexRAM “fizzled away, but now, with the emergence of novel devices such as phase-change memories, resistive RAM, and STT MRAM and strong interest in hardware accelerators for machine learning, there is a revitalization of the field … but most of the demonstrations are at a device or device-array level, not a complete accelerator, to the best of my knowledge.”
Tomi Engdahl says:
https://www.uusiteknologia.fi/2018/05/08/alykas-pilvi-vaatii-uusia-sovellustyokaluja/
Tomi Engdahl says:
http://kampanja.pwc.fi/julkaisu/artificial-intelligence-in-finland
Tomi Engdahl says:
TensorFlow in your Browser
https://hackaday.com/2018/04/16/tensorflow-in-your-browser/
If you want to explore machine learning, you can now write applications that train and deploy TensorFlow in your browser using JavaScript. We know what you are thinking. That has to be slow. Surprisingly, it isn’t, since the libraries use Graphics Processing Unit (GPU) acceleration. Of course, that assumes your browser can use your GPU. There are several demos available, include one where you train a Pac Man game to respond to gestures in your webcam to control the game.
TensorFlow.js
https://js.tensorflow.org/
A WebGL accelerated, browser based JavaScript library for training and deploying ML models.
Tomi Engdahl says:
White House Says It Won’t Use Heavy Hand With AI Rules
https://www.bloomberg.com/news/articles/2018-05-10/white-house-tells-google-goldman-it-won-t-rush-to-regulate-ai
The White House unveiled a hands-off regulatory approach to foster the development of artificial intelligence at a gathering of more than 40 companies in Washington Thursday.
A top White House technology adviser, Michael Kratsios, told representatives of companies including Alphabet Inc.’s Google, Facebook Inc., Goldman Sachs Group Inc. and Boeing Co. that they’ll have the greatest possible latitude to develop AI, according to a copy of his remarks that was provided to Bloomberg.
“We didn’t cut the lines before Alexander Graham Bell made the first telephone call,” Kratsios said in his prepared remarks. “We didn’t regulate flight before the Wright Brothers took off at Kitty Hawk.”
The Trump administration’s free-market approach to AI comes as the technology sector is facing increasing calls for regulation from lawmakers and consumer groups in addition to concerns specific to AI development that include issues around bias in data use and fierce competition with China.
Intel Corp. Chief Executive Officer Brian Krzanich
“Privacy, cybersecurity, ethics, and potential employment impact are all worthy of careful analysis,”
Tomi Engdahl says:
Talking Neural Nets
https://hackaday.com/2016/12/03/talking-neural-nets/
Speech synthesis is nothing new, but it has gotten better lately. It is about to get even better thanks to DeepMind’s WaveNet project. The Alphabet (or is it Google?) project uses neural networks to analyze audio data and it learns to speak by example. Unlike other text-to-speech systems, WaveNet creates sound one sample at a time and affords surprisingly human-sounding results.
WaveNet: A Generative Model for Raw Audio
https://deepmind.com/blog/wavenet-generative-model-raw-audio/
Tomi Engdahl says:
Mobile Machine Learning Hardware At Arm
A systems-on-chip (SoC) perspective.
https://semiengineering.com/mobile-machine-learning-hardware-at-arm/
Machine learning is playing an increasingly significant role in emerging mobile application domains such as AR/VR, ADAS, etc. Accordingly, hardware architects have designed customized hardware for machine learning algorithms, especially neural networks, to improve compute efficiency. However, machine learning is typically just one processing stage in complex end-to-end applications, which involve multiple components in a mobile Systems-on-a-chip (SoC). Focusing on just ML accelerators loses bigger optimization opportunity at the system (SoC) level. This paper argues that hardware architects should expand the optimization scope to the entire SoC.
https://arxiv.org/pdf/1801.06274.pdf
Tomi Engdahl says:
The First 8 Jobs Robots Will Take from Humans
https://www.eeweb.com/profile/jameswarner/articles/the-first-8-jobs-robots-will-take-from-humans
It’s futile to resist technological big data advancement solutions and changes as automation is set to provide increasing benefits.
A recent study suggested that artificially intelligent (AI) robots will take over 800 million jobs by 2030. Does this mean your job is at risk and that a robot will soon be performing the tasks you’re performing?
AI robots have already revamped sectors such as industrial manufacturing; now they’re stepping towards food production and restaurant kitchens. AI technology today is not only making these operations more successful, but it’s intended to replace the human mind. In fact, the list of jobs AI robots will take over in the future is expanding day by day. Here are the top eight:
Recruiters
Salespeople
Writers and Journalists
Accountants
Factory Labor
Doctors
Lawyers
Receptionists
Many big companies have already started adopting AI
AI Robot as a recruiter: A renowned food and drinks company with gross revenue of over one billion dollars has started using AI software to identify the right candidates and even interview those candidates. AI robots will not only organize telephonic interviews with the probable candidates but are also programmed to reply to queries about the job vacancies. At this time, a single AI robot has the potential to interview 1,500 candidates in just nine hours!
AI Robots to manage the “just walk out” outlets: A popular company recently opened a convenience store where customers can easily pick the products they want to purchase, scan their phones, and walk out without wasting time at the billing counter. The store is powered by a new generation of AI machines that have the capability to sense which customer is buying what and how much is to be paid. As soon as the customers walk out of the store, they receive notification of the amount deducted from their account.
Tomi Engdahl says:
Deep learning with synthetic data will democratize the tech industry
https://techcrunch.com/2018/05/11/deep-learning-with-synthetic-data-will-democratize-the-tech-industry/?utm_source=tcfbpage&utm_medium=feed&utm_campaign=Feed%3A+Techcrunch+%28TechCrunch%29&sr_share=facebook
AdChoices
Deep learning with synthetic data will democratize the tech industry
Evan Nisselson
11 hours ago
Retro Robots
Evan Nisselson
Contributor
Evan Nisselson is a partner at LDV Capital.
More posts by this contributor
The war over artificial intelligence will be won with visual data
The visual data sets of images and videos amassed by the most powerful tech companies have been a competitive advantage, a moat that keeps the advances of machine learning out of reach from many. This advantage will be overturned by the advent of synthetic data.
The world’s most valuable technology companies, such as Google, Facebook, Amazon and Baidu, among others, are applying computer vision and artificial intelligence to train their computers. They harvest immense visual data sets of images, videos and other visual data from their consumers.
These data sets have been a competitive advantage for major tech companies, keeping out of reach from many the advances of machine learning and the processes that allow computers and algorithms to learn faster.
Tomi Engdahl says:
A look at open source image recognition technology
https://opensource.com/article/18/5/state-of-image-recognition?sc_cid=7016000000127ECAAY
Image recognition technology promises great potential in areas from public safety to healthcare.
A PhD student from Louisiana State University, Shayan Shams, had set up a large monitor displaying a webcam image. Overlaid on the image were colored boxes with labels. As I looked closer, I realized the labels identified objects on a table.
Of course, I had to play with it.
When I asked Shams about the project, I was surprised to learn that he did not need to write any code to create it—the entire thing came together from open software and data. Shams used the Common Objects in Context (COCO) dataset for object recognition, reducing unnecessary classes to enable it to run on less powerful hardware. “Detecting some classes, such as airplane, car, bus, truck, and so on in the SC exhibition hall [was] not necessary,” he explained. To do the actual detection, Shams used the You Only Look Once (YOLO) real-time object detection system.
Tomi Engdahl says:
Intel Starts R&D Effort in Probabilistic Computing for AI
https://spectrum.ieee.org/automaton/robotics/artificial-intelligence/intel-starts-rd-effort-in-probabilistic-computing-for-ai
Intel announced today that it is forming a strategic research alliance to take artificial intelligence to the next level. Autonomous systems don’t have good enough ways to respond to the uncertainties of the real world, and they don’t have a good enough way to understand how the uncertainties of their sensors should factor into the decisions they need to make. According to Intel CTO Mike Mayberry the answer is “probabilistic computing”, which he says could be AI’s next wave.
Tomi Engdahl says:
The Extraordinary Link Between Deep Neural Networks and the Nature of the Universe
https://www.technologyreview.com/s/602344/the-extraordinary-link-between-deep-neural-networks-and-the-nature-of-the-universe/?utm_source=facebook.com&utm_medium=social&utm_campaign=owned_social
Nobody understands why deep neural networks are so good at solving complex problems. Now physicists say the secret is buried in the laws of physics.
In the last couple of years, deep learning techniques have transformed the world of artificial intelligence. One by one, the abilities and techniques that humans once imagined were uniquely our own have begun to fall to the onslaught of ever more powerful machines. Deep neural networks are now better than humans at tasks such as face recognition and object recognition.
But there is a problem. There is no mathematical reason why networks arranged in layers should be so good at these challenges. Mathematicians are flummoxed. Despite the huge success of deep neural networks, nobody is quite sure how they achieve their success.
Tomi Engdahl says:
https://yle.fi/aihe/artikkeli/2018/05/12/algoritmi-ennustaa-sairastumisesi-mutta-kenelle
Tomi Engdahl says:
https://www.tivi.fi/Kaikki_uutiset/nvidia-pomo-huolestui-tekoaly-suomen-puolesta-pitaisi-olla-enemman-laskentakapasiteettia-6724425
Tomi Engdahl says:
AI Systems Should Debate Each Other To Prove Themselves, Says OpenAI
https://science.slashdot.org/story/18/05/12/2134209/ai-systems-should-debate-each-other-to-prove-themselves-says-openai?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+Slashdot%2Fslashdot%2Fto+%28%28Title%29Slashdot+%28rdf%29%29
To make AI easier for humans to understand and trust, researchers at the [Elon Musk-backed] nonprofit research organization OpenAI have proposed training algorithms to not only classify data or make decisions, but to justify their decisions in debates with other AI programs in front of a human or AI judge. In an experiment described in their paper
Why Scientists Think AI Systems Should Debate Each Other
Researchers at Musk-backed OpenAI propose a new way to tell if an AI is making the right decisions, and they released an online game to demonstrate the theory.
https://www.fastcompany.com/40569116/why-scientists-think-ai-systems-should-debate-each-other
Tomi Engdahl says:
https://techcrunch.com/2018/05/12/what-do-ai-and-blockchain-mean-for-the-rule-of-law/?utm_source=tcfbpage&utm_medium=feed&utm_campaign=Feed%3A+Techcrunch+%28TechCrunch%29&sr_share=facebook
Tomi Engdahl says:
Adobe CTO leads company’s broad AI bet
https://techcrunch.com/2018/05/12/adobe-cto-leads-companys-broad-ai-bet/?sr_share=facebook&utm_source=tcfbpage
Tomi Engdahl says:
Making machine learning work for you.
https://pages.awscloud.com/AI_ML_eBook_download.html?sc_channel=psm&sc_campaign=emea-ai-ml&sc_medium=FB_Leads_FMM_EMEAAIML_042418&sc_publisher=fb&sc_country=mult&sc_geo=emea&sc_category=mult&sc_outcome=field&trkCampaign=emea-ai-ml&trk=psm_fb_followers_gated_NORDICS_emea-ai-ml
Bring AI capabilities to your business.
Whether you’re looking to improve learning techniques, solve business problems quickly, or efficiently analyze and track data patterns, it’s time to embrace AI. With data volumes continuing to rise, AI is getting smarter and learning faster, enabling organizations such as yours to leverage more-advanced Machine Learning (ML) and Deep Learning (DL) solutions.
This paper explores the three layers of the AI stack on Amazon Web Services (AWS),
Tomi Engdahl says:
Artificial Neural Nets Grow Brainlike Navigation Cells
By
JOHN RENNIE
May 9, 2018
https://www.quantamagazine.org/artificial-neural-nets-grow-brainlike-navigation-cells-20180509/
Faced with a navigational challenge, neural networks spontaneously evolved units resembling the grid cells that help living animals find their way.
Tomi Engdahl says:
https://www.tivi.fi/Kaikki_uutiset/googlella-ovi-paukkuu-vakea-irtisanoutunut-salaisen-projektin-paljastumisen-jalkeen-6724755
Tomi Engdahl says:
Google Employees Resign in Protest Against Pentagon Contract
https://gizmodo.com/google-employees-resign-in-protest-against-pentagon-con-1825729300
It’s been nearly three months since many Google employees—and the public—learned about the company’s decision to provide artificial intelligence to a controversial military pilot program known as Project Maven, which aims to speed up analysis of drone footage by automatically classifying images of objects and people. Now, about a dozen Google employees are resigning in protest over the company’s continued involvement in Maven.
The resigning employees’ frustrations range from particular ethical concerns over the use of artificial intelligence in drone warfare to broader worries about Google’s political decisions—and the erosion of user trust that could result from these actions. Many of them have written accounts of their decisions to leave the company, and their stories have been gathered and shared in an internal document, the contents of which multiple sources have described to Gizmodo.
Tomi Engdahl says:
New Deep Learning Processors, Embedded FPGA Technologies, SoC Design Solutions
#55DAC: Must-see technologies in the DAC 2018 IP track.
https://semiengineering.com/new-deep-learning-processors-embedded-fpga-technologies-soc-design-solutions/
Tomi Engdahl says:
AI may be last growth driver for Taiwan IT players, says Quanta chairman
https://www.digitimes.com/news/a20180514PD210.html
Ongoing AI (artificial intelligence) applications may provide the last wave of huge growth in business opportunities for Taiwan IT firms, as they have done little about the development of next-generation quantum computer and would have greater difficulty getting into the new field, Quanta Computer chairman Barry Lam has said.
Lam said that he has never felt so optimistic about the extremely large business growth potential AI will bring to his company and Taiwan’s IT industry, urging IT players to seize and cash in on the golden opportunity.
Lam said Quanta can precisely feel the impulse of the AI development through the production of servers and has developed new servers to allow fast real-time computing algorithms to support a wide variety of AI applications including smart vehicles, smart lawyers, smart police, smart healthcare, and smart cities, all involving great business opportunities.