There’s an old saying that the first fifty years of the car industry were about creating car companies and working out what cars should look like, and the second fifty years were about what happened once everyone had a car - they were about McDonalds and Walmart, suburbs and the remaking of the world around the car, for good and of course bad. The innovation in cars became everything around the car. One could suggest the same today about smartphones - now the innovation comes from everything else that happens around them.
Farid Mheir's insight:
WHY IT MATTERS: this post is a bit theoretical for someone not bathing in digital transformation all day as I do (mostly). It explores what could come next but more importantly stresses an important fact: we now have a new platform - the mobile phone - in our pockets that has the potential to transform our world. We have started to see this with eCommerce - physical retail stores become irrelevant and we can price compare from anywhere - but we have yet to see the real impact of eCommerce+mobile+cloud+5G. THIS is what I think about most days...
A curated list of awesome Deep Learning tutorials, projects and communities.
Farid Mheir's insight:
WHY IT MATTERS: this list is most useful as it contains hundreds of links and references curated by the community. What is also noteworthy is how it is maintained and delivered: using gitHub, the solution to manage open source projects. It serves for me as a guide about what is possible when using crowdsourcing.
A gentle introduction to the principles behind neural networks, including back propagation.
Farid Mheir's insight:
WHY IT MATTERS: this is the basic of how artificial intelligence works. Everyone should understand this to gain little bit more understanding of the digital world around us.
The AI Index is an open, not-for-profit project to track activity and progress in AI. It aspires to be a comprehensive resource of data and analysis for policymakers, researchers, executives, journalists and others to rapidly develop intuitions about the complex field of AI. AI is the new electricity, and is transforming multiple industries. The AI Index will help current generations track and navigate this societal transformations. It will also help future generations look back and understand the AI's rise.
Farid Mheir's insight:
WHY IT MATTERS: AI evolves rapidly and this report presents the state of the world (mostly the US at this time) as it relates to AI. Must read.
This guide is for anyone who is curious about machine learning but has no idea where to start. I imagine there are a lot of people who tried reading the wikipedia article, got frustrated and gave up wishing someone would just give them a high-level explanation. That’s what this is. The goal is be accessible to anyone — which means that there’s a lot of generalizations. But who cares? If this gets anyone more interested in ML, then mission accomplished.
Farid Mheir's insight:
WHY IT MATTERS: this series of article present the fundamental technologies behind AI in a very simple and easy to understand way.
What AI can — and cannot — do for your organization.
Farid Mheir's insight:
WHY IT MATTERS: advances in computer vision due to deep learning algorithms has opened a world of possibilities for companies. Read this article and accompanying HBR posts. Great series.
Find out how you can make use of Google's machine learning expertise to power your applications. Google Cloud Platform (GCP) offers five APIs that provide access to pre-trained machine learning models with a single API call: Google Cloud Vision API, Cloud Speech API, Cloud Natural Language API, Cloud Translation API and Cloud Video API. Using these APIs, you can focus on adding new features to your app rather than building and training your own custom models. In this session we'll share an overview of each API and dive into code with some live demos.
Farid Mheir's insight:
WHY IT MATTERS: Google is putting AI in everything they do and makes it really easy for developers to embed AI into their website, app, or tool. To quickly test the vision API and be amazed go here: https://cloud.google.com/vision
Creating and maintaining a platform for reliably producing and deploying machine learning models requires careful orchestration of many components—-a learner for generating models based on training data, modules for analyzing and validating both data as well as models, and finally infrastructure for serving models in production. This becomes particularly challenging when data changes over time and fresh models need to be produced continuously. Unfortunately, such orchestration is often done ad hoc using glue code and custom scripts developed by individual teams for specific use cases, leading to duplicated effort and fragile systems with high technical debt. We present the anatomy of a general-purpose machine learning platform and one implementation of such a platform at Google. By integrating the aforementioned components into one platform, we were able to standardize the components, simplify the platform configuration, and reduce the time to production from the order of months to weeks, while providing platform stability that minimizes service disruptions. We present the case study of one deployment of the platform in the Google Play app store, where the machine learning models are refreshed continuously as new data arrive. Deploying the platform led to reduced custom code, faster experiment cycles, and a 2% increase in app installs resulting from improved data and model analysis.
Farid Mheir's insight:
WHY THIS MATTERS
This article explains the different components that Google had to create in order to make their machine learning tensorFlow system production grade and ensure it could support a software engineering development process. We are entering a brave new world of new tools.
A new idea is helping to explain the puzzling success of today’s artificial-intelligence algorithms — and might also explain how human brains learn.
Farid Mheir's insight:
WHY IT MATTERS
This article explains how deep learning neural networks learn to generalize and find patterns in their inputs by initially finding patterns then compressing to forget irrelevant elements and generalize what they've "learned". Well illustrated and relatively simple to understand.
Explore TensorFlow Playground demos. See how they explain the mechanism and power of neural networks, which extract hidden insights and complex patterns.
Farid Mheir's insight:
WHY IT MATTERS
Understanding how neural networks work is difficult. This simple tool helps you visualize how they work and change parameters to see their impact on the network quality.
Instacart is saving minutes per delivery by sorting shopping lists using deep learning. Emojis help to define the problem and outline both a simple and a more complex deep learning architecture.
Farid Mheir's insight:
WHY IT MATTERS
Grocery shopping is often tedious, as most grocery stores have a slightly different layout. Instacart sends human shoppers to pick orders placed online on their website. Mapping every single one of their thousands of stores is not an option. Instead, they have trained a neural networks on millions of orders to predict the best sort order for their shopping lists. And in the meantime save precious minutes in the order picking process.
Teachable Machine, an experiment that makes it easier for anyone to explore machine learning. Teach a machine using your camera – live in the browser, no coding required.
Farid Mheir's insight:
WHY THIS MATTERS
You have to try this to better understand how AI works. Moreover, the fact that the network runs on your browser is, by itself, quite the testament to the power of deep learning and machine learning to transform every application and device we own. Now go play!
Tech giants including Baidu and Google spent between $20B to $30B on AI in 2016, with 90% of this spent on R&D and deployment, and 10% on AI acquisitions. AI investment has turned into a race for patents and intellectual property (IP) among the world’s leading tech companies.
Farid Mheir's insight:
WHY IT MATTERS
This article from McKinsey provides an overview of the investments, improvements and priorities in AI and machine learning as of mid 2017.
ARK Invest aims to identify large-scale investment opportunities resulting from technological change. We believe innovation is key to growth. From a broad spectrum of disruptive innovations, Big Ideas represents our annual breakout of technologies that we believe will accelerate significantly in the months ahead.
Farid Mheir's insight:
WHY THIS IS IMPORTANT
Investment firms tend to be very traditional in their approach. This one appears to be embracing the digital transformation and making very insightful research that - from what they claim - guides their investments. Great idea. Now we'll see if their assumptions prove right and generate more returns than other investment guidance. I have a feeling it will.
Topics covered: 3D printing, additive manufacturing, artificial intelligence, automation, autonomous taxis, autonomous vehicles, bitcoin, blockchain, CRISPR, cryptocurrency, deep learning, dna sequencing, fintech, genomics, machine learning, mobile, robotics, shared autonomous vehicles
Have you looked at Google Trends? It’s pretty cool — you enter some keywords and see how Google Searches of that term vary through time. I thought — hey, I happen to have this arxiv-sanity database of 28,303 (arxiv) Machine Learning papers over the last 5 years, so why not do something similar and take a look at how Machine Learning research has evolved over the last 5 years?
Farid Mheir's insight:
WHY THIS IS IMPORTANT
More research = more discoveries so we can expect even more practical uses of machine intelligence in the coming years...
Thanks to machine learning, it’s becoming easy to generate realistic video, and to impersonate someone.
Farid Mheir's insight:
WHY THIS IS IMPORTANT
If you thought fake news was bad, well an era of fake audio and news clips is about to dawn on us. We need regulations to govern this otherwise there will be consequences in the political, business and personal fronts for sure...
Also see this other blog post on the subject, as it related to fake actors in movies, a much less troublesome use of the technology: http://sco.lt/8Eglqj
This field is moving fast if you compare this to a blog post of January 2016: http://sco.lt/85op5l
Systems able to recognize sounds familiar to human listeners have a wide range of applications, from adding sound effect information to automatic video captions, to potentially allowing you to search videos for specific audio events. Building Deep Learning systems to do this relies heavily on both a large quantity of computing (often from highly parallel GPUs), and also – and perhaps more importantly – on significant amounts of accurately-labeled training data. However, research in environmental sound recognition is limited by currently available public datasets. In order to address this, we recently released AudioSet, a collection of over 2 million ten-second YouTube excerpts labeled with a vocabulary of 527 sound event categories, with at least 100 examples for each category.
Farid Mheir's insight:
WHY THIS IS IMPORTANT
Training data is crucial for machine intelligence. This kind of stuff will help researchers improve their AI solutions. What is most important is the data has been crowdsourced from youTube videos. It used to be we were the product, not it seems we are the data...
70% of retail decision makers globally are ready to adopt the Internet of Things to improve customer experiences.
73% of retailers rate managing big data as important or business-critical to their operations.
78% of retailers say it is important or business-critical to integrate e-commerce and in-store experiences, so an omnichannel experience is delivered to every customer.
87% of retailers will deploy mobile point-of-sale (MPOS) devices by 2021, enabling them to scan and accept credit or debit payments anywhere in the store.
90% of retailers will implement buy online, pickup in store by 2021.
Farid Mheir's insight:
WHY THIS IS IMPORTANT
Based on a detailed report, this article provides a chockful of stats about investments in certain technologies. Although based on a survey, it confirms certain trends that are widely accepted - so there should be no surprises for readers of this blog.
YouTube-BoundingBoxes is a large-scale data set of video URLs with densely-sampled high-quality single-object bounding box annotations.
The data set consists of approximately 380,000 15-20s video segments extracted from 240,000 different publicly visible YouTube videos, automatically selected to feature objects in natural settings without editing or post-processing, with a recording quality often akin to that of a hand-held cell phone camera.
All these video segments were human-annotated with high precision classifications and bounding boxes at 1 frame per second.
Farid Mheir's insight:
WHY THIS IS IMPORTANT
Training artificial intelligence systems using deep learning algorithms requires large datasets where the elements you want to recognize - cats, planes umbrellas - are already identified.
This dataset from YouTube contains 380000 videos with over 10.5M annotations to help train systems of the future. We have to remember that the more data is available for training the better the system will perform. Nevertheless, this is interesting as is shows a little bit of the behind the scene in AI and helps understand a bit more how good AI systems will be - or when they may fail.
Maybe you’ve downloaded TensorFlow and you’re ready to get started with some deep learning?
But then you wonder: What the hell is a tensor?
Perhaps you looked it up on Wikipedia and now you’re more confused than ever. Maybe you found this NASA tutorial and still have no idea what it’s talking about?
The problem is most guides talk about tensors as if you already understand all the terms they’re using to describe the math. Have no fear! I hated math as a kid, so if I can figure it out, you can too! We just have to explain everything in simpler terms.
Farid Mheir's insight:
If you want to start nibbling on artificial intelligence, start by reading this tutorial. Simple, yet complete enough to walk you through the most important stuff.
These principles were developed in conjunction with the 2017 Asilomar conference (videos here), through the process described here. Artificial intelligence has already provided beneficial tools that are used every day by people around the world. Its continued development, guided by the following principles, will offer amazing opportunities to help and empower people in the decades …
Raffi Khatchadourian on Nick Bostrom, an Oxford philosopher who asks whether inventing artificial intelligence will bring us utopia or destruction.
Farid Mheir's insight:
Perfect Sunday morning reading which is guaranteed to make you reflect and ponder for the next weeks. The article is a typical New Yorker one, very well researched and written. So captivating that it got me to start reading the book which appears to be as captivating and surprisingly easy to read and understand. I love those finds and have the feeling this book will be the best complement to "Singularity is Near" and "On Intelligence" that I wrote about in the past.
Perfect Sunday morning reading which is guaranteed to make you reflect and ponder for the next weeks. The article is a typical New Yorker one, very well researched and written. So captivating that it got me to start reading the book which appears to be as captivating and surprisingly easy to read and understand. I love those finds and have the feeling this book will be the best complement to "Singularity is Near" and "On Intelligence" that I wrote about in the past.
Raffi Khatchadourian on Nick Bostrom, an Oxford philosopher who asks whether inventing artificial intelligence will bring us utopia or destruction.
Farid Mheir's insight:
Perfect Sunday morning reading which is guaranteed to make you reflect and ponder for the next weeks. The article is a typical New Yorker one, very well researched and written. So captivating that it got me to start reading the book which appears to be as captivating and surprisingly easy to read and understand. I love those finds and have the feeling this book will be the best complement to "Singularity is Near" and "On Intelligence" that I wrote about in the past.
Perfect Sunday morning reading which is guaranteed to make you reflect and ponder for the next weeks. The article is a typical New Yorker one, very well researched and written. So captivating that it got me to start reading the book which appears to be as captivating and surprisingly easy to read and understand. I love those finds and have the feeling this book will be the best complement to "Singularity is Near" and "On Intelligence" that I wrote about in the past.
To get content containing either thought or leadership enter:
To get content containing both thought and leadership enter:
To get content containing the expression thought leadership enter:
You can enter several keywords and you can refine them whenever you want. Our suggestion engine uses more signals but entering a few keywords here will rapidly give you great content to curate.
WHY IT MATTERS: this post is a bit theoretical for someone not bathing in digital transformation all day as I do (mostly). It explores what could come next but more importantly stresses an important fact: we now have a new platform - the mobile phone - in our pockets that has the potential to transform our world. We have started to see this with eCommerce - physical retail stores become irrelevant and we can price compare from anywhere - but we have yet to see the real impact of eCommerce+mobile+cloud+5G. THIS is what I think about most days...