May 22, 2019
Category archive


Facebook releases a trio of maps to aid with fighting disease outbreaks

in AI/Artificial Intelligence/Delhi/Facebook/Health/India/Politics/public data/Public health/Social/TC/vaccines by

Facebook this morning announced a new initiative focused on using its data and technologies to help nonprofit organizations and universities working in public health better map the spread of infectious diseases around the world. Specifically, the company is introducing three new maps: population density maps with demographic estimates, movement maps, and network coverage maps. These, says Facebook, will help the health partners to understand where people live, how they’re moving, and if they have connectivity — all factors that can aid in determining how to respond to outbreaks, and where supplies should be delivered.

As Facebook explained, health organizations rely on information like this when planning public health campaigns. But much of the information they rely on is outdated, like older census data. In addition, information from more remote communities can be scarce.

By combining the new maps with other public health data, Facebook believes organizations will be better equipped to address epidemics.

The new high-resolution population density maps will estimate the number of people living within 30-meter grid tiles, and provide insights on demographics, including the number of children under five, the number of women of reproductive age, as well as young and elderly populations. These maps aren’t built using Facebook data, but are instead built by using Facebook’s A.I. capabilities with satellite imagery and census information.

Movement maps, meanwhile, track aggregate data about Facebook users’ movements via their mobile phones (when location services are enabled). At scale, health partners can combine this with other data to predict where other outbreaks may occur next.

Above: movement maps of London and surrounding areas

And network coverage maps help to show where people can be reached with online messages — like those alerting to vaccination days or other health-related communications.

Of course, it’s hard to not overlook the irony involved of Facebook data and tech being used to help with outbreaks that, in some cases, Facebook had a hand in creating. The company for years allowed the spread of vaccine misinformation to spread across its network. Irresponsibly, it also allowed Facebook Pages and Groups that promoted vaccine skepticism or outright false claims to rank at the top of Facebook searches for related terms, like “vaccines.”

After time passed, it’s not surprising to find this spread of misinformation on Facebook and elsewhere resulted in measles outbreaks in the U.S. and abroad — a disease that had been eliminated as a major U.S. public health threat around 20 years ago. And despite repeated debunkings of false claims about the link between measles and autism by scientists, Facebook users — now feeling equipped by the internet to be experts on anything they can type into a search box — continued to spread misinformation at scale.

Once looped into the anti-vax communities, Facebook’s algorithms only further ensconced the skeptics into a world where only their opinions were correct, and they were surrounded by others who felt the same. This deepened their beliefs.

Today, the World Health Organization says vaccine hesitancy is now one of the biggest threats in global public health, citing the 30 percent increase in measles cases worldwide.

Above: network coverage map in Democratic Republic of Congo following the recent Ebola outbreak

Facebook only recently announced a plan to curb the spread of vaccine misinformation on its platforms, following government inquiries. But the health crisis it helped create by way of inattention is already well underway.

To be fair, Facebook is not the only platform corrupted by anti-vax misinformation — Pinterest, Twitter, YouTube, and Facebook-owned Instagram have also just taken measures to address the problem. But Facebook is the largest social network, and a highly influential platform.

Of course, vaccine-preventable diseases are only one of many crises (and potential crises) in public health, along with threats such as Ebola, antimicrobial resistance, influenza, dengue, HIV and many others. Access to more data can help health organizations in many areas.

This is not the first time Facebook has tapped its large data stores or other technologies to aid nonprofits and other researchers. The company has also provided aid organizations with anonymized location data for users in areas affected by disasters — including where people were marking themselves as safe, and where they were fleeing. In another humanitarian-related effort, Facebook has been working with A.I. to create population density maps, including most recently one showing the population of Africa.

“Epidemics pose a growing threat to lives and livelihoods,” said Vanessa Candeias, Head of Shaping the Future of Health and Healthcare at the World Economic Forum, in a Facebook announcement about the new maps. “Mitigating their risk and impact requires every tool in the toolbox.”

News Source =

Oh no, there’s A.I. whiskey now

in AI/Artificial Intelligence/Delhi/India/Microsoft/Politics/whiskey by

Forget all those whiskey brands from musicians and celebs — there’s A.I. whiskey now. Microsoft this week announced it has teamed up with Finnish tech company Fourkind and Sweden-based distillery Mackmyra Whisky to create the “world’s first whisky developed with artificial intelligence.”

Oh no!

Here’s how it will work.

As part of the distillation process, whiskey first spends time — typically years — sitting in charred wooden casks. This turns the clear liquor a darker color, and gives it a unique flavor. How long it stays in the casks, and what the casks held before — like bourbon, wine, sherry, etc. — helps create a specific recipe. Master distillers tweak all these variables along with the different ingredients used to create the whiskey in the first place to come up with a variety of blends.

Until now, this entire process is done by humans with a specialized set of skills. For the A.I. blend, Mackmyra is turning part of the job over to the machines.

The distillery is feeding its existing recipes, sales data and customer preferences to machine learning models, so the A.I. can suggest what recipes it should make next.

The A.I., Mackmyra says, is capable of generating over 70 million different recipes. And it will highlight those it predicts will be most popular and of the highest quality, based on the cask types that are currently on hand.

These models are powered by Microsoft’s Azure cloud platform and Azure cognitive services. Fourkind developed the A.I. algorithms involved, explains Microsoft in its announcement.

However, the distillery notes it’s not actually replacing its Master Blenders with A.I. Instead, it’s using the A.I. to create the recipes which are then curated by the (still human) experts.

“The work of a Master Blender is not at risk,” insists Angela D’Orazio, Mackmyra’s Master Blender. “While the whiskey recipe is created by A.I., we still benefit from a person’s expertise and knowledge, especially the human sensory part, that can never be replaced by any program. We believe that the whiskey is A.I.-generated, but human-curated. Ultimately, the decision is made by a person.”

Microsoft says this is the first time A.I. has been used to augment the process of making whiskey. The finished product will be available in Autumn 2019.

Despite not knowing how the juice turns out, Fourkind wants to turn its algorithms to other industries where complex recipes are involved — including those for other beverages, and also things like perfumes, sweets, or sneaker designs.

This would not be the first time that A.I. has been put to work in a more artistic field.

For example, at Google’s I/O developer conference this month, the company showed off how A.I. could be used in artistic endeavors — including music, visual art, and even fashion.

Of course, when A.I. is tasked with making art, the end results tend to be strange, unworldly and sometimes a little frightening.

Which begs the question: how the hell will an A.I. whiskey taste?

(via TNW


News Source =

Iguazio brings its data science platform to Azure and Azure Stack

in AI/Artificial Intelligence/Azure/Azure Stack/Cloud/data science/Delhi/iguazio/India/machine learning/Microsoft/Politics/Serverless by

Iguazio, an end-to-end platform that allows data scientists to take machine learning models from data ingestion to training, testing and production, today announced that it is bringing its solution to Microsoft’s Azure cloud and Azure Stack on-premises platform.

The 80-person company, which has received a total of $48 million in funding to date, aims to make it easier for data scientists to do the work they are actually paid to do. The company argues that a lot of the work that data scientists do today is about managing the infrastructure and handling integrations, not building the machine learning models.

“We see that machine learning pipelines are way more complex than people think,” Iguazio CEO Asaf Somekh told me. “People think this is good stuff, but it’s actually horrible. We’re trying to simplify that.”

To do this, Iguazio is betting on open source. It uses standard tools and API to pull in data from a wide variety of sources, which is then stored in its real-time in-memory database, which can handle streaming data, as well as time series data, tables and files. It also uses standard Jupyter notebooks instead of some form of proprietary format, but what’s maybe most interesting is that the company also built and open-platform for building data science pipelines. To build the models, Iguazio also uses KubeFlow, a machine learning toolkit for the Kubernetes container platform.

Given that Azure and Azure Stack are essentially the same platform, as far as the APIs are concerned, Iguazio can then take its software and run it both in the cloud and on premises. Soon, it’ll also bring its service to Microsoft’s Azure Data Box Edge, Microsoft’s hardware solution for storing and analyzing data at the edge, which can be equipped with FPGAs for deploying machine learning models.

“Partnering with Iguazio, we can offer additional options for AI applications in the cloud to also run on the edge. Iguazio provides an additional path to run AI on the edge beyond our current Microsoft Azure Machine Learning inferencing on the edge,” said Henry Jerez, Principal Group Product Manager at Microsoft’s Intelligent Edge Solutions Platform Group. “This new marketplace option provides an additional alternate path for our customers to bring intelligence close to the data sources for applications such as predictive maintenance and real-time recommendation engines.”

The Azure solution joins Iguazio’s existing options to deploy its services on top of AWS and Google Cloud Platform.

News Source =

Word’s new AI editor will improve your writing

in AI/Artificial Intelligence/Delhi/India/Microsoft/microsoft word/MS Build 2019/Politics by

If you write in Microsoft Word Online, you’ll soon have an AI-powered editor at your side. As the company announced today, Word will soon get a new feature called “Ideas” that will offer writers all kinds of help with their documents.

If writing is a struggle for you, the most important feature of Ideas is surely its ability to help you write more concise and readable text. You can think of this as a grammar checker on steroids, as it goes beyond fixing obvious mistakes and focuses on making your writing better. It uses machine learning, for example, to suggest a rewrite when you mangled a complex phrase. Ideas will also help you write more inclusive texts.

The cloud-based tool will also give you information about the estimated reading times and decode acronyms for you, based on data it has about your company in the Microsoft Graph.

Ideas can also automatically extract key points from a document. That’s probably more interesting to a reader than a writer, though, so I expect that’s something users will use when somebody sends them a 67-page news summary.

Microsoft also notes that Ideas will bring something called the “Word Designer” to the word processor, which will help you style different parts of a document, including tables.

These new features will come to Office Insiders in June and will become generally available to all users in the fall.

News Source =

Facebook updates PyTorch with a focus on production use

in AI/Artificial Intelligence/Delhi/Developer/F8 2019/Facebook/India/Politics/PyTorch by

During last year’s F8 developer conference, Facebook announced the 1.0 launch of PyTorch, the company’s open source deep learning platform. At this year’s F8, the company launched version 1.1. The small increase in version numbers belies the importance of this release, which focuses on making the tool more appropriate for production usage, including improvements to how the tool handles distributed training.

“What we’re seeing with PyTorch is incredible moment internally at Facebook to ship it and then an echo of that externally with large companies,” Joe Spisak, Facebook AI’s product manager for PyTorch, told me. “Make no mistake, we’re not trying to monetize PyTorch […] but we want to see PyTorch have a community. And that community is starting to shift from a very research-centric community — and that continues to grow fast — into the production world.”

So with this release, the team and the over 1,000 open-source committers that have worked on this project are addressing the shortcoming of the earlier release as users continue to push the limits. Some of those users, for example, include Microsoft, which is using PyTorch for its language models that scale to a billion words and Toyota, which is using it for some of its driver assistance features.

As Spisak told me, one of the most important new features in PyTorch 1.1 is support for TensorBoard, Google’s visualization tool for TensorFlow that helps developers evaluate and inspect models. Spisak noted that Google and Facebook worked together very closely on building this integration. “Demand from developers has been incredible and we’re going to contribute back to Tensorboard as a project and bring new capabilities to it,” he said.

Also new are improvements to the PyTorch just-in-time compiler, which now supports dictionaries, user classes and attributes, for example, as well as the addition of new APIs to PyTorch that support Boolean tensors and support for custom recurrent neural networks.

What’s most important for many production users, though, is the improvements the team made to PyTorch’s distributed training capabilities. These include the ability to split large models across GPUs and various other tweaks that’ll make training large models faster when you have access to a cluster of machines.

News Source =

1 2 3 13
Go to Top