Details on Microsoft Singularity for AI Workflow

Worth a look, I will.  Note indication of scheduling AI workloads.

Microsoft goes public with details on its 'Singularity' AI infrastructure service

Microsoft's Azure and Research teams are working together on the 'Singularity' AI infrastructure service.

By Mary Jo Foley

Posted in All About Microsoft on February 21, 2022 | Topic: AI & Robotics

Microsoft's Azure and Research teams are working together to build a new AI infrastructure service, codenamed "Singularity." The Singularity team is working to build what Microsoft describes in some of its job postings for the group as "a new AI platform service ground-up from scratch that will become a major driver for AI, both inside Microsoft and outside."

A group of those working on the project have published a paper entitled "Singularity: Planet-Scale, Preemptible and Elastic Scheduling of AI Workloads,"  which provides technical details about the Singularity effort. The Singularity service is about providing data scientists and AI practitioners with a way to build, scale, experiment and iterate on their models on a Microsoft-provided distributed infrastructure service built specifically for AI.

Authors listed on the newly published paper include Azure Chief Technical Officer Mark Russinovich; Partner Architect Rimma Nehme, who worked on Azure Cosmos DB until moving to Azure to work on AI and deep learning in 2019; and Technical Fellow Dharma Shukla. From that paper:

"At the heart of Singularity is a novel, workload-aware scheduler that can transparently preempt and elastically scale deep learning workloads to drive high utilization without impacting their correctness or performance, across a global fleet of accelerators (e.g., GPUs, FPGAs)."  (Details at her ZDnet site linked to above) .... ' 

Considering Robot Consciousness

OK, we are far from this, still sci fi. but worth a thought.  So we will be ready then.  

If a Robot Is Conscious, Is It OK to Turn It Off? The Moral Implications of Building True AI  By Anand Vaidya

In the Star Trek: The Next Generation episode “The Measure of a Man,” Data, an android crew member of the Enterprise, is to be dismantled for research purposes unless Captain Picard can argue that Data deserves the same rights as a human being. Naturally the question arises: What is the basis upon which something has rights? What gives an entity moral standing?

The philosopher Peter Singer argues that creatures that can feel pain or suffer have a claim to moral standing. He argues that nonhuman animals have moral standing, since they can feel pain and suffer. Limiting it to people would be a form of speciesism, something akin to racism and sexism.

Without endorsing Singer’s line of reasoning, we might wonder if it can be extended further to an android robot like Data. It would require that Data can either feel pain or suffer. And how you answer that depends on how you understand consciousness and intelligence.  ... " 

Most Valuable AI Companies

Fascinating look from the Singularity Hub.  Who are the 32 AI Unicorn Hubs?   Useful to see the kinds of  things that are being worked on.

The World’s Most Valuable AI Companies, and What They’re Working On
By Peter Rejcek In the Singularity Hub

It recognizes our faces. It knows the videos we might like. And it can even, perhaps, recommend the best course of action to take to maximize our personal health.

Artificial intelligence and its subset of disciplines—such as machine learning, natural language processing, and computer vision—are seemingly becoming integrated into our daily lives whether we like it or not. What was once sci-fi is now ubiquitous research and development in company and university labs around the world.

Similarly, the startups working on many of these AI technologies have seen their proverbial stock rise. More than 30 of these companies are now valued at over a billion dollars, according to data research firm CB Insights, which itself employs algorithms to provide insights into the tech business world.  ... "

Singularity by 2045?

Says Kurzweil,  Google's Director of Engineering.  Will that mean Google Drive will be the approved recepticle?  Google Analytics the means of  understanding?  Deep Mind the means for running offline? Very ambitious I think.  Google presents the Singularity Cloud! But if so, there is lots to figure out before then.

Origin of Trancendant Singularity

I had thought that Kurzweil was the originator of the mathematical term singularity in the Transcendence sense, but the Wikipedia writes:

" ... The first use of the term "singularity" in this context was by mathematician John von Neumann. In 1958, regarding a summary of a conversation with von Neumann, Stanislaw Ulam described "ever accelerating progress of technology and changes in the mode of human life, which gives the appearance of approaching some essential singularity in the history of the race beyond which human affairs, as we know them, could not continue".[2] The term was popularized by science fiction writer Vernor Vinge, who argues that artificial intelligence, human biological enhancement, or brain-computer interfaces could be possible causes of the singularity.[3] Futurist Ray Kurzweil cited von Neumann's use of the term in a foreword to von Neumann's classic The Computer and the Brain. ... " 

In the enterprise we worked for years withVon Neumann's daughter, Marina. 

Machines Taking Over

Are we making it too easy for machines to take over?  In ReadwriteWeb.  Do we see early symptoms when we listen for every smartphone ring?   I have been reading a novel recently that includes, in part,  this theme.  More on that to come.  Not out of the question.  " ... What if computer code could write itself? What if robots could think for themselves and continuously learn from their environment while being fed contextual information from a vast global network of data? What if the machines could build themselves and propagate, much in the same way that mammals give birth to new mammals? ... " 

Blog Tags are Here

Starting now you can utilize blog tags here, also called labels.  They are found at the bottom of each post.  These tags classify the blog post with others that are of related topics.  You can click on these at the bottom of the post to get all blog posts classified this way.  You can also use this syntax - label: tag in the search block at the upper left to search for particular categories. (case sensitive)

For multiple tags use label:tag1 label:tag2 (case sensitive).  Many past key blog posts, but not all, have been tagged.  All new posts from now on will be tagged.  Your comments are welcome.

Team Blogging Practices

When I ran a multiple internal blog operation in the enterprise, the goals were knowledge retention, sharing and collaboration.  We did both internal and external facing work.  This new article looks at the practices involved.  In particular as it addresses team operation.  We used a wiki style architecture for permanence.   Largely obvious, but worth the review.

Being Postdigital

A new idea to me, In a newly distruptuve era of analytics are we  ready to be 'postdigital'?  I think we are a distance from that, singularity or not.  But an interesting exercise in preparation.

Smart Machines and World Without Work

General AP article that covers a number of developments in the last decade.  Perhaps leading to less labor.  This kind of labor singularity is still some distance away.  Goods need to be created, transported and maintained.   Work has and will change. That you can expect soon.

Singularity Rising

Moore's law and Artificial Intelligence.  What does it mean to be 'smart' for machines?  A new book:    Singularity Rising. describes the problem. What will the singularity, in whatever form it may take,  mean to human beings to human civilization?   And an interview with author James D Miller.

Boosting Productivity with Social Media

In the HBR Blogs:  A video interview with Alexandra Samuel. Some interesting points here.   Yes, I believe this can be done, selectively.  I do not yet believe that managing the process can be effectively done with today's social media, which disrupts more than it focuses.  See also Vision Critical, which does predictive analysis via social conversations.

Kurzweil Joins Google

The futurist proposer of the approach of an inevitable people-machine singularity, Ray Kurzweil, has joined Google as an employee.   In GigaOM.   Apparently to work on machine learning and language processing.      Given the long time interest of the Google founders and the company itself on AI related topics, this sounds like a logical step.  Note also, the Singularity University.

Commenting on Communications

I am getting increasing number of Spam Comments with an author of  'Anonymous'.  It is getting difficult to even scan all messages.  If you would like to improve your chances of being approved for publication include an author name.  If you want to get to me privately for a speaking or consulting session, use the contact information in the left hand column.  I am speaking at the Rotman School of Business in Toronto this week.  I will always respond to requests promptly.  I will not approve comments that are not on point to this blog and the message commented on.

Unlocking the Predictive Power of the Web

A piece in Singularity Hub on the temporal search in the forward looking search engine Recordedfuture, which I have now followed for some time.   It now has a new user interface in Beta, which provides a timeline-style interaction, improving its temporal style.

Subscribe to My Feed

This blog has always had an RSS feed for syndication.  If you find it useful please do subscribe.  This blog has been around for a number of years.  It parallels my consulting and technical interests.  I have made hundreds of connections through communications via this blog and I am glad to facilitate others.

Innovation and Invention Not the Same

Good Techdirt piece on the exploration of innovation.  We met with PARC and had a lunch with Engelbart and discussed how the mouse eventually emerged, hoping to apply come of the ideas to internal innovation promotion and transmission.  Aim to read the whole thing, with some caution given the author ...

" ... In his latest piece, Gladwell goes a step further in his exploration of innovation, in writing about the difference between invention and innovation, picking apart the classic story of Steve Jobs seeing the GUI/mouse combo at Xerox PARC and "copying" it for the Macintosh. Gladwell points out that the lessons that some take from the story aren't really correct. Specifically, one of the standard lessons is the idea that Xerox had the personal computer revolution in its hands and let it slip away. But Gladwell points out that this isn't really true. While PARC showed Jobs that idea (much of which was copied itself from Doug Engelbart and his famous work at SRI), it really was the implementation that mattered, and Jobs and Apple (along with Ideo) had to work quite hard to take the idea of the mouse -- which cost hundreds of dollars and was fragile in the Xerox version -- and make it cheap, reliable and easy to use ... "

ScanBuy at Home Depot

Do it yourself stores have a particular challenge, they often need to train shoppers in the use of many products that they might buy at a store for the first time.  Sometimes called 'extended packaging', the idea is to link a shopper to necessary information, specifications, sales-pitches, promotions and use videos that might be available.  We talked to Scanbuy about their 2-D barcode reading system several years ago. You can experiment with their Scanlife concept online.  They now support all kinds of barcode types.

 More shoppers understand the concept today and there are more smartphones out there.  Yet it still requires a level of 'training' to do the scanning.   It was an early implementation of reading barcodes from a smartphone.   Different than a location based system. The codes can be read at home in printed advertisements, online or at the store.  More about what Home Depot is doing in Mashable.  ScanBuy was impressive in attempting to address the problem of understanding shopper behavior.

Smartphone Enabled Cancer Detection

A protype that is a good example of mobile sensor-based data display and analysis:  At the Singularity Hub: " ... Massachusetts General Hospital that have knocked it out of the park by integrating a microNMR device that accurately detects cancer cells to a smartphone. Though just a prototype, this device enables a clinician to extract small amounts of cells from a mass inside of a patient, analyze the sample on the spot, acquire the results in an hour, and pass the results to other clinicians and into medical records rapidly. How much does the device cost to make? $200.   ... "

At the Singularity Hub

Retinal implants at the newly discovered Singularity Hub blog.  I have my doubts about broad singularity type predictions, but this blog covers lot of related topics that are useful.