Amzon AWS Cost Control Insights
June 29, 2020
Amazon’s AWS is a fascinating business case. On one hand, AWS reduces some of the hurdles to modern solution development. On the other hand, it is easy — even for an experienced Certified AWS expert — to forget what’s running, whether a particular service is unnecessary, or what processes are tucked into the corner of Jeff Bezos’ profit making machine. “Our AWS Bill is ~ 2% of revenue. Here’s How We Did It” provides a run down of the money gobblers and provides some helpful guidance. There are screenshots in the Gulf racing colors of orange and blue. There are explanations. Plus, there are useful insights; for example:
Our application is a Shopify app and during the process of building the application we created a Shopify store. Every Shopify store gets its own personal CDN where you can manually upload anything and it will be served over the Shopify CDN. So we minified and uploaded our JS file to the CDN of our Shopify store and now we serve 20000 Shopify stores using this method at zero cost.
One problem: There are more ways for Mr. Bezos to suck cash from eager and willing customers than helpful explanations of how to keep expenses low.
Stephen E Arnold, June 29, 2020
Google: The Me Too Innovation Juggernaut
June 28, 2020
Like Microsoft, Google will have an opportunity to explain its business practices. Perhaps the company will explain how its magical black box interacts with the layers of software wrapped around the smart software too? Maybe, maybe not.
Turning to more practical matters, Microsoft’s decision to kill off the empty spaces called Microsoft Stores illustrates that me too innovation does not work reliably. Hey, Microsoft tried. Also, Microsoft’s interesting attempt to clone Amazon Twitch has ended by creeping to Facebook on little cat’s feet. Hasta la vista, Sr. Mixer. The hope is that Facebook’s magic returns and converts Mixer into a Zucking winner. Hope is useful for some.
The real news, however, concerns Google’s embrace of me too innovation. “Group Video Calls Now Arrive on Google Nest Hub Max.” I don’t know what a Nest Hub is. I don’t know what a Nest Max is. I don’t know what a Hub Max is. What I do know is that Google wants to be JLZ. That’s an acronym for “just like Zoom.” News flash: Zoom has a one word product name, “Zoom.” Google is four times more creative because the GOOG uses four words. Efficient, clear, and memorable: Google Nest Hub Max. Who will be the first Stadia addict to have the letters GNHM tattooed on his or her mouse hand?
The second Google item concerns Google’s acquisition of a me too company which developed some glasses like Google Glass. That was a product that sparked one wit to coin the term “glasshole”.
Google’s Parent Alphabet Is Reportedly in Final Stages to Buy Smart Glasses Maker North, As the Augmented Reality Race with Apple Heats up” reveals that Google is acquiring a company which practices the me too approach to product innovation. What happens when two me two innovation teams collaborate? That sounds like one of those discussion questions bandied about in Dr. Francis Chivers’ phenomenological existentialism classes at Duquesne University. Exciting, eh.
Although not on the scale of virtue signaling practiced by other high technology companies, Google wants to be more diverse. Okay, that’s original.
“Google Execs Say We Need a Plan to Stop A.I. Algorithms from Amplifying Racism” reports:
Two Google executives said Friday that bias in artificial intelligence is hurting already marginalized communities in America, and that more needs to be done to ensure that this does not happen.
Haven’t I read this sentiment before? You?
See. Me too works!
Stephen E Arnold, June 28, 2020
Policeware: Fascinating Real Journalists Again
June 27, 2020
Imagine writing about policeware — software and specialized services tailored to the needs of enforcement authorities — this way.
You learn about a quinoa farmer in rural Virginia. You look into the farmer’s activities and find that the farmer sells produce to locals heading toward North Carolina. You add flavor to your story the way a cook in Lima converts quinoa into a gourmet treat for travel weary tourists. The farmer is an interesting person. The farmer is struggling to survive. The farmer labels the quinoa as “world’s best” and “super healthy.” The farmer becomes famous because he tells you, “I sell more quinoa despite the local regulations and the Food Lion supermarket.” The problem is that the story’s author is unaware of Archer Daniels Midlands, an outfit with an interest in quinoa.
The story is a human interest write up particularized to a single quinoa farmer in a state known for a mall, traffic jams, and government contractors. Micro story gives the impression that Virginia is a great place for quinoa. Accurate? A reflection of the business environment? A clear reflection of local ordinances?
Nah.
I thought about the difference between a quinoa farmer’s story and a general lack of awareness about Archer Daniel Midlands when I read “Firm That Tracked Protesters Targeted Evangelicals During 2016 Election.” The outfit providing data may have more in common with the hypothetical quinoa story that meets the eye. Coverage of the policeware or intelware market sector invites micro examples used to support large scale generalizations about the use of data from mobile phones or open source information like public posts on a social media site.
Furthermore, small companies like the one described in Vice Motherboard article exist in every business sector. Focusing on a single firm — whether a quinoa farmer or a commercial data provider — may not provide a representative description of the market.
News flash: Data are available to companies, government agencies, and not for profit organizations from hundreds of companies. Some of these are tiny like Mobilewalla. Others are beefy; for example, Oracle BlueKai. Still others occupy a middle ground like Dataminr. Others are loosely affiliated with other countries’ government entities; possibly Innity.
The fixation on policeware appears to be a desire on the part of “real” journalists to tell mobile phone users that the essential device is gathering data about the user.
News flash: Mobile devices which seek cell towers and WiFi connections emit data as part of their normal functioning. Individuals who use mobile devices to look at ads on ManyVids, surf the Dark Web from a mobile device, and use the gizmos to buy contraband and pay with Bitcoin are skywriting. Big messages are available to those with access to different sets of data.
Some of the data flows into the stellar giants of the online world; for instance, Facebook and Google. Other data gathers in the telcos. Quite useful data floods from online mobile game enthusiasts. Granny in the retirement home happily provides companies like Amazon with a flow of information about what’s hot from her quite particular point of view.
My thought is that chasing quinoa farmer stories is a new and exciting angle for some “real” journalists. But is there a different story to be researched, understood, and communicated.
“Real” journalists might begin by asking and answering with facts, not anecdotes, these questions:
What organizations are the equivalent of the agribusiness giants just in the commercial database sector? How are these data gathered, verified, and made available? What people, companies, and organizations license these data? Why does a commercial database business exist? When did data morph into mechanism for dealing with certain types of events? How many government agencies integrate these types of data into their “feet on the street” activities? What’s the upside to these data and their use? What are the downsides to these data and their use?
The stories about the quinoa farmer are okay. Moving beyond the anecdote to the foundation of commercial data licensing is more meaningful and more interesting.
The problem may be that moving beyond the quinoa approach takes work, time, and understanding. Hey, “real” journalists have to log into Slack and then jump on a Zoom call. This “go beyond quinoa” is just too much like “real work.”
That’s a problem I assert for individuals uninterested in what happened when trans-Atlantic telegraph messages began to flow. Why not look into that type of history?
Stephen E Arnold, June 27, 2020
Policeware: Making Headlines
June 26, 2020
DarkCyber noted “Machines with Brains.” The article includes a category or “pre title” with the phrase “From Our Obsession.” The “our” is ambiguous. Is it the “our” of the Silicon Valley real news team members or is it the “our” of the zip zip technology craving social milieu?
The point of the write up is that policeware using pattern recognition and other assorted technologies are not ready for prime time. The article identifies several companies as providing solutions that create problems, not ones that solve them. These entities are:
- Amazon, IBM, and Microsoft, deeply suspect companies but ones that had the common sense to drop out of the facial recognition marathon. “Yes, quitters can be winners” in the “From Our Obsession” point of view.
- DataWorks, “one of the biggest resellers of facial recognition technology to US Police departments.” The company allegedly has “contracts with police in Detroit, Chicago, New York City, Santa Barbara, South Carolina, Pennsylvania, and Maryland.” DarkCyber admires alphabetization and mixing states and municipalities. Without context, how large are the contracts? What are the terms of the deal? Are these proofs of concept or full-blown integrated solutions humming 24×7 or some other type of installation.
- Cognitech
- NEC
- Rank One Computing
The policeware market is one which most “real” journalists struggle to understand. Yesterday, in a conversation with a “real” journalist employed by the one and only Rupert Murdoch organization, I chortled during the “interview/conversation” as the young “real journalist” struggled to understand why law enforcement and intelligence professionals try out new technologies.
My comments about the companies providing policeware did not compute for the sincere and apparently fascinated news hunter. The idea that vendors provide news technology, make modifications as technical problems arise, and alter systems as users – yes, real enforcement officials – struggle to apply technology to the challenges enforcement presents.
Several observations:
First, the policeware and intelware markets, companies, and technologies are unknown territories for most technology professionals and terra incognita for a large percentage of “real” journalists. This means the individuals do not know of what they news gather. Out of context is the principal method employed.
Second, the solutions developed for enforcement and intelligence officials are a surprise to the uninformed. No one likes surprises; for example, the idea that a cherished group of Facebook friends may harbor a child molester or a contraband dealer. How does one mitigate surprise? Easy? Sensationalism, finger pointing, and generalizations. Facial recognition sucks. Easy. Does the Amazon-powered Ripper technology suck? What’s that? Ignorance is bliss for some.
Third, exactly what bureaucratic solutions exist to deal with technology? (Oxymoron alert: Bureaucracies are subject to Parkinson’s Law and Augustine’s Laws.) Some “real” journalists enthusiastically embrace mobile devices, online hook up services pretending to be video dating services, and the Twitter lifestyle. Maybe the newly minted experts in policeware have some ideas other than “don’t use technology”? Wait. That won’t work because fairy land, in case one has been oblivious to the social construct, seems to be emulating the world of Road Warrior.
Net net: Information in context and perspective are useful when writing about a technology sector with which one is not familiar. Just a thought because the morphing of “Machines with Brains” into “humans with brains” is an interesting idea to contemplate.
Perhaps an “obsession” with perspective, context, knowledge, and less sensationalistic short cuts would be helpful?
Policeware is becoming a beat. Good. Let’s strive for context, not shouting “Fire” in a socially distanced movie theater.
Stephen E Arnold, June 26, 2020
Facebook: Trust Crisis? You Must Be Joking, Never
June 26, 2020
I read “Facebook Faces Trust Crisis As Ad boycott Grows.” The lovable college drop out who founded Facebook seems to be in pickle. The write up reveals that the company Facebook has to mend some fences with advertisers.
Specifically:
In a call with over 200 advertisers Tuesday, Facebook’s head of trust and safety policy Neil Potts “acknowledged that the company suffered from a trust deficit,” according to the Financial Times. A source familiar with the meeting confirmed the comment. The conversation occurred amid a growing boycott of Facebook and Facebook-owned Instagram’s ad platform by roughly a dozen brands.
In a moment of insight, the author of the write up states:
The political and social pressure on Facebook is ramping up, but the tech giant doesn’t show any signs of seriously changing its policies in response to the mounting pressure, as most politicians and marketers seem to benefit too much from Facebook advertising to really give it up long-term.
What’s this mean? There are some good reasons to allow Facebook to just keep being Facebook. One of them is the data Facebook gathers has value to some individuals in government agencies. Losing Backpage was a set back, but losing Facebook, hey, let’s talk about this.
Second, where there are eyeballs, there are advertisers. The ethical compass of advertisers spins toward selling and making money. That pull is strong enough to light up some folks’ Faraday effect.
Stephen E Arnold, June 26, 2020
Lucky Ukraine: A Data Bomb Test Site
June 26, 2020
Russia surprised the world when Putin ordered his soldiers to invade Ukraine and annex Crimea. Putin’s actions against the Ukraine are not the only modern war stories circling Russia. The Small Wars Journal examines how the Great Bear could be conducting a futuristic warfare using technology: “Russia In Ukraine 2013-2016: The Application Of New Type Warfare Maximizing The Exploitation Of Cyber, IO, and Media.”
Russia could be masters of cyber and information warfare tp support militaristic/political objectives against domestic and international enemies. The thesis study reads logically, but also Russia’s recent actions support it:
“The Russians were able to use Ukraine operations as a test for New Generation Warfare (NGW) to enhance the deep battle concept. Russia has adeptly executed deep battle, creating time and space to effectively employ limited ground forces and special operations to achieve desired effects. The employment of the cyber domain created windows of opportunity for success and simultaneous execution of offensive and defensive tasks across the strategic and operational levels and other domains. Additionally, the cyber capabilities employed allowed the Russians to achieve three critical strategic effects; 1) troop levels were minimized through integrated cyber operations and operational advantage gained; 2) Russian leadership maintained plausible deniability through effective cyber and information operations delaying international intervention; 3) cyber operations achieved desired effects and kept the threshold for violence below an international outcry for intervention or interference allowing the Russians to achieve the strategic objective to control key terrain in Ukraine.”
While Russia remains the punch line for jokes about international affairs, the country is not a laughing matter as history shows. Under Putin’s leadership, Russia proves to be masterful at manipulating multiple information sources: TV, Internet, radio, etc. to cover their rears while executing desired. Russia has invested capital in homegrown technology, instead of relying on foreign made.
Russia used its cyber forces to overwhelm the Ukraine with malware and disinformation through media channels to annex the Crimean territory. It was a brilliant, mostly bloodless tactic, because Ukraine does not have the technology nor physical forces to fend off the Great Bear. Smaller countries, especially in Eastern Europe and Asia, remain sitting ducks if the enter Russia’s crosshairs.
The biggest issue is proving Russia’s culpability and whether the country will be held accountable. Russia’s more militaristic past still casts shadows on its current society, but Russian citizens are not in favor of being a military power again. Like the rest of the world, they want to live a steady, peaceful life.
Whitney Grace, June 26, 2020
About That Degree in Real Journalism?
June 26, 2020
Now that humans and algorithms share the job of curating online news, how do the two compare? Curious, Northwestern University’s Jack Bandy and Nicholas Diakopoulos examined one news service and did the math. Mac O’Clock shares Bandy’s summary, “What We Learned About Editors vs. Algorithms from 4,000 Stories in Apple News.”
In the case of Apple News, which boasts 125 million monthly users, human editors pick the “top stories” while AI chooses the “trending stories.” Bandy created a program to track the articles curated by each for two months. The researchers came to three conclusions. First, human editors chose pieces more evenly across news sources. Second, humans chose a wider range of sources. Interestingly, the narrower group of sources favored by the algorithms tended toward topics like celebrities and entertainment. This observation pointed the pair to their final conclusion—that human editors chose fewer “soft news” stories and more articles on serious topics. See the illustrated write-up for more on each of these points.
Bandy follows up:
“Our results highlight the trade-offs between human curation and algorithmic curation. While our study only looked at one platform, it shows that human editors were ‘much more subtly following the news cycle and what’s important,’ as Lauren Kern (editor in chief at Apple News) put it. For many readers and publishers, this is good news. The data shows that editors choose stories about important topics from a diverse set of sources, and choose those sources quite evenly. This is less true of the algorithmic Trending Stories, where readers will see more ‘soft news,’ and just a few major publishers tend to make the cut.”
It is a good idea to frequent a site at which humans still choose the top stories, as one example illustrates. Bandy uses the Wayback Machine to see the Google News headlines from the end of February, and was grateful he had not relied on that AI-centric page for his news at the time. He writes:
“All of the stories mention two things: coronavirus and Donald Trump. If you read them, you may glimpse some information about the impending pandemic — a ‘severity warning’ from the CDC, for example. The headlines probably grab your attention, but they do not provide meaningful information. Apple’s editors had a different approach that day, featuring an article with the headline ‘Coronavirus’s spread in U.S. is “inevitable,” CDC warns.’ It was a formal, descriptive piece from the Washington Post that quoted several officials at the Center for Disease Control. … I remember one quote from the article that changed my expectations for the coming months: ‘Disruptions to everyday life may be severe, but people might want to start thinking about that now’.”
Perhaps one day algorithms will be able to learn that sort of discernment, but now is not that time. Who, or what, is curating your news? Cynthia Murrell, June 26, 2020
News Flash: SEO Leads to Buying Ads with or without Semantic Blabber
June 26, 2020
A Search Engine Optimization blog offers some axioms on semantic search for a crowd used to manipulating keywords, backlinks, URL structure, and the like. David Amerland posts, “Five Semantic Search Principles to Help Organize your Content and Marketing.” To us, the result seems like a reconstruction of an Incan incantation with a mystical diagram tossed in for added magic. Amerland writes:
“Semantic search is as open to analysis and interpretation of the elements that govern it as the good ol’ Boolean search of the past was. Yet the effort required to achieve a positive outcome (i.e. higher visibility in search) is now every bit as labor and cost intensive as doing the right thing. Semantic search, in other words, does not automatically make us all behave in a morally better way because it is the right thing to do. It makes us behave morally better because there is no viable alternative.”
So far so good. The piece then gets into search as psychology. We’re told the structure of search has always shaped users’ perceptions of the information presented and, by extension, their behaviors. We cannot argue with that much. Then Amerland continues:
“Semantic search has much in common with Gestalt psychology. It looks at the phenomena it studies as organized and structured wholes rather than the sum of their parts and, like semantic search, it deals with entities and how we perceive them. The question that arises with semantic search, now, is that since there are so many elements that drive it and since many of them are roughly equal so that none has a significant advantage over the other, how can we create a strategy that actually works? This is where Gestalt psychology comes into its own.”
Gestalt psychology as an SEO strategy—interesting. See the article to go further down the rabbit hole, where it discusses, with illustrations, its five principles: the law of proximity, the law of similarity, the law of perceptual organization, the law of symmetry, and the law of closure. We grant that SEO professionals are nothing if not creative, but perhaps there is such a thing as over thinking one’s approach to algorithm manipulation.
Cynthia Murrell, June 11, 2020
Business Intelligence: When Case Studies Are Not
June 25, 2020
A case study in the good old days was a little soft, a little firm, and a lot mushy. The precise definition of a case study is “it depends.” The problem is that case studies are often not easily duplicated. The data collection methods vary because many organizations do not keep data or, if kept, do not maintain data in a consistent manner. There’s a bright young sprout who wants data in a format unintelligible to other people and maybe systems.
Other minor potholes wander through thickets of subjectivity and into the mysterious world of sparse data. Ever heard, “Well, we don’t have that data, but we can take the inverse of these data and use them.”
The silly idea of answer who, what, why, where, where, and how are often discarded because the information is not available, secret, or just too much work. Just because you know. Meetings.
I thought about the murky world of case studies when I read “5 Valuable Business Intelligence Use Cases for Organizations.” First, there is the word “valuable.” Second, there is the phrase “business intelligence.” Third, there is the jargon “use cases.” Examples is a useful word. Why not employ it?
What caught my attention was that the examples illustrate the type of effort a group of volunteers make when no one wants to work very hard. You may have participated in filling a food basket with canned goods which few would actually consume.
Let’s look at one “use case,” and I will leave it to you to dig through the other four.
Use case number 2 explains how business intelligence can speed up and make better decision making manifest themselves. Okay, we have this pandemic thing. We have a bit of a financial downturn. We have the disruption of supply chains. We have the work from home method. We have Zoom solutions to knit together humans who like to hang out in break rooms and share gossip.
The fix is to use business intelligence to bring
together data mining, data analysis and data visualization to give executives and other business users a comprehensive view of enterprise data, which they can then use to make business decisions in a more informed way.
Now what do these terms mean? Data mining, data analysis, and data visualization. Where to the data come from? Are the data valid? Are the data comprehensive?
The “evidence” in the example is a survey conducted in a sample of an unknown number. The sample which may or may not be representative reports that “reliable data” is a hurdle. No kidding.
The case explains that the shift to real time data is important. Plus real time data piped into predictive analytics allows “fast action.”
The conclusion: Instantaneous decisions are possible.
Net net: The write up is a fluffy promotion of a nebulous concept. Use case my foot! I made an instant decision. Business intelligence like knowledge management and content management is a confection.
That’s why crazy “use case” explanations are needed. The other four examples in the article? Similar. Disconnected. A food basket filled with stuff no one will consume.
Stephen E Arnold, June 25, 2020
Search for the Young in Heart and Mind
June 25, 2020
An expert in search at an outfit called Blue Fountain Media understands search and retrieval. The insights about matching a user query with relevant content almost caused me to weep. See and experience the impact of high quality information yourself by navigating to “Using AI-Powered Visual Search to Enhance CX.” Note: I had to click past warnings about possible malware, complete two captchas about vehicles, and put up with a weird mobile oriented output in order to access this magnificent research paper.
Was the effort worth the intellectual reward? Sure, like a Twinkie for a starving college student addled by a long weekend of social distancing in Florida.
The main point of the write up is:
The old saying “a picture is worth a thousand words” takes on all new relevance when we consider the next big feature coming over the ecommerce horizon: AI-powered visual search.
That’s insightful. Like TV, right?
The write up states that AI has some other benefits:
- The missing agent has been artificial intelligence which has become increasingly more adept at image recognition.
- Beyond image recognition, AI makes it possible to learn a person’s taste and style preferences.
And there’s a stunning revelation about “most shoppers”. Sure, this is a generalization, but what search expert wants to deal with verifiable data:
Let’s face it, most shoppers know what they want – when they see it – but find themselves completely dumbfounded when it comes to describing what they’re after using the traditional text-driven search engine.
What about a student wanting a copy of Jacques Ellul’s Technological Bluff. Would a visual search be feasible on a mobile device while social distancing on a Zoom video call?
Of course, of course.
Plus, the search expert reveals some next steps. How does this match up with the wonks trying to improve the precision and recall for stress analysis engineers dealing with the failure of an emergency core cooling system valve?
The key, of course, is to mark up these images appropriately. Ironically, yes, this will still include keywords. But if you follow Googles [sic] best practices for images, you’ll find values such as placement and metadata are increasingly more important than simple text descriptors.
A diagram is okay, but the issue is calculating engineering stresses with a dash of math amongst the photos.
Amazing what one learns about search and retrieval when one consults true experts. And those engineers struggling with the analytic job. Check out Google images. It’s a best practice, particularly when the valve is part of the reactor on a nuclear submarine. There’s nothing quite like Google metadata. Are search engine optimization professionals who are also search experts up to task. Sure.
Stephen E Arnold, June 25, 2020
-
- Subscribe to Beyond Search
Feature archive
News archive
-
