Is It Party Time for STM Professional Publishers?
March 4, 2022
I spotted a TorrentFreak write up called “FBI Gains Access to Sci-Hub Founder’s Google Account Data.” The article explains that investigators are gathering information about Alexandra Elbakyan, the founder of what the article references as the “Pirate Bay of Science.”
The idea behind the service is to make paywall protected content available without the paywalls. The article explains what agencies have been involved and some of the legal procedures followed. These are routine but may be surprising to those who think about new recreational vehicles and the new pizza place.
What makes the investigation interesting is that references are made to Ms. Elbakyan’s alleged links to other governmental entities.
Several observations:
- Alleged links to a foreign power engaged in hostile actions move the story from scientific, technical and medical content made available without the pro9fessional publishers permission to a higher level of security concern.
- Professional publishers have not been happy campers since Sci-Hub became available. (Is this because the service has chewed into some revenues for these commercial enterprises? My guess is, “Yep.”)
- Allegedly, Ms. Elbakyan lives in Russia and, if the Wikipedia is spot on, she is studying philosophy at the Russian Academy of Sciences. (Will extradition be possible? My view is that the process will be interesting.)
When I read the story, I thought about one professional publishing big wig who said off the record, “That crazy Kazakh has to be shut down?”
Is it party time in the world of STM professional publishing? Not yet, but some may want to buy foil party hats and cheap kazoos.
Stephen E Arnold, March 4, 2022
IBM: Big Blue May Have Some Digital Re-Engineering to Explain
March 4, 2022
Yo, I am a dinobaby, and I am proud of that fact. You want proof. I know what a rotary dial phone is. I know how to use a facsimile machine. Heck, I can still crank out a mimeograph document. I even know how to get a drink from a terracotta jar in rural Brazil. (Love those chemicals and that wonky purple-blue color which reminds me of Big Blue.)
Several years ago, I read a blog by some IBM people which documented the harvesting of old workers. That blog disappeared, of course. It named managers, disclosed snippets of email, and did a fine job to making clear that oldsters had one function. The idea was that before finding their future elsewhere, the old employees would train their replacements. This is a variation on copying data from a DASD to a zippy new storage device, just with humanoids, not silicon.
I have been following the word dinobaby. I entered it into my log of jazzy new terms coined by millennials and GenXers. I put dinobaby between grosso modo memetic learning and vibe shift. This is not alphabetical I know, but I like the rhythm of the words when offered in a dinner conversation about technology.
The word appeared in “IBM Executives Planned to Rid the Company of Older, Dinobaby Employees and Replace Them with Millennials, Lawsuit Alleges.” I thought the lawsuit was an interesting opportunity for legal eagles to generate some money.
Then I read the February 26, 2022, story “IBM Cannot Kill This Age-Discrimination Lawsuit Linked to CEO.” Despite Covid, financial turmoil, and the unfortunate events in Eastern Europe:
The judge overseeing an age-discrimination case against IBM has denied the IT giant’s motion to dismiss the lawsuit, citing evidence supporting plaintiff Eugen Schenfeld’s claim that CEO Arvind Krishna, then director of IBM research, made the decision to fire him.
The write up includes a link to a legal document and some snazzy code names; for example, Project Concord, Project Baccarat, and Project Ruby. It appears that each project was intended to get the big, noisy, weird dinobabies out of IBM’s life.
Not happening yet.
The write up asserts that there are more than 10,000 mainframe capable dinobabies vaporized by the “projects” implemented during the scintillating tenure of Ginni Rometty, former president and CEO of Big Blue. (Did you know that Ms. Rometty worked at General Motors, an esteemed automobile company which developed the Chevrolet Bolt, a model which caught on fire? The owner was not Ginni Rometty. The burning GM vehicle was owned by an elected official in Vermont.)
IBM may escape punishment for its alleged conversion of humanoids into dinobabies. But it will be interesting to follow the legal machinations which now seeks to transform dinobabies into hamsters and gerbils with mainframe and other esoteric skills.
Plus the lawyers can consult IBM Watson for inputs!
Stephen E Arnold,March 4, 2022
Oracle Enters the Classified Cloud Arena
March 4, 2022
Oracle is keeping secrets—those of the Air Force, to be specific. Nextgov reports, “Oracle’s Cloud Can Now Host Select Top Secret Defense Data.” The sensitive data will reside in a classified Cloud. Reporter Frank Kinkel writes:
“According to the company, Defense Department assessors granted it an authority to operate, or ATO, for secure processing of some of the Air Force’s most sensitive data in what the company calls Oracle National Security Regions. These air-gapped computing regions are connected only to government networks and not to the rest of the internet. … The ATO applies only to Air Force data now, but the company expects to host classified data from other agencies over time through future accreditations. Thus far, only two companies—Amazon Web Services and Microsoft—have achieved the Defense Information Systems Agency’s Impact Level 6 accreditation to host data at all government classification levels, yet Oracle’s latest accreditation continues the competition among several cloud giants that continually vie for important and lucrative Defense Department and intelligence community contracts.”
Gotta love this me-too innovation. Will Oracle reach the coveted Impact Level 6 accreditation? Perhaps. The company’s Glen Dodson touts the firm’s long history working with the DoD and intelligence community and boasts about its data management, analytics, and AI tools. The write-up reminds us that last year, along with AWS, Microsoft, Google, and IBM, Oracle was awarded part of the CIA’s Commercial Cloud Enterprise contract. That seems like a good sign for the company. Movin’ slow. But movin’.
Cynthia Murrell, March 4, 2022
Clearview Aims to Collect Every Face, and More
March 4, 2022
Chances are, Clearview already has a record of your face. In fact, reports Silicon Republic, “Clearview AI Plans to Put Almost Every Human Face in its Database.” At least that is what it has told its investors, according to documents obtained by The Washington Post. Writer Leigh Mc Gowran reports:
“Clearview AI, which describes itself as ‘the world’s largest facial network’, has built a database that currently holds more than 10bn ‘publicly available facial images’ taken from the web. It works with customers such as law enforcement agencies to compare facial data against its database. The US-based company has said this database is the ‘largest known of its kind in its industry’. A financial presentation the company created last December goes further than this publicly available statement. In this document, Clearview claimed it already has 11 times more facial recognition data than any government or non-government entity today. The facial recognition company claimed to be ‘achieving rapid international expansion’. It said it has more than 3,000 security and law enforcement customers in the US, including the FBI and ICE, according to documents shared by Washington Post tech reporter Drew Harwell on Twitter. … Clearview’s technology roadmap goes even further, with plans to develop services such as licence plate recognition, movement tracking and contactless fingerprint recognition. Last month, Clearview AI announced that it was awarded a US patent for a facial recognition capability that performed ‘nearly flawlessly’ in vendor tests.”
It sounds like the company is on a roll. All this despite increased regulatory pressure in several countries. The ACLU and authorities in Australia, Canada, and the UK have all taken action of one sort or another against the company. Meanwhile, mass biometric surveillance in general is being challenged in the EU. A couple companies have reversed course on the technology—Meta (formerly known as Facebook) pledged to delete the facial recognition data it had collected, and IBM promised to jettison its facial recognition and analysis software. For those firms, however, creepy AI was just part of the mix. Such software is Clearview’s entire game, and it seems determined to forge ahead with no regard for attempts to rein it in.
Cynthia Murrell, March 4, 2022
Google: Defines Excellence for Android Users
March 3, 2022
I read a hoot of a story. “Data Stealing App Found in Google Play Downloaded Thousands of Times.” The idea for branded stores is consistency, compatibility, and trust. No one wants to buy an air fryer that explodes and maims an influencer. Why would one want to download a mobile app which allows a bad actor to seize data or control of one’s mobile device.
The write up reports:
A notorious Android banking trojan designed to steal user data, like passwords and text messages, has been discovered in Google Play and downloaded thousands of times. The TeaBot banking trojan, also known as Anatsa and Toddler, was first observed in May 2021 targeting European banks by stealing two-factor authentication codes sent by text message.
Yep, malware direct from the Google. Let’s rundown those qualities of a branded store:
- Consistency
- Compatibility
- Trust
Check, check, and check.
Ah, Google, are you entering a security drag race against the Softies?
Stephen E Arnold, March 3, 2022
A Meta Burger Surprise? Nope, Seems Like a Standard Operating Procedure
March 3, 2022
I love the thinking of high school science club managers. I think I have spotted an example clearly spelled out in “Facebook Misled Investors on Scope of Misinformation Problems, Whistleblower Says.” The key point seems to be that Meta (maybe meat) says one thing and does another — often with world class ineptitude.
The write up states:
Haugen’s new complaints say that while Facebook/Meta executives trumpeted their efforts to tamp down misinformation about climate change and COVID in earnings calls and elsewhere, internally, the company knew it was falling short.
The saying one thing and doing another approach is okay until a certain someone steps forward and says, “Not so fast.” That someone is Frances Haugen, the former Facebooker turned whistle blower. The secret is that Meta (maybe meat) could not chop liver.
I loved this approach to grilling the Meta outfit:
Using whistleblower complaints to address the misinformation problem is “creative,” Nathaniel Persily, a professor at Stanford Law School and director of the Stanford Cyber Policy Center, told The Washington Post. “You cannot pass a law in the US banning disinformation,” he said. “So what can you do? You can hold the platforms accountable to promises they make. Those promises could be made to users, to the government, to shareholders.” The strategy could work, given many investors’ appetites for focusing on environmental, social, and governance investment strategies (so-called ESG investors). For years, the SEC has told publicly traded companies that they need to make clear and accurate disclosures, Jane Norberg, a partner at Arnold & Porter who recently ran the SEC’s whistleblower program, told the Post. “If the company says one thing to investors but internal documents show that what they were saying is untrue, that could be something the SEC would look at,” she said.
Would the head Meta person pull a sophomoric stunt like obfuscate, fiddle with words, and prevaricate?
Yep, just like 14 years olds explaining the chemistry experiment was not intended to blow up the lab table.
Stephen E Arnold,March 3, 2022
More Print Magazines Jobs Bite Dust As They Go Digital
March 3, 2022
More printed magazines and newspapers are disappearing says Al Jazeera in the article, “Barry Diller-Owned Dotdash Ends Print Editions Of Six Magazines.” Print media is not entirely dead. Newspapers are thinner than most waistlines, phone books can still be found if you look hard enough, and magazines still sell well at stores. Dotdash Meredith will remove some of those magazines from the racks, including the popular Entertainment Weekly and InStyle. Other popular titles are People en Español, Parents, Health, and EatingWell. Dotdash wants to concentrate on digital publishing.
Unfortunately two hundred jobs are cut in the process, which is 5% of the company’s entire workforce. April 2022 will be the final issues for the six aforementioned magazines. Meredith Corp. purchased Dotdash in October 2021, but the purpose was not to buy print media instead Meredith wanted brands. Meredith Corp. is seriously invested in growing and invested $80 million in content for its brands.
Despite people reading more during the pandemic, print media saw decreasing revenues:
“Newspaper and magazine publishers have faced pressure in recent years from a decline in print-advertising revenues, while tech giants like Google and Facebook have pocketed most of the ad dollars in the digital space.
The COVID-19 pandemic has added to the industry’s woes as it accelerated a shift to online news platforms, bringing newsstand sales to a halt.”
Dotdash is the digital arm of IAC and owns fourteen media branches. Its fastest growing branches are health, lifestyle, and finance. The company reaches 100 million online consumers on a monthly basis.
Print is not dead yet, but its digital counterpart is certainly taking over!
Whitney Grace, March 3, 2022
Think You Can Identify a Faux Face? Think Again
March 3, 2022
We have reached the point where people cannot tell an AI-generated human face from the real thing. The only difference, according to a recent study, is that apparently the fake versions look more like someone you could trust. How ironic. The brief report from the Proceedings of the National Academy of Sciences is titled, “AI-Synthesized Faces are Indistinguishable from Real Faces and More Trustworthy.” The report includes images of real and synthetic faces used in the study, so curious readers may want to check those out. The abstract states:
“Artificial intelligence (AI)–synthesized text, audio, image, and video are being weaponized for the purposes of nonconsensual intimate imagery, financial fraud, and disinformation campaigns. Our evaluation of the photorealism of AI-synthesized faces indicates that synthesis engines have passed through the uncanny valley and are capable of creating faces that are indistinguishable—and more trustworthy—than real faces.”
Later, the authors add:
“Perhaps most pernicious is the consequence that, in a digital world in which any image or video can be faked, the authenticity of any inconvenient or unwelcome recording can be called into question.”
Indeed. So no longer can we trust our own eyes (or ears), and the result of doing so (or not) could be devastating for the target(s). Though some techniques have been developed to auto-detect deep fakes, none can yet keep up with the massive amount of faked content uploaded daily. The researchers set out to see whether humans could detect the truth for themselves, perhaps searching for a glimmer of hope. If so, they were disappointed. See the report for details of their methodology and results.
The authors describe the popular method of synthesizing images they used in their experiments: generative adversarial networks (GANs). We learn:
“A GAN pits two neural networks—a generator and discriminator—against each other. To synthesize an image of a fictional person, the generator starts with a random array of pixels and iteratively learns to synthesize a realistic face. On each iteration, the discriminator learns to distinguish the synthesized face from a corpus of real faces; if the synthesized face is distinguishable from the real faces, then the discriminator penalizes the generator. Over multiple iterations, the generator learns to synthesize increasingly more realistic faces until the discriminator is unable to distinguish it from real faces.”
The authors note their findings are consistent with those of two previous studies, though this one goes further to include racial and gender diversity and to minimize the chance of inadvertent cues. They also point out how easy it is for anyone to access GAN technology and advise the graphics community to develop ethical guidelines for content creators, researchers, publishers, and media distributors. Posthaste, please.
The study was authored by Sophie J. Nightingale of Lancaster University’s Psychology Department and Hany Farid of Department of Electrical Engineering and Computer Sciences at the University of California, Berkeley.
Cynthia Murrell, March 3, 2022
Microsoft and Security: A Probably Trivial Item
March 2, 2022
An online publication called Venture Beat published “Russia May Use SolarWinds-Like Hacks in Cyberwar over Ukraine.” The article contained a paragraph I found suggestive. Here’s the passage:
…the attackers are believed to have gained access for as much as nine months to numerous companies and government agencies, including FireEye, Microsoft and the Departments of Defense, State and Treasury.
The point for me is that the extent of the breaches is not fully known. It is easier to issue news releases and make high-profile marketing moves than come to grips with the allegedly accurate information in the Venture Beat article.
Stephen E Arnold,March 2, 2022
Interesting Chatbot Findings
March 2, 2022
I am an old geezer. I remember when one could call or stop by a store and ask a humanoid a question. If that humanoid did not know the answer, there was a chance that the clueless humanoid could ask another allegedly more informed humanoid over to consider my question.
Not any more. I have to figure out how to find a company online. This is not often easy for a wide range of reasons. Then I have to locate the “contact” button. Then I have to identify crosswalks or work a math problem. And then I can fill out a form? At no point in the process is a humanoid offered as a solution.
Many, many Web sites are embracing smart software which either talk like the weird pitches on junk telemarketing calls or through a pop up box into which one types. The smart software on the receiving end determines how to answer. Some of the really nifty approaches include referring me to the firm’s Web site again or displaying a list of links which the smart software has determined with 55 percent confidence will address my issue. Sure, sure.
I was surprised when I read “Report: Two-Thirds of Consumers Would Rather Use a Chatbot Than Browse a Website.” Here’s the key statement in the write up:
According to Botco.ai, not only do eight out of 10 people say they’ve used a chatbot in the past, but the vast majority of consumers — 70% — say that chatbots typically answer all or most of their questions satisfactorily.
And who sponsored the study? And what does the sponsor do to make money? You get one guess.
Did you say, “This outfit is in the chatbot business”? If so, you are sort of correct.
Just one question: How does one locate and use a chatbot?
Did you say, “On a company’s Web site”? If so, you are right again.
Do I have confidence in this finding?
Nope. Why not label this type of document “Marketing”?
Stephen E Arnold, March 2, 2022
-
- Subscribe to Beyond Search
Feature archive
News archive
-
