Meta Zuck: AIR SC Sort of Sketched Out

January 25, 2022

I read Facebook’s (Meta’s) blog post called “Introducing the AI Research SuperCluster — Meta’s Cutting-Edge AI Supercomputer for AI Research.” The AIR SC states:

Today, Meta is announcing that we’ve designed and built the AI Research SuperCluster (RSC) — which we believe is among the fastest AI supercomputers running today and will be the fastest AI supercomputer in the world when it’s fully built out in mid-2022.

Then this statement:

Ultimately, the work done with RSC will pave the way toward building technologies for the next major computing platform — the metaverse, where AI-driven applications and products will play an important role.

So the AIR SC is sort of real. The applications for the AIR SC are sort of metaverse. That’s not here either in my opinion.

So what’s going on? Here are my thoughts:

  1. Facebook wants to stake out conceptual territory claims as AT&T did with its non 5G announcements about the under construction 5G capabilities.
  2. Facebook wants to show that its AIR SC is bigger, better, faster, and more super than anything from the Amazon, Google, or other quasi-monopolies who want systems that will dominate the super computer league table for now and possibly forever unless government regulators or user behavior changes the game plan.
  3. Facebook believes the Silicon Valley marketing mantra, “Fake it until you make it” with a possible change. I interpret the announcement to say, “Over promise and under deliver.” I admit I have become jaded with the antics of these corporate giants who have been able to operate without meaningful oversight or what some might call ethical guidelines for a couple of decades.

In the old days, companies in the Silicon Valley mode did vaporware. The tradition continues? Sure, why not? There’s even a TikTok style video to get the AIR SC message across.

Stephen E Arnold, January 25, 2022

European Parliament Embraces the Regulatory PEZ Dispenser Model for Fines on Big Tech

January 24, 2022

I read about the Digital Services Act. “European Parliament Passes Huge Clampdown on Tracking Ads” states:

The European Parliament, the legislative body for the European Union (EU), has voted in favor of its Digital Services Act (DSA), which seeks to limit the power of American internet giants such as Facebook, Amazon and Google.

That’s mostly on the money. What’s not spelled out is that the procedure of identifying a tracking instance, building a case, adjudicating, appealing, and levying a fine is now official. It’s a procedure. Perhaps a bright French artificial intelligence professional will use Facebook or Google AI components to make the entire process automatic, efficient, and – obviously – without bias. No discrimination! But the DSA is aimed at outfits like Amazon, Facebook, and Google. Nope. Not discriminatory and also not yet a really official thing…yet.

I found this paragraph memorable:

According to the EU, the DSA covers several key areas, including introducing mechanisms by which companies have to remove “illegal” content in a timely manner in a bid to reduce misinformation, increasing requirements on so-called very large online platforms (VLOPs), regulating online ad targeting, and clamping down on dark patterns. The scope and scale of the DSA (and associated DMA) are huge, perhaps the biggest effort yet by a substantial world power (outside of China) to regulate what happens in cyberspace.

How does one redistribute “wealth”? Easy. Create a legal PEZ dispenser, push the plastic likeness of Mr. Bezos, Mr. Zuckerberg, or Mr. Pichai (who is the only one of the PEZ dispensers with AI in his name).

Stephen E Arnold, January 24, 2022

A News Blog Complains about Facebook Content Policies

January 20, 2022

Did you know that the BMJ (in 1840 known as the Provincial Medical and Surgical Journal and then after some organizational and administrative cartwheels emerged in 1857 as the British Medical Journal? Now the $64 question, “Did you know that Facebook appears to consider the BMJ as a Web log or blog?” Quite a surprise to me and probably to quite a few others who have worked in the snooty world of professional publishing.

The most recent summary of the dust up between the Meta Zuck outfit and the “news blog” BMJ appears in “Facebook Versus The BMJ: When Fact Checking Goes Wrong.” The write up contains a number of Meta gems, and a read of the “news blog” item is a good use of time.

I want to highlight one items from the write up:

Cochrane, the international provider of high quality systematic reviews of medical evidence, has experienced similar treatment by Instagram, which, like Facebook, is owned by the parent company Meta. A Cochrane spokesperson said that in October its Instagram account was “shadow banned” for two weeks, meaning that “when other users tried to tag Cochrane, a message popped up saying @cochraneorg had posted material that goes against ‘false content’ guidelines” (fig 1). Shadow banning may lead to posts, comments, or activities being hidden or obscured and stop appearing in searches. After Cochrane posted on Instagram and Twitter about the ban, its usual service was eventually restored, although it has not received an explanation for why it fell foul of the guidelines in the first place.

I like this shadow banning thing.

How did the Meta Zuck respond? According to the “news blog”:

Meta directed The BMJ to its advice page, which said that publishers can appeal a rating directly with the relevant fact checking organization within a week of being notified of it. “Fact checkers are responsible for reviewing content and applying ratings, and this process is independent from Meta,” it said. This means that, as in The BMJ’s case, if the fact checking organization declines to change a rating after an appeal from a publisher, the publisher has little recourse. The lack of an independent appeals process raises concerns, given that fact checking organizations have been accused of bias.

There are other interesting factoids in the “news blog’s” write up.

Quickly, several observations:

  1. Opaque actions plague the “news blog”, the British Medical Journal and other luminaries; for example, the plight of the esteemed performer Amouranth of the Inflate-a-Pool on Amazon Twitch. Double talking and fancy dancing from Meta- and Amazon-type outfits just call attention to the sophomoric and Ted Mack Amateur Hour approach to an important function of a publicly-traded organization with global influence.
  2. A failure of “self regulation” can cause airplanes to crash and financial disruption to occur. Now knowledge is the likely casualty of a lack of a backbone and an ethical compass. Right now I am thinking of a ethics free, shape shifting octopus like character with zero interest in other creatures except their function as money generators.
  3. A combination of “act now, apologize if necessary” has fundamentally altered the social contract among corporations, governments, and individuals.

So now the BMJ (founded in 1840) has been morphed into a “news blog” pitching cow doody?

Imposed change is warranted perhaps? Adulting is long overdue at a certain high-tech outfit and a number of others of this ilk.

Stephen E Arnold, January 20, 2022

UK Legal Eagles Circle the Facebook ATM

January 17, 2022

I read “Facebook Parent Meta Faces $3.1B UK Class Action for Breach of Competition Law.” In a blistering rapid response to the Cambridge Analytica misstep, legal eagles have realized that Facebook may have taken some liberties with the notion of market dominance, information collection, and downstream use of those data about its much loved customers.

The article states (once one gets past the pop up ads from everyone’s favorite moguls at the News Corp.):

The suit. If successful, would have Facebook paying $3.1 billion in damages to Facebook U.K. users. The lawsuit was filed with the U.K.’s Competition Appeal Tribunal in London.

From my point of view, the action is a way to get a large, US technology company to output a bale of cash for its behaviors. Although $3.1 billion is a respectable number, Facebook’s approximate daily revenue intake is in the neighborhood of a $100 billion, give or take a few billion. That works out to a week and a half of revenue.

Bad but not that bad. Will the treasured Facebook customers get the money if the class action suit prevails? Sure, absolutely just after fees and other costs. Such a deal and one that will definitely chasten Facebook, Meta, whatever.

Stephen E Arnold, January 17, 2022

Alleged Collusion Between Meta and Google: Shocking Sort Of

January 17, 2022

Google and Facebook’s Top Execs Allegedly Approved Dividing Ad Market among Themselves” reports:

The alleged 2017 deal between Google and Facebook to kill header bidding, a way for multiple ad exchanges to compete fairly in automated ad auctions, was negotiated by Facebook COO Sheryl Sandberg, and endorsed by both Facebook CEO Mark Zuckerberg (now with Meta) and Google CEO Sundar Pichai, according to an updated complaint filed in the Texas-led antitrust lawsuit against Google.

Fans of primary research can read the 242 page amended filing at this link.

One question arises: How could two separate companies engage is discussions to divide a market? Perhaps one clue is the presence of the estimable lean in professional Sheryl Sandberg, who joined Google 2001 after blazing a trail in economics, McKinsey-type thinking familiar to many today as the pharma brain machine, and then some highly productive US government work.

At the Google she was a general manager. Her Googley behavior earned her a promotion. She was one of the thinkers shaping the outstanding revenue generation system known as AdWords. She added her special touch of McKinsey-ness to AdSense to the Gil Ebaz smart system packaged as Applied Semantics aka Oingo. The important point about applied semantics is that the technology included what I think of as steering or directionality; that is, one uses semantic information to herd the doggies (users) down the trail (consumption of ad inventory. For more on this notion of steering yo8u will want to listen to my interview with Dr. Donna Ingram who addresses this issue in the DarkCyber, 4th series, Number 1 video program to be released on January 18, 2022.

In 2007, chatting at the party helped her migrate from the Google to the company formerly known as Facebook. Ms. Sandberg, Harvard graduate with a chubby contact list, joined the scintillating management team as the social network engineering machine. In 2012, she became a member of the company’s board of directors. She leaned in to her role until some “real” news outfits flipped over the mossy rock of Cambridge Analytica’s benchmark marketing methods.

Ms. Sandberg was recognized by Professor Shoshana Zuboff as the Typhoid Mary of surveillance capitalism. Is that a Meta T shirt yet? He book is a must read. It is called Lean In: Women, Work, and the Will to Lead. It appeared in 2013 and may be due for an update to include the Cambridge Analytica misunderstanding, the Frances Haugen revelations, and, of course, the current Texas-sized legal matter.

The write up cited above points out a statement from the Google. The main idea is that the idea is “full of inaccuracies and lacks legal merit.”

I believe everything I read on the Internet. I accept the Google search output when I query “Silicon Valley ethics” – Theranos. I trust in the Meta thing because how could two outfits collude? I think such interactions are highly improbable in Silicon Valley, the home of straight shooting.

Stephen E Arnold, January 17, 2022

Facebook: Be Proud

December 31, 2021

Gee, who could have predicted it? MacRumors reports, “Facebook Crowned ‘Worst Company of 2021’ by New Survey.” Each December Yahoo Finance picks a Company of the Year based on factors like achievements and market performance, and stalwart Microsoft has come out on top for 2021. The less coveted worst company designation is determined by a survey of Yahoo Finance’s audience. Writer Sami Fathi reports:

“According to the survey, which polled over 1,000 individuals, Facebook received 50% more votes for the spot compared to Alibaba, a Chinese e-commerce platform. Those surveyed have a ‘litany of grievances’ towards Facebook, including but not limited to concerns over censorship, reports about Instagram’s impact on mental health, and privacy. While the survey results are grim and not in the company’s favor, 30% of the participants responded positively to whether Facebook could ‘redeem itself.’ … Facebook has long been embroiled in public concerns over the privacy of its users. Facebook has notably fought with Apple over changes in iOS and iPadOS that make it harder for companies to track users across other apps and websites. Facebook has claimed the new change, App Tracking Transparency, would hurt small businesses that rely on advertising to attract new customers. Mark Zuckerberg has gone as far as to say that Facebook’s lackluster growth in the last quarter of the year was partly to blame on ATT (App Tracking Transparency).”

For his part, Apple CEO Tim Cook flings criticism right back at the Zuck. Which is worse—giving up a little revenue or fomenting polarization and violence? I suppose it depends on one’s perspective. The company formerly known as Facebook is hoping its Meta rebrand will distract everyone from its woes, but going by this survey that has yet to happen.

Cynthia Murrell, December 31, 2021

Facebook: Making Friends in the USAF

December 31, 2021

This is a short post sparked by this Financial Times’ article: “Facebook to Build Metaverse with Start-Up That Had US Military Contracts.”

The main idea is that Facebook bought a company. The firm — Reverie— will work with Meta Facebook thing’s Reality Labs. But the bonus move is that the Meta Facebook thing was terminated when the Meta Facebook thing bought Reverie. The venerable and generally respectable Financial Times pointed out that the Meta Facebook thing would “not be involved with any future defense or military AI development.”

Okay. My hunch is that a Meta Facebook thing employee whose child seeks to enter the Air Force Academy may find that some of those involved in the selection process may remember this “not be involved with any future defense or military AI development.”

Who likes this type of business decision? Maybe the Chinese and Russian military leadership? But that’s just a thought from the wilds of rural Kentucky. The Meta Facebook thing knows what’s best for itself and, of course, the US government.

Stephen E Arnold, December 31, 2021

Reputation Repair Via Content Moderation? Possibly a Long Shot

December 30, 2021

Meta (formerly known as Facebook) is launching another shot in the AI war. CNet reports, “Facebook Parent Meta Uses AI to Tackle New Types of Harmful Content.” The new tool is intended to flag posts containing misinformation and those promoting violence. It also seems designed to offset recent criticism of the company, especially charges it is not doing enough to catch fake COVID-19 news.

As Meta moves forward with its grand plans for the metaverse, it is worth noting the company predicts this tech will also work on complex virtual reality content. Eventually. Writer Queenie Wong tells us:

“Generally, AI systems learn new tasks from examples, but the process of gathering and labeling a massive amount of data typically takes months. Using technology Meta calls Few-Shot Learner, the new AI system needs only a small amount of training data so it can adjust to combat new types of harmful content within weeks instead of months. The social network, for example, has rules against posting harmful COVID-19 vaccine misinformation, including false claims that the vaccine alters DNA. But users sometimes phrase their remarks as a question like ‘Vaccine or DNA changer?’ or even use code words to try to evade detection. The new technology, Meta says, will help the company catch content it might miss. … Meta said it tested the new system and it was able to identify offensive content that conventional AI systems might not catch. After rolling out the new system on Facebook and its photo-service Instagram, the percentage of views of harmful content users saw decreased, Meta said. Few-Shot Learner works in more than 100 languages.”

Yep, another monopoly type outfit doing the better, faster, cheaper thing while positioning the move as a boon for users. Will Few-Shot help Meta salvage its reputation?

Cynthia Murrell, December 30, 2021

Facebook Innovates: Beating Heart Emojis

December 29, 2021

I could not resist citing this write up: “WhatsApp Working on Animated Heart Emojis for Android, iOS: Report.” What’s the big news for 2022 from the most loved, oops sorry, worst company in the United States? Here’s the answer according to Gadgets360:

WhatsApp is reportedly planning to add animation to all the heart emojis of various colors for Android and iOS. This could be linked to the message reaction feature that the platform is said to be working on. The feature has been already added to WhatsApp Web/ Desktop via a stable update.

The beating hearts are chock full of meaning. The pulsing image files provide notification information. The compelling news story added:

WhatsApp is rumored to allow users to react to a specific message in a chat with specific emojis. There is also a reaction info tab to show who reacted to a message. Message reactions are reported to be rolled out to individual chat threads and group chat threads.

Definitely impressive. What use will bad actors using WhatsApp for interesting use cases find for pulsing hearts or other quivering emojis?

Stephen E Arnold, December 29, 2021

Meta Mark Gets an F from the British Medical Journal

December 20, 2021

I don’t know anything about Covid, medical data, or Facebook. I do recognize a failing “mark” when I see one. I noted “Researcher Blows the Whistle on Data Integrity Issues…” [Note: the editor has trimmed certain stop words because trigger warning software is a fascinating part of life these days.’’]

The Harvard drop out who has garnered a few dollars via a “friend”, “like”, and “social online” service is unlikely to be personally affected by the big red F.

The write up states:

We are aware that The BMJ is not the only high quality information provider to have been affected by the incompetence of Meta’s fact checking regime. To give one other example, we would highlight the treatment by Instagram (also owned by Meta) of Cochrane, the international provider of high quality systematic reviews of the medical evidence.[3] Rather than investing a proportion of Meta’s substantial profits to help ensure the accuracy of medical information shared through social media, you have apparently delegated responsibility to people incompetent in carrying out this crucial task. Fact checking has been a staple of good journalism for decades. What has happened in this instance should be of concern to anyone who values and relies on sources such as The BMJ. We hope you will act swiftly: specifically to correct the error relating to The BMJ’s article and to review the processes that led to the error; and generally to reconsider your investment in and approach to fact checking overall.

I was disappointed to see the letter’s close; that is, “best wishes.” A more British expression could have been “Excuse me.” But excusing a stupid “mark” is impolite.

Stephen E Arnold, December 20, 2021

« Previous PageNext Page »

  • Archives

  • Recent Posts

  • Meta