An Epidemic of Whistle Blowing?
October 22, 2021
Are organizations prepared for an epidemic of whistle blowing?
This question struck me as I read “How One Facebook Worker Unfriended the Giant Social Network.” Here’s the statement in the article which caught my attention:
“There has just been a general awakening amongst workers at the tech companies asking, `What am I doing here?’” said Jonas Kron of Trillium Investment Management, which has pushed Google to increase protection for employees who raise the alarm about corporate misdeeds. “When you have hundreds of thousands of people asking that question, it’s inevitable you’ll get more whistleblowing,” he said.
The comment touched upon two issues which I don’t think have been resolved.
The first is the “awakening.” The idea that workers are “woke” is interesting. My reaction is that the flood of information about social unraveling, breakdowns in what were supposed to be reliable services, and the waves of disturbing news have broken down the “I’m entitled” Drosophila in these folks’ brains. Woke is not a good word. I think something along the lines “I understand now” is more accurate.
The second is the statement that a particular individual who allegedly “has pushed Google to increase protection for employees who raise the alarm about corporate misdeeds.” Okay, that’s interesting. How is that working out for those of the Timnit Gebru ilk?
Net net: Whistle blowers can present different reasons for their actions. The write up makes clear that the “cult of me” is alive and well. Some “me’s” are into dumping documents and information which are confidential. These actions take place even though the person has signed an agreement to keep an organization’s data secret. Guess that piece of paper is not working too well, right? Plus, an investment professional urging the Google to alter its DNA is a helpful endorsement of an individual’s valiant effort to induce change and get some good vibes for the action. Whistle blowing may be little more than an extension of an individual’s need to be the nail that sticks up.
Stephen E Arnold, October 22, 2021
The Bezos Bulldozer and One of Its Charming Quirks
October 22, 2021
Amazon is the Bezos bulldozer. I know. I know. He’s into space and making the world better. Nevertheless, the “trust” outfit Reuters is not buying the PR. “Amazon Copied Products and Rigged Search Results to Promote Its Own Brands, Documents Show” provides an interesting look at Amazon’s ecommerce business strategy.
The write up asserts:
… Thousands of pages of internal Amazon documents examined by Reuters – including emails, strategy papers and business plans – show the company ran a systematic campaign of creating knockoffs and manipulating search results to boost its own product lines in India, one of the company’s largest growth markets. The documents reveal how Amazon’s private-brands team in India secretly exploited internal data from Amazon.in to copy products sold by other companies, and then offered them on its platform.
Navigate to the source document for quotes, names of bulldozer drivers, and the specifics of the retail ants crushed under the steel tracks of the snorting behemoth.
Why would Amazon copy and boost its own products?
Gee, that’s a tough question. Pick from these possible reasons:
[a] Executive compensation incentives engineer rapacious methods into the ecommerce processes
[b] Because Amazon could. Hey, what’s power for if one doesn’t use it.
[c] Increasing profit results in higher stock prices and juicier bonuses for high-performing Amazon professionals
[d] It’s fun because business is a game
[e] The companies and products are little more than tests for Amazon. Follow the data.
I like the “It’s fun” answer. Because business is a game to be won.
Stephen E Arnold, October 22, 2021
What Can Slow Down the GOOG? Lawyers Reviewing AI Research Papers
October 21, 2021
I spotted an allegedly true factoid in “Google’s AI Researchers Say Their Output Is Being Slowed by Lawyers after a String of High Level Exits : Getting Published Really Is a Nightmare Right Now.” Here is the paywalled item:
According to Google’s own online records, the company published 925 pieces of AI research in 2019, and a further 962 in 2020. But the company looks to have experienced a moderate slowdown this
year, publishing just 618 research papers in all of 2021 thus far.
Quite a decrease, particularly in the rarified atmosphere of the smartest people in the world who want to be in a position to train, test, deploy, and benefit from their smart software.
With management and legal cooks in the Google AI kitchen, the production of AI delicacies seems to be going down. Bad for careers? Good for lawyers? Yes and yes.
Is this a surprise? It depends on whom one asks.
At a time when there is chatter that VCs want to pump money into smart software and when some high profile individuals suggest China is the leader in artificial intelligence, the Google downturn in this facet of research is not good news for the GOOG.
Is there a fix? Sure, but none is going to include turning back the hands of time to undo what I call the Battle of Timnit. The decision to try and swizzle around the issue of baked in algorithmic bias appears to have blocked some Google researchers’ snorkels. Deep dives without free flowing research oxygen can be debilitating.
Stephen E Arnold, October 21, 2021
Is Self Driving Ready for Regular Humanoids?
October 21, 2021
Teslas are popular and expensive cars. Tesla owners love their cars with the same fervor as Prius owners, except with a more elitist attitude. Fears of Elon Musk acting as Big Brother have been placated, but Electrek shares how that subject comes into question again in: “Tesla Will Make Sure You Are A Good Driver Before Giving You Access To Fill Self-Driving Beta.”
Musk said that his company will use telemetry data to guarantee its customers are “good” drivers before giving them access to the cars’ self-driving option. That is a problematic approach, because Tesla owners already paid for the software. In September, Tesla released an update to its Full Self-Driving Beta v10 software. The software realizes the dream of cars self-driving, however, drivers are still required to pay attention at all times and keep their hands on the steering wheel.
Teslas still have bugs with its self-driving feature, but Musk promises the upgrade is “mind-blowing.” But Musk only wants “good” drivers with a high safety record to use the beta. Since Teslas are linked to a “hive mind,” Musk has access to their driving data. The “good” driving requirement is a way for Musk to prevent accidents and deaths, but it begs the question if it is legal.
Insurance companies already monitor their customers with safe driving applications to receive discounts. Law enforcement also install breathalyzers in cars to prevent drunk driving. Limiting the self-driving beta is in the same vein, but the pros and cons must be investigated.
It also brings Musk’s intentions into question. Will he take responsibility if a Tesla terminates an annoying humanoid?
Whitney Grace, October 21, 2021
China, Smart Software, and Different Opinions
October 21, 2021
I spotted “China Isn’t the AI Juggernaut the West Fears.” The main idea for the story is that China has cornered smart software applications and innovation. Therefore, the future — at least some of it — is firmly in the grip of the Chinese Communist Party.
My hunch is that this article in the Japan Times is a response to articles like “Former Senior Pentagon Official Says China is Kicking Our Ass in Artificial Intelligence.” Nicolas Chaillan, a former Pentagon official, suggested that China is making significant progress in AI. If China continues on its present path, that country may surpass the US and its allies in smart software.
What’s interesting is that quite different viewpoints are zooming around the interwebs.
The Japan Times’ take which channels Bloomberg includes this statement:
On paper, the U.S. and China appear neck and neck in artificial intelligence. China leads in the share of journal citations — helped by the fact that it also publishes more — while the U.S. is far ahead in the more qualitative metric of cited conference papers, according to a recent report compiled by Stanford University. So while the world’s most populous country is an AI superpower, investors and China watchers shouldn’t put too much stock in the notion that its position is unassailable or that the U.S. is weaker. By miscalculating the others’ abilities, both superpowers risk overestimating their adversary’s strengths and overcompensating in a way that could lead to a Cold War-style AI arms race.
Yep, citation analysis.
I don’t have a dog in this fight. I want to point out that citation analysis, like patent documents, may not tell a comprehensive story.
I would suggest that citation analysis may be distorted by the search engine optimization techniques used by some academics and government-funded researchers. In addition, the publication flow from what I call AI cabals — loose federations of like minded researchers who cross cite one another — provide a fun house mirror opportunity.
That is, what’s reflected is a version of reality, not the reality that a person like myself would perceive without the mirrors.
Net net: The Japan Times’ write up may be off the mark. As a result, the view point of Nicolas Chaillan may warrant serious consideration.
Stephen E Arnold, October 21, 2021
Google Launches Even More Personalized Search Upgrade
October 21, 2021
Google is already the most used search engine on the planet and delivers fairly accurate results. Like many companies, Google continues to push innovation and The National News shares the latest search upgrade in: “Google To Introduce Search 1,000 Times More Powerful Than Current Engine.” Google’s new search technology leverages AI that combines search criteria for more personalized and accurate results.
Google revealed its latest search achievement at the Search On ’21 event, where executives discussed how they plan to use their AI research to stop the spread of misinformation and make information on the Web more useful. Google also wants to regain shopping traffic from Amazon, Alibaba, Lazada, and other commerce Web sites. The new search technology aims to improve the shopping search experience:
“Google teased the MUM technology during its annual I/O summit last May. It uses its so-called T5 – Text-To-Text Transfer Transformer – framework and is said to be 1,000 times more powerful than the Bert (Bidirectional Encoder Representations from Transformers) technology the company currently uses.
The revamped search technology, using the company’s image-recognition tool Google Lens, will combine data from text, images and even videos, which would then provide more accurate and tailor-made results. Lens has been updated with new AI-powered language features that will narrow searches further. ‘For example, when you search for ‘cropped jackets’, we’ll show you a visual feed of jackets in various colors and styles, alongside other helpful information like local shops, style guides and videos,’ Bill Ready, president of commerce, payments and next billion users at Google, said.”
Google will also include a “wildfire layer” on its Maps to keep track of forest fires in real time. To combat misinformation, search results will include an “About This Result”option that cites the result’s sources and what others users think of it.
Google designed a picture search engine for shopping and is actually citing sources for search results? Yes, please!
Whitney Grace, October 20, 2021
Microsoft: A Legitimate Point about Good Enough
October 20, 2021
A post by Stefan Kanthak caught my attention. The reason was an assertion that highlights what may be the “good enough” approach to software. The article is “Defense in Depth — the Microsoft Way (Part 78): Completely Outdated, Vulnerable Open Source Component(s) Shipped with Windows 10&11.” I am in the ethical epicenter of the US not too far from some imposing buildings in Washington, DC. This means I have not been able to get one of my researchers to verify the information in the Stefan Kanthak post. I, therefore, want to point out that it may be horse feathers.
Here’s the point I noted in the write up:
Most obviously Microsoft’s processes are so bad that they can’t build a current version and have to ship ROTTEN software instead!
What’s “rotten”?
The super security conscious outfit is shipping outdated versions of two open source software components: Curl.exe and Tar.exe.
If true, Stefan Kanthak may have identified another example of the “good enough” approach to software. If not true, Microsoft is making sure its software is really super duper secure.
Stephen E Arnold, October 20, 2021
Mapping the Earth: A Big Game?
October 20, 2021
I read “Was Google Earth Stolen?” I have not thought about making a map of the earth game-like for many years. I read the article by Avi Bar-Zeev, one of the individuals close to the Keyhole approach. Interesting stuff.
I want to underscore the fact that Microsoft was noodling around in this geographic earth space as well. There is a short item on the Microsoft Web site called “The Microsoft TerraServer.” The write up states:
The Microsoft TerraServer stores aerial and satellite images of the earth in a SQL Server Database served to the public via the Internet. It is the world’s largest atlas, combining five terabytes of image data from the United States Geodetic Survey, Sovinformsputnik, and Encarta Virtual Globe™. Internet browsers provide intuitive spatial and gazetteer interfaces to the data. The TerraServer demonstrates the scalability of Microsoft’s Windows NT Server and SQL Server running on Compaq AlphaServer 8400 and StorageWorks™ hardware. The TerraServer is also an E-Commerce application. Users can buy the right to use the imagery using Microsoft Site Servers managed by the USGS and Aerial Images. This paper describes the TerraServer’s design and implementation.
The link to download the 23 year old Microsoft document is still valid, believe it or not!
Other outfits were into fancy maps as well; for example, the US government entity in Bethesda and some of the folks at Boeing.
Is this germane to the Bar-Zeev write up? Nah, probably no one cares. I find stories about technology “origins” quite interesting for what each includes and what each omits. Quite game-like, right?
Stephen E Arnold, October 20, 2021
Enterprise Search: Will It Prove Sweet for MeiliSearch?
October 20, 2021
Developers looking for a search solution may want to check out MeiliSearch. The customizable search engine recently released a new version that boasts a refined indexer, customizable ranking rules, and a sort function that works as users type their query. Their website promises:
“For developers: Scalable, maintainable, customizable. MeiliSearch provides an extensive toolset for customization. Unlike with other search engines, these customization options are just that: optional. It works out-of-the-box with a preset that easily answers the needs of most applications. Communication is done with a RESTful API because most developers are already familiar with its norms.
And we noted:
“For users: Blazing fast, relevant, typo tolerant. The search experience feels simple and intuitive. It’s all too common for search bars to make users feel like they have to learn a new language just to get the best results – or worse, that they have to jump back and forth between their search and Google just to get the right spelling or product UID.MeiliSearch makes searching simple and responsive, so the user can stay focused on the results.”
The company’s blog post, “What’s New in v0.22,” emphasizes the new sort-as-you-type feature. Writer Gui Machiavelli tells us one could previously achieve a similar result with custom ranking rules applied during index configuration, but this feature makes it easier. MeiliSearch is an open source project. The company was founded in 2018 and is based in Paris, France.
Cynthia Murrell October 20, 2021
No Click Excitement: Interaction-Less Vulnerabilities in Messaging Apps
October 20, 2021
Google researcher Natalie Silvanovich has made it her mission to investigate one particular type of vulnerability—one that allows attackers to access video and/or audio without the victim so much as clicking a link. Wired discusses her unnerving findings in, “Messaging Apps Have an Eavesdropping Problem.” Writer Lily Hay Newman tells us:
“Silvanovich has spent years studying “interaction-less” vulnerabilities, hacks that don’t require their targets to click a malicious link, download an attachment, enter a password in the wrong place, or participate in any way. Those attacks have taken on increasing significance as targeted mobile surveillance explodes around the world.”
The resolute researcher presented her findings at the recent Black Hat security conference in Las Vegas. Her search turned up bugs in apps domestic and foreign, from Facebook Messenger, Google Duo, and Signal to JioChat and Viettel Mocha. The vulnerabilities she found were eagerly patched by the respective developers once she notified them, but her discoveries reveal a problem more widespread than had been suspected. It seems that some of the vulnerabilities resulted from honest mistakes by developers using the open source communication tool WebRTC. Other times, though, it had to do with how an app connects calls. We learn:
“When someone calls you on an internet-based communication app, the system can start setting up the connection between your devices right away, a process known as ‘establishment,’ so the call can start instantly when you hit accept. Another option is for the app to hang back a bit, wait to see if you accept the call, and then take a couple of seconds to establish the communication channel once it knows your preference. … Most mainstream services take the other route, though, setting up the communication channel and even starting to send data like audio and video streams in advance to offer a near-instantaneous connection should the call’s recipient pick up. Doing that prep work doesn’t inherently introduce vulnerabilities, and it can be done in a privacy-preserving way. But it does create more opportunities for mistakes.”
Concerned users may want to favor Telegram—Silvanovich found that app takes the slower but safer route. Though the snippets hackers can capture with these vulnerabilities may or may not be valuable, many find it worth a try—such attacks are difficult to detect and to trace. Careful design and implementation on the part of app developers are the keys to avoiding such breaches, she tells us.
Cynthia Murrell October 20, 2021
-
- Subscribe to Beyond Search
Feature archive
News archive
-
