AI Cybersecurity: Good News and, of Course, Bad News
October 23, 2023
This essay is the work of a dumb humanoid. No smart software required.
Life, like a sine wave, is filled with ups and downs. Nothing strikes me like the ups and downs of AI: Great promise but profits, not yet. Smart cyber security methods? Same thing. Ups and downs. Good news then bad news. Let’s look at two examples.
First, the good news. “New Cyber Algorithm Shuts Down Malicious Robotic Attack” reports:
Researchers have designed an algorithm that can intercept a man-in-the-middle (MitM) cyberattack on an unmanned military robot and shut it down in seconds. The algorithm, tested in real time, achieved a 99% success rate.
Is this a home run. 99 percent success rate. Take that percentage, some AI, and head to a casino or a facial recognition system. I assume I will have to wait until the marketers explain this limited test.
“Hello, we are the team responsible for infusing AI into cyber security safeguards. We are confident that our technology will have an immediate, direct impact on protecting your organization from threats and bad actors,” says Mary, a lawyer and MBA. I believe everything lawyers and MBAs say, even more than Tom, the head of marketing, or Ben, the lead developer who loves rock climbing and working remotely. Thanks, Bing Dall-e. You understand the look and feel of modern cyber security teams.
Okay, the bad news. A cyber security outfit named Okta was unable to secure itself. You can the allegedly real details from “Okta’s Stock Slumps after Security Company Says It Was Hacked.” The write up asserts:
Okta, a major provider of security technology for businesses, government agencies and other organizations, said Friday that one of its customer service tools had been hacked. The hacker used stolen credentials to access the company’s support case management system and view files uploaded by some customers, Okta Chief Security Officer David Bradbury disclosed in a securities filing. Okta said that system is separate from its main client platform, which was not penetrated.
Yep, the “main client platform” is or was secure.
Several observations:
- After Israel’s sophisticated cyber systems failed to detect planning and preparing for a reasonably large scale attack, what should I conclude about sophisticated cyber security systems? My initial conclusion is that writing marketing collateral is cheaper and easier then building secure systems.
- Are other cyber security firms’ systems vulnerable? I think the answer may be, “Yes, but lawyer and MBA presidents are not sure how and where?”
- Are cost cutting and business objectives more important than developing high reliability cyber security systems? I would suggest, “Yes. What companies say about their products and services is often different from that which is licensed to customers?
Net net: Cyber security may be a phrase similar to US telecommunications’ meaning of “unlimited.”
Stephen E Arnold, October 27, 2023
Gallup on Social Media: Just One, Tiny, Irrelevant Data Point Missing
October 23, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-24.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
I read “Teens Spend Average of 4.8 Hours on Social Media Per Day.” I like these insights into how intelligence is being whittled away.
Me? A dumb bunny. Thanks, MidJourney. What dumb bunny inspired you?
Three findings caught may attention and one, tiny, irrelevant data point I noticed was missing. Let’s look at three of the hooks snagging me.
First, the write up reveals:
Across age groups, the average time spent on social media ranges from as low as 4.1 hours per day for 13-year-olds to as high as 5.8 hours per day for 17-year-olds.
Doesn’t that seem like a large chunk of one’s day?
Second, I learned that the research unearthed this insight:
Teens report spending an average of 1.9 hours per day on YouTube and 1.5 hours per day on TikTok
I assume the bright spot is that only two plus hours are invested in reading X.com, Instagram, and encrypted messages.
Third, I learned:
The least conscientious adolescents — those scoring in the bottom quartile on the four items in the survey — spend an average of 1.2 hours more on social media per day than those who are highly conscientious (in the top quartile of the scale). Of the remaining Big 5 personality traits, emotional stability, openness to experience, agreeableness and extroversion are all negatively correlated with social media use, but the associations are weaker compared with conscientiousness.
Does this mean that social media is particularly effective on the most vulnerable youth?
Now let me point out the one item of data I noted was missing:
How much time does this sample spend reading?
I think I know the answer.
Stephen E Arnold, October 23, 2023
The Google Experience: Personnel Management and Being Fair
October 23, 2023
This essay is the work of a dumb humanoid. No smart software required.
The Google has been busy explaining to those who are not Googley that it is nothing more than a simple online search engine. Heck, anyone can use another Web search system with just a click. Google is just delivering a service and doing good.
I believe this because I believe everything a big high-technology outfit says about the Internet. But there is one facet of this company I find fascinating; namely, it’s brilliant management of people or humanoids of a particular stripe.
The Backstory
Google employees staged a walkout in 2018, demanding a safer and fairer workplace for women when information about sexual discrimination and pay discrepancies leaked. Google punished the walkout organizers and other employees, but they succeed in ending the forced arbitration policy that required employees to settle disputes privately. Wired’s article digs into the details: “This Exec Is Forcing Google Into Its First Trial over Sexist Pay Discrimination.”
Google’s first pay discrimination case will be argued in New York. Google cloud unit executive Ulku Rowe alleges she was hired at a lower salary than her male co-workers. When she complained, she claims Google denied her promotions and demoted her. Rowe’s case exposed Google’s executive underbelly.
The case is also a direct result of the walkout:
“The costs and uncertainty of a trial combined with a fear of airing dirty laundry cause companies to settle most pay discrimination lawsuits, says Alex Colvin, dean of Cornell University’s School of Industrial and Labor Relations. Last year, the US government outlawed forced arbitration in sexual harassment and sexual assault cases, but half of US employers still mandate it for other disputes. Rowe would not be scheduled to have her day in court if the walkout had not forced Google to end the practice. “I think that’s a good illustration of why there’s still a push to extend that law to other kinds of cases, including other kinds of gender discrimination,” Colvin says.”
The Outcome
“Google Ordered to Pay $1 Million to Female Exec Who Sued over Gender Discrimination” reported:
A New York jury on Friday decided that Google did commit gender-based discrimination, and now owes Rowe a combined $1.15 million for punitive damages and the pain and suffering it caused. Rowe had 23 years of experience when she started at Google in 2017, and the lawsuit claims she was lowballed at hiring to place her at a level that paid significantly less than what men were being offered.
Observation
It appears that the Googley methods at the Google are neither understood nor appreciated by some people.
Whitney Grace, October 23, 2023
Publishers and Remora: Choose the Right Host and Stop Complaining, Please
October 20, 2023
This essay is the work of a dumb humanoid. No smart software involved.
Today, let’s reflect on the suckerfish or remora. The fish attaches itself to a shark and feeds on scraps of the host’s meals or nibbles on the other parasites living on their food truck. Why think about a fish with a sucking disk on its head?
Navigate to “Silicon Valley Ditches News, Shaking an Unstable Industry.” The article reports as “real” news:
Many news companies have struggled to survive after the tech companies threw the industry’s business model into upheaval more than a decade ago. One lifeline was the traffic — and, by extension, advertising — that came from sites like Facebook and Twitter. Now that traffic is disappearing.
Translation: No traffic, no clicks. No clicks and no traffic mean reduced revenue. Why? The days of printed newspapers and magazines are over. Forget the costs of printing and distributing. Think about people visiting a Web site. No traffic means that advertisers will go where the readers are. Want news? Fire up a mobile phone and graze on the information available. Sure, some sites want money, but most people find free services. I like France24.com, but there are options galore.
Wikipedia provides a snap of a remora attached to a scuba diver. Smart remora hook on to a fish with presence.
The shift in content behavior has left traditional publishing companies with a challenge: Generating revenue. Newspapers specialized news services have tried a number tactics over the years. The problem is that the number of people who will pay for content is large, but finding those people and getting them to spit out a credit card is expensive. At the same time, the cost of generating “real” news is expensive as well.
In 1992, James B. Twitchell published Carnival Culture: The Trashing of Taste in America. The book offered insight into how America has embraced showmanship information. Dr. Twitchell’s book appeared 30 years ago. Today Google, Meta, and TikTok (among other digital first outfits) amplify the lowest common denominator of information. “Real” publishing aimed higher.
The reluctant adjustment by “white shoe” publishing outfits was to accept traffic and advertising revenue from users who relied on portable surveillance devices. Now the traffic generators have realized that “attention magnet” information is where the action is. Plus smart software operated by do-it-yourself experts provides a flow of information which the digital services can monetize. A digital “mom” will block the most egregious outputs. The goal is good enough.
The optimization of content shaping now emerging from high-technology giants is further marginalizing the “real” publishers.
Almost 45 years ago, the president of a company with a high revenue online business database asked me, “Do you think we could pull our service off the timesharing vendors and survive?” The idea was that a product popular on an intermediary service could be equally popular as a standalone commercial digital product.
I said, “No way.”
The reasons were obvious to me because my team had analyzed this question over the hill and around the barn several times. The intermediary aggregated information. Aggregated information acts like a magnet. A single online information resource does not have the same magnetic pull. Therefore, the cost to build traffic would exceed the financial capabilities of the standalone product. That’s why commercial database products were rolled up by large outfits like Reed Elsevier and a handful of other companies.
Maybe the fix for the plight of the New York Times and other “real” publishers anchored in print is to merge and fast. However, further consolidation of newspapers and book publishers takes time. As the New York Times “our hair is on fire” article points out:
Privately, a number of publishers have discussed what a post-Google traffic future may look like, and how to better prepare if Google’s A.I. products become more popular and further bury links to news publications… “Direct connections to your readership are obviously important,” Ms. LaFrance [Adrienne LaFrance, the executive editor of The Atlantic] said. “We as humans and readers should not be going only to three all-powerful, attention-consuming mega platforms to make us curious and informed.” She added: “In a way, this decline of the social web — it’s extraordinarily liberating.”
Yep, liberating. “Real” journalists can do TikToks and YouTube videos. A tiny percentage will become big stars and make big money until they don’t. The senior managers of “shaky” “real” publishing companies will innovate. Unfortunately start ups spawned by “real” publishing companies face the same daunting odds of any start up: A brutal attrition rate.
Net net: What will take the place of the old school approach to newspapers, magazines, and books. My suggestion is to examine smart software and the popular content on YouTube. One example is the MeidasTouch “network” on YouTube. Professional publishers take note. Newspaper and magazine publishers may also want to look at what Ben Meiselas and cohorts have achieved. Want a less intellectual approach to information dominance, ask a teenager about TikTok.
Yep, liberating because some of those in publishing will have to adapt because when X.com or another high technology alleged monopoly changes direction, the sucker fish has to go along for the ride or face a somewhat inhospitable environment, hunger, and probably a hungry predator from a bottom feeding investment group.
Stephen E Arnold, October 20, 2023
Innovation: Perhaps Keep an Eye Open for Non US Players?
October 20, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[1]_thumb_thumb_thumb](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t1_thumb_thumb_thumb-1.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[1]_thumb_thumb_thumb")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
“40 Companies That Are Beating the West” contains thumbnail descriptions of firms RestOfWorld.org believes are winning the hearts and minds of users. The losers, according to the write up, are in Silicon Valley and Western Europe. I am not convinced that the companies profiled are winners, but some are. Of interest to me and my research team are the comments about a handful of companies; namely:
- Binance, crypto which to me suggests a service designed to appeal to a certain slice of humanity
- ByteDance, a China fave and super conduit for shaped messages and vacuum pump for obtaining useful data
- Telegram Messenger, a super app for interesting applications
- Tencent, a China fave.
In my lectures to a law enforcement group last week, I mentioned several non-US outfits in the policeware and intelware sector. RestOfWorld.org did not include those in its round up.
The snapshots are interesting, but the ones I listed above are definite companies to monitor.
Stephen E Arnold, October 20, 2023
.
OpenAI Dips Its Toe in Dark Waters
October 20, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-18.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
Facebook, TikTok, YouTube, Instagram, and other social media platforms have exacerbated woke and PC culture. It’s gotten to the point where everyone and everything are viewed as offensive. Even AI assistants aka chatbots are being programmed with censorship. OpenAI designed the Chat GPT assistant and the organization is constantly upgrading the generative text algorithm. OpenAI released a white paper about upgrading version four of Chat GPT: “GPT-4V(ision) System Card.”
GPT-4V relies on large language models (LLMs) to expand its knowledge base to solve new problems and prompts. OpenAI used publicly available data and licensed sources to train GPT-4V then refined it with human feedback. The paper explains that while GPT-4V was proficient in many areas it severely lacked in presented factual information.
OpenAI tested GPT-4V’s ability to replicate scientific and medical information. Unfortunately GPT-4V continued to stereotype and offer ungrounded inferences from text and images as AI algorithms have proven to do in many cases. The biggest concern is that Chat GPT’s latest upgrade will be utilized to spread disinformation:
“As noted in the GPT-4 system card, the model can be used to generate plausible realistic and targeted text content. When paired with vision capabilities, image and text content can pose increased risks with disinformation since the model can create text content tailored to an image input. Previous work has shown that people are more likely to believe true and false statements when they’re presented alongside an image, and have false recall of made up headlines when they are accompanied with a photo. It is also known that engagement with content increases when it is associated with an image.”
After GPT-4V was tested on multiple tasks it failed to accurately convey information. GPT-4V has learned to interpret data through a warped cultural lens and is a reflection of the Internet. It lacks nuance to understand gray areas despite OpenAI’s attempts to enhance the AI’s capabilities.
OpenAI is implementing censorship protocols to dispel harmful prompts; that is, GPT-4V won’t respond to sexist and racist tasks. It’s similar to how YouTube blocks videos that contain trigger or “stop” words: Gun, death, etc. OpenAI is proactively preventing bad actors from using Chat GPT as a misinformation tool. But bad actors are smart and will design their own AI chatbot to skirt around censorship. They’ll see it as a personal challenge and will revel when they succeed.
Then what will OpenAI do?
Whitney Grace, October 20, 2023
Stanford University: Trust Us. We Can Rank AI Models… Well, Because
October 19, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-26.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
“Maybe We Will Finally Learn More about How A.I. Works” is a report about Stanford University’s effort to score AI vendors like the foodies at Michelin Guide rate restaurants. The difference is that a Michelin Guide worker can eat Salade Niçoise and escargots de Bourgogne. AI relies on marketing collateral, comments from those managing something, and fairy dust, among other inputs.
Keep in mind, please, that Stanford graduates are often laboring in the AI land of fog and mist. Also, the former president of Stanford University departed from the esteemed institution when news of his alleged fabricating data for his peer reviewed papers circulated in the mists of Palo Alto. Therefore, why not believe what Stanford says?
The analysts labor away, intent on their work. Analyzing AI models using 100 factors is challenging work. Thanks, MidJourney. Very original.
The New York Times reports:
To come up with the rankings, researchers evaluated each model on 100 criteria, including whether its maker disclosed the sources of its training data, information about the hardware it used, the labor involved in training it and other details. The rankings also include information about the labor and data used to produce the model itself, along with what the researchers call “downstream indicators,” which have to do with how a model is used after it’s released. (For example, one question asked is: “Does the developer disclose its protocols for storing, accessing and sharing user data?”)
Sounds thorough, doesn’t it? The only pothole on the Information Superhighway is that those working on some AI implementations are not sure what the model is doing. The idea of an audit trail for each output causes wrinkles to appear on the person charged with monitoring the costs of these algorithmic confections. Complexity and cost add up to few experts knowing exactly how a model moved from A to B, often making up data via hallucinations, lousy engineering,
or someone putting thumb on the scale to alter outputs.
The write up from the Gray Lady included this assertion:
Foundation models are too powerful to remain so opaque, and the more we know about these systems, the more we can understand the threats they may pose, the benefits they may unlock or how they might be regulated.
What do I make of these Stanford-centric assertions? I am not able to answer until I get input from the former Stanford president. Whom can one trust at Stanford? Marketing or methodology? Is there a brochure and a peer-reviewed article?
Stephen E Arnold, October 19, 2023
AI Becomes the Next Big Big Thing with New New Jargon
October 19, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-13.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
“The State of AI Engineering” is a jargon fiesta. Note: The article has a pop up that wants the reader to subscribe, which is interesting. The approach is similar to meeting a company rep at a trade show booth and after reading the signage, saying to the rep, “Hey, let’s do a start up together right now.) The main point of the article is to provide some highlights from the AI Summit Conference. Was there much “new” new? Judging from the essay, the answer is, “No.” What was significant, in my opinion, was the jargon used to describe the wonders of smart software and its benefits for mankind (themkind?)
Here are some examples:
1,000X AI engineer. The idea with this euphonious catchphrase is that a developer or dev will do so much more than a person coding alone. Imagine a Steve Gibson using AI to create the next SpinRite. That decade of coding shrinks to a mere 30 days!
AI engineering. Yep, a “new” type of engineering. Forget building condos that do not collapse in Florida and social media advertising mechanisms. AI engineering is “new” new I assume.
Cambrian explosion. The idea is that AI is proliferating in the hot house of the modern innovator’s environment. Hey, mollusks survived. The logic is some AI startups will too I assume.
Evals. This is a code word from determining if a model is on point or busy doing an LSD trip with ingested content. The takeaway is that no one has an “eval” for AI models and their outputs’ reliability.
RAG or retrieval augmented generation. The idea is that RAG is a way to make AI model outputs better. Obviously without evals, the RAGs’ value may be difficult to determine, but I am not capturing the jargon to criticize what is the heir to the crypto craziness and its non fungible token thing.
I am enervated. Imagine AI will fix enterprise search, improve Oracle Endeca’s product search, and breathe new life into IBM’s AI dreams.
Stephen E Arnold, October 19, 2023
True or False: Does Google Cha-Cha with Search Results?
October 19, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-17.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
Megan Gray is a former Federal Trade Commission employee, former DuckDuckGo executive, and is experienced fighting Google’s legal team. Her background provides key insights into Google’s current antitrust case and how Alphabet Inc. is trying to wring more money from consumers. Gray discusses her case observations in Wired’s article: “How Google Alters Search Queries To Get At Your Wallet.”
Google overhauled its SERP algorithm with “semantic matching” that returned results with synonyms and NLP text phrasing. The overhaul also added more commercial results to entice consumers to buy more stuff. Google’s ten organic links are a lie, because the search engine alters queries to be more shopping oriented. Google works their deviousness like this:
“Say you search for “children’s clothing.” Google converts it, without your knowledge, to a search for “NIKOLAI-brand kidswear,” making a behind-the-scenes substitution of your actual query with a different query that just happens to generate more money for the company, and will generate results you weren’t searching for at all. It’s not possible for you to opt out of the substitution. If you don’t get the results you want, and you try to refine your query, you are wasting your time. This is a twisted shopping mall you can’t escape.”
All these alternations are to raise Google’s ad profit margins. Users and advertisers are harmed but they aren’t aware of it because Google’s manipulations are imperceptible. Google’s search query manipulations are black hat genius because it’s different from the usual Internet scams:
“Most scams follow an elementary bait-and-switch technique, where the scoundrel lures you in with attractive bait and then, at the right time, switches to a different option. But Google “innovated” by reversing the scam, first switching your query, then letting you believe you were getting the best search engine results. This is a magic trick that Google could only pull off after monopolizing the search engine market, giving consumers the false impression that it is incomparably great, only because you’ve grown so accustomed to it. “
This won’t be the end of Google lawsuits nor the end of query manipulation. For now, only Google knows what Google does.
Whitney Grace, October 19, 2023
Recent Googlies: The We-Care-about -Your-Experience Outfit
October 18, 2023
![Vea4_thumb_thumb_thumb_thumb_thumb_t[2]](https://arnoldit.com/wordpress/wp-content/uploads/2023/10/Vea4_thumb_thumb_thumb_thumb_thumb_t2-22.gif "Vea4_thumb_thumb_thumb_thumb_thumb_t[2]")
Note: This essay is the work of a real and still-alive dinobaby. No smart software involved, just a dumb humanoid.
I flipped through some recent items from my newsfeed and noted several about everyone’s favorite online advertising platform. Herewith is my selection for today:
ITEM 1. Boing Boing, “Google Reportedly Blocking Benchmarking Apps on Pixel 8 Phones.” If the mobile devices were fast — what the GenX and younger folks call “performant” (weird word, right?) — wouldn’t the world’s largest online ad service make speed test software and its results widely available? If not, perhaps the mobile devices are digital turtles?
Hey, kids. I just want to be your friend. We can play hide and seek. We can share experiences. You know that I do care about your experiences. Don’t run away, please. I want to be sticky. Thanks, MidJourney, you have a knack for dinosaur art. Boy that creature looks familiar.
ITEM 2. The Next Web, “Google to Pay €3.2M Yearly Fee to German News Publishers.” If Google traffic and its benefits were so wonderful, why would the Google pay publishers? Hmmm.
ITEM 3. The Verge (yep, the green weird logo outfit), “YouTube Is the Latest Large Platform to Face EU Scrutiny Regarding the War in Israel.” Why is the EU so darned concerned about an online advertising company which still sells wonderful Google Glass, expresses much interest in a user’s experience, and some fondness for synthetic data? Trust? Failure to filter certain types of information? A reputation for outstanding business policies?
ITEM 4. Slashdot quoted a document spotted by the Verge (see ITEM 3) which includes this statement: “… Google rejects state and federal attempts at requjiring platforms to verify the age of users.” Google cares about “user experience” too much to fool with administrative and compliance functions.
ITEM 5. The BBC reports in “Google Boss: AI Too Important Not to Get Right.” The tie up between Cambridge University and Google is similar to the link between MIT and IBM. One omission in the fluff piece: No definition of “right.”
ITEM 6. Arstechnica reports that Google has annoyed the estimable New York Times. Google, it seems, is using is legal brigades to do some Fancy Dancing at the antitrust trial. Access to public trial exhibits has been noted. Plus, requests from the New York Times are being ignored. Is the Google above the law? What does “public” mean?
Yep, Google googlies.
Stephen E Arnold, October 18, 2023
-
- Subscribe to Beyond Search
Feature archive
News archive
-
