Smart Software: Just Keep Adding Layers between the Data and the Coder
September 6, 2018
What could be easier? Clicking or coding.
Give up. Clicking wins.
A purist might suggest that training smart software requires an individual with math and data analysis skills. A modern hippy dippy approach is to suggest that pointing and clicking is the way of the future.
Amazon is embracing that approach and other firms are too.
I read “Baidu Launches EZDL, an AI Model Training Platform That Requires No Coding Experience.” Even in China where technical talent is slightly more abundant than in Harrod’s Creek, Kentucky, is on the bandwagon.
I learned:
Baidu this week launched an online tool in beta — EZDL — that makes it easy for virtually anyone to build, design, and deploy artificial intelligence (AI) models without writing a single line of code.
Why slog through courses? Point and click. The future.
There’s not much detail in the write up, but I get the general idea of what’s up from this passage from the write up:
To train a model, EZDL requires 20-100 images, or more than 50 audio files, assigned to each label, and training takes between 15 minutes and an hour. (Baidu claims that more than two-thirds of models get accuracy scores higher than 90 percent.) Generated algorithms can be deployed in the cloud and accessed via an API, or downloaded in the form of a software development kit that supports iOS, Android, and other operating systems.
Oh, oh. The “API” buzzword is in the paragraph, so life is not completely code free.
Baidu, like Amazon, has a bit of the competitive spirit. The write up explains:
Baidu’s made its AI ambitions clear in the two years since it launched Baidu Brain, its eponymous platform for enterprise AI. The company says more than 600,000 developers are currently using Brain 3.0 — the newest version, released in July 2018 — for 110 AI services across 20 industries.
What could go wrong? Nothing, I assume. Absolutely nothing.
Stephen E Arnold, September 6, 2018
From the Gloom, a Ray of Management Light
September 5, 2018
Forget the contentious Congressional hearings. Ignore the excitement of technology companies explaining. Forget the throttling of bandwidth.
Navigate to this management gem: “The #1 Office Perk? Natural Light.”
The Harvard Business Review, one of the fonts of management wisdom, explains:
… A new survey by my HR advisory firm Future Workplace called “The Employee Experience” reveals the reality is that employees crave something far more fundamental and essential to human needs. In a research poll of 1,614 North American employees, we found that access to natural light and views of the outdoors are the number one attribute of the workplace environment, outranking stalwarts like onsite cafeterias, fitness centers, and premium perks including on-site childcare…
Now I am not sure about the composition of the sample. For decades, I have worked in offices with zero natural light. If I had windows in my sometimes overly large offices, I blocked the natural sunlight. I put cardboard in the overhead fixtures. In short, I was happy as a clam in semi darkness. My monitors provided the light and warmth I needed.
But you, gentle reader, may be different. So take heed. Forget high school science club management methods and create natural light for your employees. If the building or loft or renovated warehouse lacks windows, do what the Romans did. Chop a hole in the room.
Don’t forget the impluvium, however. A happy company should not have workers in wet shoes.
Stephen E Arnold, September 5, 2018
India and Social Media: A Bit of an Issue
September 5, 2018
Spam is a pain for your inbox, feed, social network messages, and pretty much anything else you do online. One of the worse things about spam messages is when someone does not know how to identify spam from the real stuff. According to Reuters, the Indian government is getting fed up with spam, says the article, “WhatsApp To Clamp Down On ‘Sinister’ Messages In India: Ravi Shankar Prasad.”
Facebook apparently said it would develop tools to help the Indian government detect spam and other content with the purpose of sparking mass hysteria. India is not any different from other countries when it is whipped into a frenzy: people get angry, there is collateral damage, and people get hurt. WhatsApp CEO Chris Daniels commiserated with India’s chief information technology minister Ravi Shankar Prasad. Prasad wants Facebook to design a way to track rabble rousing messages’ origins. The IT minister does not think it is rocket science to figure a message’s origins, seemingly not knowing what work is required in order to read the metadata and program the code.
WhatsApp’s biggest market is India with a 200 million strong market and where, quite astonishingly, people forward more content than any other country. We learned:
“There are also concerns that supporters of political parties could use social media platforms such as WhatsApp to spread false messages in the run-up to India’s national elections in 2019.In July, WhatsApp said message forwards will be limited to five chats at a time, whether among individuals or groups, and said it will remove the quick forward button placed next to media messages.”
India is also a very social country, meaning your reputation, education, connections, and family status can mean the difference between success and failure. Social networks are more complex than anything you could find in the West, China might be the only country that compares. It does not come as a surprise WhatsApp could be used as a political tool and also to incite violence.
Whitney Grace, September 5, 2018
Factualities for September 5, 2018
September 5, 2018
Believe ‘em or not.
- $13,470. Cost for 30 minutes of T Mobile Internet access. Source: Hardocp News
- 2.2 million customers’ data lost in data breach by T Mobile. Source: Naked Security
- $60 billion. The size of the global blockchain market in 2024. Source: Blokt
- 20%. Number of employees who share their email password with co workers. Source: ZDNet
- 2 percent. The number of teens who read a newspaper. Source: Science
- 64 billion. Number of fake emails sent each day. Source: Dark Reading
Stephen E Arnold, September 5, 2018
High School Science Club Management: Volunteer Workers Point Out Flaws
September 4, 2018
I know that HSSCM methods is a fresh concept. I used the phrase in a few lectures years ago. I am delighted that “high school science club management methods” or HSSCM continues to be a useful concept.
One recent example is the information (which I assume to be spot on) from an online information service and a YouTube centric outfit. The source is “Unpaid and Abused: Moderators Speak Out Against Reddit.”
The point of the write up is that Reddit, itself an online social information service, reports:
Engadget and Point spoke to 10 Reddit moderators, and all of them complained that Reddit is systematically failing to tackle the abuse they suffer. Keeping the front page of the internet clean has become a thankless and abusive task, and yet Reddit’s administration has repeatedly neglected to respond to moderators who report offenses.
Reddit is owned by Advance Publications. Therefore, the write up is focusing on the management methods of a large corporation.
If the information in the write up is accurate, the large corporation has allowed HSSCM methods to help keep Reddit on track.
One example:
Subreddits can have their own bylaws, but racism, sexism and hate speech are targeted by moderators on pretty much any thread. The clear majority of abuse is in response to moderators calling users out when they break the rules.
The result? Death threats, burn out, and what seem to be a lack of management systems which can:
- Work constructively with volunteers who express concerns about their perceptions about death threats, stress, etc.
- Present a positive image of the management for the organization unit
- Make clear that innovative management methods are indeed having a positive effect on volunteers, employees, users, and stakeholders.
My view is that HSSCM methods continue to be the path of least resistance.
The problem is that those individuals who perceive the ineffectiveness of those HSSCM methods are beginning to speak up. Employee push back at a number of companies make it clear that traditional management methods are not a viable option for the owners of these outfits and that the HSSCM methods used because they are easy also lack suitable benefits.
In short, a crisis for MBA truisms and for the HSSCM approaches has arrived.
Of course, if the information in the Engadget article is not representative or not up to the accuracy standards I assume to be in play, MBA and HSSCM methods are working just fine.
Stephen E Arnold, September 4, 2018
IBM and Its Employee Assessment Idea
September 4, 2018
If you are a programmer at IBM, you might be sitting on a powderkeg of ecpnomic and credibility challenges. No, it has nothing to do with national security or privacy. It has to do with your performance review. We learned about his strange AI wrinkle from a recent Computing story, “IBM is Automating Employee Performance Analysis with Watson AI.”
According to the story:
“The AI also takes information from IBM’s databases to find out if those workers have boosted their skills. It then serves a performance rating up to managers who can look at the score and decide what pay, bonuses and promotions the employees might be up for.”
Can you imagine writing the code for the AI that judges whether or not you get an annual raise? We suspect that this plan has a lot of potential for excitement. It would not be unlike if the criminals were allowed to write the programs that handle AI sentencing in the courts. Thankfully, that has not happened, but if we were advising IBM, the Beyond Search team might want to keep humans in the loop.
Watson could not cure cancer. Therefore, Watson can do compensation? Even in Harrod’s Creek, the two statements are difficult to reconcile.
Patrick Roland, September 4, 2018
Blockchain Bridges the Crypto Gap of Legality
September 4, 2018
Cryptocurrency like Bitcoin has long been the Dark Web’s favorite way of doing illegal business transactions. However, the technology that it is built upon is opening up and providing law enforcement with an interesting weapon, as we discovered in a recent CoinPick story, “Use of BitCoin Over Dark Web Has Dropped, but DEA Wants Criminals to Keep Using Cryptocurrencies.”
According to the piece:
“This is where blockchain plays a very important part. Even though Bitcoin does not carry IDs, the transactions being available on a distributive ledger are accessible to the public. The investigators can track the funds and apprehend Individuals related to criminal activity this way. Infante further stated: The blockchain actually gives us a lot of tools to be able to identify people.”
By exploiting the platform’s weaknesses, law enforcement is zeroing in on illegal activity. We expect this to become more and more common thanks to the fact that police are beginning to familiarize themselves with it for run-of-the-mill internal programs as well as high level crimefighting. This gap is being bridged and one of two things will happen: Either cryptocurrency crime will be wiped out, or the bad guys will have to find a new way to stay hidden.
In Stephen E Arnold’s upcoming lecture in Washington, DC, attendees will learn that bad actors need to be aware of a new intelligence service. The provider? Amazon. More details will be shared in a DarkCyber video after Stephen returns.
Patrick Roland, September 4, 2018
DarkCyber for September 4, 2018, Now Available
September 4, 2018
DarkCyber for September 4, 2018, is now available at www.arnoldit.com/wordpress and on Vimeo at https://vimeo.com/287783314.
Stephen E Arnold’s DarkCyber is a weekly video news and analysis program about the Dark Web and lesser known Internet services.
This week’s program covers three Dark Web and security related stories.
The first story addresses Gephi, an open graph visualization platform. Unlike Excel, Gephi is a platform. The software system can be a useful complement to blockchain deanonymization tasks. An analyst can perform link analysis; that is, what hyperlink or item leads to another. The Gephi One feature allows the user to turn a graphical representation and explore it in three dimensions. A user can interact with the data, drilling down into a cluster or popping up one or more levels to see how a particular item relates to a broader grouping of data. The system can manipulate up to one million nodes and edges. Some commercial tools struggle to deal with more than a handful of nodes and edges. The video includes a link at which Gephi can be downloaded.
The second story describes a vehicle tracking and surveillance innovation called Zoomed. Developed by Cameroonia computer whiz Zuo Bruno, the system does not require the Internet. Instead, Zuo Bruno devised a system which operates via SMS. Once the device is placed in a vehicle, the location of the vehicle can be determined by placing a mobile call to the Zoomed device. The device drops the call and messages the location and other data of the vehicle. The Zoomed technology can perform other functions as well; for example, the audio in the vehicle can be recorded and the vehicle can be disabled.
The third story describes a free account takeover alerting service or ATO from Truthfinder. The idea is that after a person registers for the service, Truthfinder will notify that individual when his personal information is discovered by the monitoring service. DarkCyber explains how to sign up for the service and how to disable the notifications if they become a burden.
Kenny Toth, September 4, 2018
Gray Literature May Face a Backlash from Professional Publishing Companies
September 3, 2018
I read “Read Research Papers Trapped Behind a Paywall With This Chrome Extension.” The write up describes a Google Chrome extension which helps a user locate a pre-publication version of a journal article. This makes veering around a paywall a little easier for some people. Here in Harrod’s Creek, we just ignore articles for which someone has assigned a price tag. We believe that ignorance is bliss, and we wouldn’t want to have our halcyon life disrupted.
The write up explains:
the extension searches for the article to see if the author has posted a free version anywhere as well.
Bingo.
The write up explains in a somewhat labored manner:
Important to note: This isn’t getting you around the paywall or stealing the content. Instead, what it’s doing is seeing if there’s another legal version online that you can read instead. Emphasis on the “legal” in that sentence. If there is another copy online, you’ll see a green tab on your screen letting you know.
The visual clue is a green tag. Click it and, in theory, you will see the so called pre-publication, gray version of the write up.
Beyond Search believes that some of the professional publishers will note this innovation. We hypothesize that some of these savvy executives will come up with some ideas to nuke these unauthorized versions of the publishers’ for fee articles.
Some professional publishers charge authors for the work required to take their drafts and convert them into professional publishing recurring revenues.
Authors, some of whom are academics, often have zero choice about publishing in journals owned by professional publishing outfits. Do gooder researchers often have different ideas.
Professional publishers can make life difficult because most have outstanding legal teams, a keen desire to keep the recurring revenue flowing, and a need to make sure another Google-type existential threat does not blindside them. To be fair, some professional publishers were not able to perceive how disruptive, Google-type outfits could affect their businesses until they were in the hospital after being hit by the digital trains.
If you want the Chrome extension, navigate to the source article. We don’t use Chrome, and, as I said, we don’t use content in peer reviewed journals. Why?
Sorry, like some of the work I did in my 50 year working career, I won’t talk. Move forward I say. But here’s a question for you, “How much are those must have medical and engineering journals either in dead tree or digital form?” That’s a fact worth knowing.
Stephen E Arnold, September 3, 2018
EU Not Taking Terror Lying Down
September 3, 2018
Applause are due to the European Union, who have instituted a unique plan for battling online extremism. We learned more about this tactic and were able to think about its potential stress points from reading an interesting Inquirer article, “EU Wants Tech Firms to Remove Terror Content Within an Hour of Facetimes.”
According to the story:
“In March, legislation passed that presented the hour-long purge window, but it was only marked under voluntary guidelines. But it looks like tech firms aren’t volunteering to work to these guidelines so the EU wants to force them to do so by imposing fines on firms that are blasé about terrorism-related materials on their services, networks and platforms.”
This is a very exciting prospect in the war on disinformation and violence planned online. However, it’s not time to celebrate quite yet. As reported recently, the EU is a complex machine itself and several countries failed to step in line with the cybersecurity deadline that loomed this summer. Beyond that, we are very curious to see how the various nations and governing bodies work together to oversee this new security measure. We hope it makes some headway, but will be watching from afar.
Patrick Roland, September 3, 2018
-
- Subscribe to Beyond Search
Feature archive
News archive
-
