Amazon Zoom: This Is Just Not Important, Right?
December 2, 2020
“Zoom and Amazon Extend Partnership for the Future of Online Communication” explains that “Zoom is coming to an Amazon Echo Show.” Makes sense, right? The article reports:
The new collaboration extends the companies’ existing relationship, which will now see AWS working to expand and scale Zoom’s services to meet customer demand.
What does the write up omit? A couple of things; for instance:
- What type of data will Amazon AWS log files capture?
- How quickly will AWS “zoomification” diffuse to other video applications?
- What Zoom data will be subject to cross correlation in Amazon’s streaming data marketplace?
No answers in the write up or from either Amazon or Zoom yet. This announcement is not important, right?
Stephen E Arnold, December 2, 2020
Artificial Intelligence: A Semi Cool Loop?
December 1, 2020
What metaphorical disaster awaits when artificial intelligence diffuses? “An AI Winter May Be Inevitable. What We Should Fear More: An AI Ice Age” suggests that no one is paying attention to the increased use of smart software. I did not think of artificial intelligence in terms of the winter or an ice age. However, Here’s the passage I noted:
We’re only seeing the most progress in learning because that’s where all the investment is going. For example: we’ve only been focused on learning the last 15 years. But AI done properly needs to cover a range of intelligence capabilities, of which being able to learn and spot patterns is just one; there’s reasoning, there’s understanding, there’s a lot of other types of intelligence capabilities that should be part of an overall Artificial Intelligence practice or capability. We know why that is—we’re focused on learning because we got good traction with that and made solid progress. But there’s all the other AI capabilities that we should be also looking at and investigating, and we’re just not. It’s a Catch-22: all the smart money is going into Machine Learning because that’s where we’re seeing the most progress, but we’re only seeing the most progress in learning because that’s where all the investment is going!
The circularity and feedback loop is a characteristic of some technology and digital operations. Perhaps the datasphere allows iterative processes to manifest themselves and then amplify their effects? I am not sure this hypothetical behavior fits into the winter and ice age metaphor.
Narrowing one’s focus may accelerate the fragmentation of an already fuzzy concept like smart software.
Stephen E Arnold, December 1, 2020
Fess Up: Elasticsearch Is a Threat to Proprietary Search and Retrieval
December 1, 2020
We have been poking around the world of Elasticsearch-based information retrieval systems. There are some interesting plays; that is, entrepreneurs use Elasticsearch (Shay Banon’s open source system) as a platform.
Fess provides Elasticsearch for personal use, although one can employ the system for an organization. The system is:
Fess is Elasticsearch-based search server, but knowledge/experience about Elasticsearch is NOT needed because of All-in-One Enterprise Search Server. Fess provides Administration GUI to configure the system on your browser. Fess also contains a crawler, which can crawl documents on Web/File System/DB and support many file formats, such as MS Office, pdf and zip.
Fess became available in 2019. The CEO of the N2SM, Inc. company is Masaharu Manabe. Demonstrations and links to the code are available at this link. A fee-based version of the software is provided under the name N2 Search. More information about the for fee version is here. A discussion forum is available at this link.
Observation: The Elasticsearch ecosystem is providing alternatives to the proprietary search systems. Beyond Search thinks that some vendors of proprietary search software are likely to be see Elasticsearch as digital kudzu. Good news or bad news for the Coveos, Fabasofts, and Microsoft Fast type folks? That’s a question some of these types of vendors stakeholders may be asking as they beat the bushes for deals in customer service, chatbots, business intelligence, and smart software services.
Stephen E Arnold, December 1, 2020
Shocker: Online Learning Teaches Little
December 1, 2020
I may be misunderstanding “Failing Grades Spike in Virginia’s Largest School System as Online Learning Gap Emerges Nationwide,” but I think the main idea is that online learning does not teach the way students-teachers in an old-fashioned class do. You will have to pay to read this most recent report from a Captain Obvious “real news” outfit.
Back to the “news” flash.
The write up states:
But one Fairfax high school teacher, who spoke on the condition of anonymity for fear of retribution from the school system, said he is doing all of these things — and still, 50 to 70 percent of his 150 students are achieving D’s and F’s, whereas before they had earned B’s and C’s.
There you go. We’re teaching students something, just not what the school hopes will be learned. What subject do students learn? Inattention perhaps.
Another factoid. Sit down and take a deep meditative breath before reading:
Younger Fairfax students are struggling more than older ones: The percentage of middle-schoolers receiving at least two F’s quadrupled, while the percentage of high-schoolers scoring at least two F’s increased by 50 percent. The percentage of students with disabilities earning at least two F’s, meanwhile, more than doubled, while the percentage of children for whom English is a second language receiving at least two F’s rose by 106 percent to account for 35 percent of all children in this group. Among racial groups, Hispanic students were most affected: The percentage of these students with at least two F’s jumped from 13 to 25 percent. Comparing grades achieved in past years with grades this year showed that the drop in passing grades is significant and unprecedented.
Had enough? I haven’t. Several observations:
- Traditional educational methods evolved toward a human “teacher” presenting information.
- Students were monitored and tested.
- Peer pressure operated in a social setting like an old-fashioned school room.
- Peer mediated instruction took place in non-class settings; for example, at a lunch table or talking with a friend at a school locker.
- Old-fashioned family structures often reinforced “learning.” Example: Consequences if lessons were not completed.
Thumb typers now have to face up to a reality in which their expertise at inattention creates a false sense of knowledge.
The problem is that moving learning to Zoom or some other online platform has a shallow experiential pool. Traditional education benefits from a long history. Maybe online will catch up, but if the students are ill prepared, inattentive, and unable to draw upon a knowledge framework — not likely.
Anyone ready for the new Dark Ages? Whoops. News flash. We are in them. Plague, social unrest, and students who are not acquiring equipment for reading.
Hey, everyone has a smartphone. What could go wrong? TikTok and YouTube autosuggest are just super.
Stephen E Arnold, December 1, 2020
DarkCyber for December 1, 2020, Now Available
December 1, 2020
DarkCyber reports about Maltrail, an open source cyber tool for detecting malicious traffic. Crime as a Service matures. Now anyone can point-and-click through a ransomware attack. Bad actors helpfully make cyber crime less of a hassle. Insider threats — what DarkCyber calls “the Snowden play” — are becoming more prevalent. Why? A need for money, revenge, or a dose of that old Silicon Valley attitude.
The feature in this episode is a summary of the next-generation in entity recognition from videos and still images. Face recognition is not the most reliable technology in the world; however, researchers from China and Japan have figured out how to match a person’s gait to an individual. Ergo gait recognition. A link to the technical details appears in the program.
The program features a brief extract from a conversation between Robert David Steele, a former CIA professional, and Stephen E Arnold (owner of Dark Cyber). Arnold describes some of the less appreciated reasons why digital information creates new challenges for law enforcement and intelligence professionals. Good news? Not really.
The final story in the program addresses the urgent need for counter unmanned aerial systems by local, county, and statement law enforcement agencies. Individuals are ramming drones into police helicopters. The DarkCyber discussion of this problem includes a link to a series of recommendations promulgated by the British government to address this kinetic use of drones.
DarkCyber is produced by Beyond Search. The video program appears every two weeks. The third season of DarkCyber begins in January 2021. The program is non-commercial, does not accept advertising, and does not beg for dollars. How is this possible? DarkCyber is not sure.
You can view the program at this link.
Kenny Toth, December 1, 2020
-
- Subscribe to Beyond Search
Feature archive
News archive
-
