Weakly Watson: A Smart Hat
June 14, 2016
I am not making this up. Navigate to “A Dad Made a Real-Life ‘Harry Potter’ Sorting Hat Using IBM’s Watson — Here’s How It Works.” Lucky kids. I bet everyone in their school understands the importance of IBM Watson. And the fashion. Stellar. I would reproduce a photo of the IBM Watson hat, but I am fearful that [a] a legal eagle would swoop upon me and [b] my gentle readers would think I photoshopped the image.
According to the write up:
The hat works simply enough. You place it on your head (that part is actually for fun, you could just talk to it) and tell the sorting hat a few things about yourself so it can sort you appropriately.
And Watson:
But running on the Natural Language Classifyer [sic] language alone would have forced users to describe themselves over written text. And we all know that’s not how the real sorting hat works! So Anderson also used Watson’s Speech to Text feature so you can speak to the hat.
You will find more details at this link. I want to reiterate that I am quite impressed. The lucky children will be the talk of anyone who sees them wearing this gizmo.
IBM Watson. More than recipes.
Stephen E Arnold, June 14, 2016
Alphabet Google: Quantum Computing in View
June 14, 2016
I read “Google Hartmut Neven Predicts That within 10 Years There Will Only Be Quantum Machine Learning and No Machine Learning on Classical Computers.” That works out to 3,600 days, give or take a few days. Years ago I included the Big O issue in my lectures about the algorithms used by most search and content processing vendors. Without wandering through that field again, traditional computing devices are not too helpful when asked to perform certain types of calculations.
The Googler in the “will be only” write up is predicting that Von Neumann’s goodies will be replaced with quantum gear. Keep in mind the “will be only.” That confidence and the categorical affirmative strike me as a bit on the wild side. I agree that new methods will arrive, but in 3,600 days?
Here’s what the write up tells me:
Google says it can scale up to that point relatively quickly, and other researchers in the field say it’s credible. It would likely take scaling up a little further to do useful work with an analog quantum computer. If and when Google or some other company does that, the devices could be used to crack tough chemistry problems in health or energy by simulating atoms to a level of realism impossible today.
I think the Alphabet Google thing is pretty good at technology. The company apparently has solved death, worked out how to put a computing device in one’s eye, and modified the Overture/GoTo advertising system to generate oodles of money.
But the “will be only” thing is a fresh new approach. Buy Alphabet Google stock now.
Stephen E Arnold, June 14, 2016
Ransomware as a Service Deals in Bitcoins of Course
June 14, 2016
Countless “as-a-service” models exist online. A piece from SCMagazine, Dark web forums found offering Cerber ‘ransomware as a service’, reveals more information about one such service called ransomware-as-a-service (RaaS), which we’ve heard about now for quite some time. Ransomware injects a virus onto a machine that encrypts the user’s files where they remain inaccessible until the victim pays for a key. Apparently, an Eastern European ransomware, Cerber, has been offering RaaS on Russian Dark Web forums. According to a cyber intelligence firm Sensecy, this ransomware was setup to include “blacklisted” countries so the malware does not execute on computers in certain locations. The article shares,
“Malwarebytes Labs senior security researcher Jerome Segura said the blacklisted geographies – most of which are Eastern European countries – provide “an indication of where the malware originated.” However, he said Malwarebytes Labs has not seen an indication that the ransomware is connected to the famed APT28 group, which is widely believed to be tied to the Russian government. The recent attacks demonstrate a proliferation of ransomware attacks targeting institutions in the U.S. and Western nations, as recent reports have warned. Last week, the Institute for Critical Infrastructure Technology (ICIT) released a study that predicted previously exploited vulnerabilities will soon be utilized to extract ransom.”
Another interesting bit of information to note from this piece is the going ransom is one bitcoin. Segura mentions the value ransomers ask for may be changing as he has seen some cases where the ransomer works to identify whether the user may be able to pay more. Regardless of the location of a RaaS provider, these technological feats are nothing new. The interesting piece is the supposedly untraceable ransom medium supplanting cash.
Megan Feil, June 14, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
SLI Systems Hopeful as Losses Narrow and Revenue Grows
June 14, 2016
The article titled SLI Systems Narrows First-Half Loss on Scoop reports revenue growth and plans to mitigate losses. SLI Systems is a New Zealand-based software as a service (SaaS) business that provides cloud-based search resources to online retailers. Founded in 2001, SLI Systems has already weathered a great deal of storms in the form of the dot-com crash that threatened to stall the core technology (developed at GlobalBrain.) According to a statement from the company, last year’s loss of $502K was an improvement from the loss of $4.1M in 2014. The article states,
“SLI shares have dropped 18 percent in the past 12 months, to trade recently at 76 cents, about half the level of the 2013 initial public offering price of $1.50. The software developer missed its sales forecast for the second half of the 2015 year but is optimistic new chief executive Chris Brennan and Martin Onofrio as chief revenue officer, both Silicon Valley veterans, can drive growth in revenue and earnings.”
The SLI of SLI stands for Search, Learn and (appropriately) Improve. The company hopes to achieve sustainable growth without raising additional capital by continuing to focus on innovation and customer retention rates, which slipped from 90% to 87% recently. Major clients include Lenovo, David Jones, Harvey Norman, and Paul Smith.
Chelsea Kerwin, June 14, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Sillycon Valley Antics: Hulk, Hillary
June 13, 2016
I noted two items which reminded me why I enjoy Sillycon Valley techno wizardry. The first item concerns the Hulk Hogan Gawker matter. The story “Gawker Files for Bankruptcy and Says It Will Sell the Company to Ziff Davis or Someone Else” converted to a quasi emoji in my addled goose brain; to wit:
My hunch is that anyone who wants to annoy the founder of Palantir Technologies, may want to consider the risks. That splat is ugly and may be blended with an aniline dye.
The other item makes clear that the Alphabet Google thing is an objective algorithmic construct, kissed by the golden Sillycon Valley sun. Navigate to “There’s No Evidence That Google Is Manipulating Searches to Help Hillary Clinton.” Therein resides the truth. I learned:
Apparently, Google has a policy of not suggesting that customers do searches on people’s crimes. I have no inside knowledge of why it runs its search engine this way. Maybe Google is just uncomfortable with having an algorithm suggesting that people search for other people’s crimes. In any event, there’s no evidence that this is specific to Hillary Clinton, and therefore no reason to think this is a conspiracy by Google to help Clinton win the election.
Definitely rock solid from a person whose brother works at Google. Even more reason to accept the Sillycon Valley objectivity argument.
Stephen E Arnold, June 13, 2016
Security, Privacy Struggle with Change
June 13, 2016
I read “NSA Looking to Exploit Internet of Things, Including Biomedical Devices, Official Says.” Let’s assume that the information in the write up is accurate. The message the article seems to convey is that “investing” in systems and methods created a problem for investigators.
I read the article as saying, “Hey, this change stuff is a problem.” Perhaps the solution to the privacy and security tug of war is a flood of changes from software and hardware vendors?
Just a thought.
I don’t like change and for good reasons.
Stephen E Arnold, June 13, 2016
The Time Google Flagged Itself for Potentially Malicious Content
June 13, 2016
Did you know Google recently labeled itself as ‘partially dangerous’? Fortune released a story, Google Has Stopped Rating ‘Google.com’ as ‘Partially Dangerous’, which covers what happened. Google has a Safe Browsing tool which identifies potentially harmful websites by scanning URLs. Users noticed that Google itself was flagged for a short time. Was there a rational explanation? This article offers a technology-based reason for the rating,
“Fortune noted that Google’s Safe Browsing tool had stopped grading its flagship site as a hazard on Wednesday morning. A Google spokesperson told Fortune that the alert abated late last night, and that the Safe Browsing service is always on the hunt for security issues that might need fixing. The issue is likely the result of some Google web properties hosting risky user-generated content. The safety details of the warning specifically called out Google Groups, a service that provides online discussion boards and forums. If a user posted something harmful there, Google’s tool would have factored that in when assessing the security of the google.com domain as a whole, a person familiar with the matter told Fortune.”
We bet some are wondering whether this is a reflection of Google management or the wonkiness of Google’s artificial intelligence? Considering hacked accounts alone, it seems like malicious content would be posted in Google Groups fairly regularly. This flag seems to be a flag for more than the “partially dangerous” message spells out. The only question remaining is, a flag for what?
Megan Feil, June 13, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Axcelerate Focuses on Control and Visibility
June 13, 2016
The article on CMSWire titled Recommind Adds Muscle to Cloud e-Discovery relates the upgrades to the Axcelerate e-Discovery platform from Recommind. The muscle referred to in the article title is the new Efficiency Scoring feature offered to increase e-discovery review process transparency by tracking efficiency and facilitating a consistent assessment. The article explains,
“Axcelerate Cloud is built on Recommind’s interactive business intelligence layer to give legal professionals a depth of insight into the e-discovery process that Recommind says they have previously lacked. Behind all the talk of agility and visibility, there is one goal here: control. The company hopes this release allays the fears of legal firms, who traditionally have been reluctant to use cloud-based software for fear of compromising data.”
Hal Marcus, Director of Product Marketing at Recommind, suggested that in spite of early hesitancy by legal professional to embrace the cloud, current legal teams are more open to the possibilities available through consolidation of discovery requirements in the cloud. According to research, there are no enterprise legal departments without cloud-based legal resources related to contract management, billing, or e-discovery. Axcelerate Cloud aims to promote visibility into discovery practices to address the major concern among legal professionals: insufficient insight and transparency.
Chelsea Kerwin, June 13, 2016
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Quote to Note: Flip the Switch
June 12, 2016
I am not sure if this statement is in context or is 100 percent accurate. However, I found the information in “Eric Schmidt Dismissed the AI fears Raised by Stephen Hawking and Elon Musk” interesting. I love it when “smart” people disagree. Allegedly the electric auto person and the person holding the Newton chair see smart software as a potential problem for humanoids.
Not the Alphabet Google thing. Eric Schmidt compared fears of these two allegedly smart humans’ ideas to a movie. The “reality”, according to the Alphabet Google thing, is:
The scenario you’re just describing is the one where the computers get so smart is that they want to destroy us at some point in their evolving intelligence due to some bug. My question to you is: don’t you think the humans would notice this, and start turning off the computers?
That’s right. A big red button. When I try to kill a Microsoft Windows process or struggle with a GPS which cannot locate me, I just turn off the system. Works every time. Almost. What was that song HAL enjoyed in 2001, Dave?
Systems work really well. For example, lines at airports, Flint water systems, the DC red line, trickle down economics, Itanium CPUs.
Stephen E Arnold, June 12, 2016
Google Nuggets for Summer 2016
June 11, 2016
I printed out several articles which contained information new to me. Here’s the run down:
- I found the suggestion that artificial intelligence could go off the rails amusing. The Alphabet Google thing is one the job, however. Here’s the story about a kill switch for smart software: “Google Is Working on a Kill Switch to Prevent an AI Uprising.” Like the Boston Dynamics robots and the many assertions that Android was not fragmented, the “working on a kill switch” sounds like smart software is not a problem. Terminator thoughts, anyone?
- Google was not included in the write up “Tech Moguls Such as Musk and Bezos Declare Era of Artificial Intelligence.” Poor Google. It sees an opportunity.
- I like it when companies assert that they are able to deliver. Self referential recursiveness public relations, perhaps? Navigate to “You Can Totally Trust Your Privacy with Google, Says Google.” When I worked in Washington, I thought I spotted some Google types at a certain three letter agency. Must have been my imagination. When I read about “totally”, I think about the Google Yahoo GoTo.com/Overture matter. Totally.
- The wonderful Daily Mail in merrie old England reported “Google Is Accused of Burying Results for Popular Pro Brexit Website on Second Page of Search Results.” I thought smart software and algorithms handled relevance. Maybe some real journalists in the UK are not aware of Google’s objective systems and methods.
Interesting PR flow from the Alphabet Google thing. Is it possible for a smart software company to protest too much?
Stephen E Arnold, June 11. 2016
-
- Subscribe to Beyond Search
Feature archive
News archive
-
