Tuesday, November 26, 2019


You knew this, right? Will it change under CCPA?
The California DMV Is Making $50M a Year Selling Drivers’ Personal Information
DMVs across the country are selling data that drivers are required to provide to the organization in order to obtain a license. This information includes names, physical addresses, and car registration information. California’s sales come from a state which generally scrutinizes privacy to a higher degree than the rest of the country.




Hacking AI. (Huge volumes of data, pointing in all directions. Add a small volume that points to the “answer” you want.)
Tiny alterations in training data can introduce "backdoors" into machine learning models
The attack is related to adversarial examples, a class of attacks that involve probing a machine-learning model to find "blind spots" -- very small changes (usually imperceptible to humans) that cause machine learning classifiers' accuracy to shelve off rapidly (for example, a small change to a model of a gun can make an otherwise reliable classifier think it's looking at a helicopter ).




For my students. (Not sure why a search on my school email says I use it in Columbia.)
Personal Email Security Guide
Everyone uses email, it’s incredibly useful, and you can’t sign-up for accounts or do much of anything online without one. However, email is still a hacker’s favorite route to attacking a target, because most users don’t bother to secure their accounts.
Stay Safe: Make sure to double-check the sender’s email address and domain name for signs of forgery or misspellings. Reputable businesses and banks will never ask for personal and sensitive information via email. If a message asks for your password, credit card details or social security number, that’s a phishing email. The most basic step to do in such an uncertain situation is to use reverse lookup service to find out more information about sender.




Suggesting that regulators are beginning to understand search algorithms?
Google Will Restrict Sharing of User Data for Google Ads Under EU Privacy Pressure
Google is taking yet another step to guarantee that a key revenue generator for the company – its programmatic advertising platform based on Real-Time Bidding (RTB) technology – remains compliant with EU privacy regulations. Under the terms of the European General Data Protection Regulation (GDPR), Google must minimize the amount of user data that it collects and then shares with third parties such as advertisers looking to buy Google Ads. With that in mind, Google’s ad exchange will stop telling advertisers what categories of websites users are visiting starting in February 2020.
Ireland’s top data protection regulator, for example, is taking a closer look at Google’s programmatic advertising technology, based on concerns that data shared with advertisers might enable them to create comprehensive user profiles once they combine their own user data sets with the user data being shared by Google.




Perhaps they haven’t read the Asimov stories where the three laws fail?
Scientists developed a new AI framework to prevent machines from misbehaving
In what seems like dialogue lifted straight from the pages of a post-apocalyptic science fiction novel, researchers from the University of Massachusetts Amherst and Stanford claim they’ve developed an algorithmic framework that guarantees AI won’t misbehave.
The framework uses ‘Seldonian’ algorithms, named for the protagonist of Isaac Asimov’s “Foundation” series, a continuation of the fictional universe where the author’s “Laws of Robotics” first appeared.
According to the team’s research, the Seldonian architecture allows developers to define their own operating conditions in order to prevent systems from crossing certain thresholds while training or optimizing. In essence, this should allow developers to keep AI systems from harming or discriminating against humans.
The current AI development paradigm places the burden of combating bias on the end user. For example, Amazon’s Rekognition software, a facial recognition technology used by law enforcement, works best if the accuracy threshold is turned down but it demonstrates clear racial bias at such levels. Cops using the software have to choose whether they want to use the technology ethically or successfully.
The Seldonian framework should take this burden off the end-user and place it where it belongs: on the developers.




Requires broad access to health records. Introduces new opportunities for bias.
Startup Deep 6 Lands $17 Million To Use AI To Help Find Patients For Clinical Trials
Of the many inefficiencies throughout the long, expensive drug-discovery process, patient recruitment for clinical trials can be a particularly brutal bottleneck. For example, various studies have shown that as many as 40% of trials fail to meet their enrollment goals.
Enter Deep 6, a Pasadena-based startup announcing $17 million in fresh funding at a $50 million valuation for its artificial intelligence-powered technology that can suggest candidates for clinical trials in “minutes instead of months.” It has raised $22 million total.
One of the reasons that clinical trial recruitment takes so long is that the patient health information necessary to determine if someone is a good match for a given trial is spread across electronic medical records (EMRs), physicians notes, pathology reports and other forms of documentation.




Perspective. Why AI is growing?
AI Stats News: Chatbots Increase Sales By 67% But 87% Of Consumers Prefer Humans
Business leaders saved an average of $300,000 in 2019 from their chatbots, with the greatest impact occurring across support and sales teams; the sales function is the most common use case for chatbots (41%), followed closely by support (37%) and marketing (17%); chatbots increased sales by an average of 67%, with 26% of all sales starting through a chatbot interaction; 35% of business leaders said chatbots helped them close sales deals; top automated tasks performed by chatbots are routing website visitors, collecting information, and qualifying leads; chatbots speed up response times by an average of 4x and increase customer support satisfaction scores by 24% [Intercom survey of 500 business leaders]




The WSJ got it wrong? Impossible!
Why many in the search community don’t believe the WSJ about Google search
Follow up to previous posting – an article by WSJ.com – How Google Interferes With Its Search Algorithms and Changes Your Results please read – Search is complicated. The WSJ appeared set on seeing that complexity through a conspiratorial lens. Barry Schwartz on November 18, 2019. “…At first, I thought maybe the Wall Street Journal had uncovered something. But as I read through page after page while being shuttled down the West Side Highway towards my office in West Nyack, New York, I was in disbelief. Not disbelief over anything Google may have done, but disbelief in how the Wall Street Journal could publish such a scathing story about this when they had absolutely nothing to back it up. The subtitle of the story read, “The internet giant uses blacklists, algorithm tweaks and an army of contractors to shape what you see.” This line alone shows a lack of understanding on how search works and why the WSJ report on Google got a lot wrong, as my colleague Greg Sterling reported last week. Google is not certainly perfect, but almost everything in the Wall Street Journal report is incorrect. I’ll go through many of the points [in this article]…”




Another peek at Brave.
There are more competing web browsers than ever, with many serving different niches. One example is Brave, which has an unapologetic focus on user privacy and comes with a radical reimagining of how online advertising ought to work.
Brave was one of the first browsers to include built advertisement and tracker blockers, leapfrogging the likes of Opera. It also came with its own cryptocurrency, called BAT (or Basic Attention Token), allowing users to reimburse the sites and creators they like.
Essentially, Brave wants to re-imagine how the Internet works: not just on a usability level, but on an economic level. It’s an undeniably radical vision, but you wouldn’t expect any less, given its founding team.



No comments: