IBM Watson and Health: Take Two Aspirin, Do Not Call Me in the Morning
March 8, 2021
IBM Watson was going to put cancer in the cupboard with AS/400 manuals. Then the billion dollar brainiac was going to deal with the Covid Rona thing. Neither worked out.
Ever since Watson blew the competition away on Jeopardy, IBM boasted that their supercomputer would enhance and/or repair industries. The biggest mountain IBM wanted Watson to scale was healthcare and MarketScreener shares: “International Business Machines: IBM’s Retreat From Watson Highlights Broader AI Struggles In Health.”
IBM speculated that AI and machine learning would revolutionize the healthcare industry, so they invested billions in Watson Health. Watson Health was a unit dedicated to developing an AI product that could diagnose and cure cancer. The unit was not profitable and IBM is now selling it.
Google’s DeepMind also invested in healthcare AI programs, but they too lost money and privacy on health data was a big concern.
The biggest roadblock, like all AI endeavors, is the lack of data and insights into the healthcare field:
“The stumbles highlight the challenges of attempting to apply AI to treating complex medical conditions, healthcare experts said. The hurdles include human, financial and technological barriers, they said. Having access to data that represents patient populations broadly has been a challenge, the experts say, as have gaps in knowledge about complex diseases whose outcomes often depend on many factors that may not be fully captured in clinical databases.
Tech companies also sometimes lack deep expertise in how healthcare works, adding to the challenge of implementing AI in patient settings, according to Thomas J. Fuchs, Mount Sinai Health System’s dean of artificial intelligence and human health.”
IBM has not given up on healthcare entirely. Watson Health did have some small successes, but in order to nab a profit IBM needs to sell its excess and concentrate on smaller initiatives.
IBM tried to make sweeping changes by casting a wide net, instead of focusing on smaller steps towards the big picture. Marketing is easier than building systems that live up to the collateral written by MBAs and art history majors it seems.
Whitney Grace, March 8, 2021
Racist Algorithms Sure but Chess?
March 8, 2021
When training software to hunt for offensive speech, some fine-tuning is to be expected. Science Magazine reports that “AI May Mistake Chess Discussions as Racist Talk.” When popular YouTube chess player Antonio Radi? found his account blocked for “harmful and dangerous” content, researchers at Carnegie Mellon University’s Language Technologies Institute (LTI) suspected YouTube’s algorithm may have been confused by the language of “black vs. white” used in discussions of the game. (Radi?’s account was reinstated within 24 hours.) The brief write-up tells us:
“To see if this was feasible, [LTI project scientist Ashiqur R.] KhudaBukhsh and Rupak Sarkar, an LTI course research engineer, tested two state-of-the-art speech classifiers — a type of AI software that can be trained to detect indications of hate speech. They used the classifiers to screen more than 680,000 comments gathered from five popular chess-focused YouTube channels. They then randomly sampled 1,000 comments that at least one of the classifiers had flagged as hate speech. When they manually reviewed those comments, they found that the vast majority — 82% — did not include hate speech. Words such as black, white, attack and threat seemed to be triggers, they said…. “As with other AI programs that depend on machine learning, these classifiers are trained with large numbers of examples and their accuracy can vary depending on the set of examples used. For instance, KhudaBukhsh recalled an exercise he encountered as a student, in which the goal was to identify ‘lazy dogs’ and ‘active dogs’ in a set of photos. Many of the training photos of active dogs showed broad expanses of grass because running dogs often were in the distance. As a result, the program sometimes identified photos containing large amounts of grass as examples of active dogs, even if the photos didn’t include any dogs.”
And an AI trained with few examples of chess discussions may identify passages as offensive, even if the conversation doesn’t include any racists. As KhudaBukhsh observes, we heard about Antonio Radi?’s experience because he is a relatively high-profile YouTube creator. How many lesser-known chess enthusiasts have been similarly shut down on that platform or others? We also wonder what similar mistakes algorithms might be making with other topics. Ah, smart software in action.
Cynthia Murrell, March 8, 2021
Facebook WhatsApp, No Code Ecommerce, and Google: What Could Go Wrong?
March 5, 2021
The Dark Web continues to capture the attention of some individuals. The little secret few pursue is that much of the Dark Web action has shifted to encrypted messaging applications. Even Signal gets coverage in pot boiler novels. Why? Encrypted messaging apps are quite robust convenience stores? Why go to Ikea when one can scoot into a lightweight, mobile app and do “business.” How hard is it to set up a store, make its products like malware or other questionable items available in WhatsApp, and start gathering customers? Not hard at all. In fact, there is a no code wrapper available. With a few mouse clicks, a handful of images, and a product or service to sell, one can be in business. The developer – an outfit called Wati – provides exactly when the enterprising marketer requires. None of that Tor stuff. None of the Amazon police chasing down knock off products from the world’s most prolific manufacturers. New territory, so what could go wrong. If you are interested in using WhatsApp as an ecommerce vehicle, you can point your browser to this Google Workspace Marketplace. You will need both a Google account and a WhatsApp account. Then you can us “a simple and powerful Google Sheet add-on to launch an online store from Google Sheets and take orders on WhatsApp.” How much does this service cost? The developer asserts, “It’s free forever.” There is even a video explaining what one does to become a WhatsApp merchant. Are there legitimate uses for this Google Sheets add on? Sure. Will bad actors give this type of service a whirl? Sure. Will Google police the service? Sure. Will Facebook provide oversight? Sure. That’s a lot of sures. Why not be optimistic? For me, the Wati wrapper is a flashing yellow light that a challenge to law enforcement is moving from the Dark Web to apps which are equally opaque. Progress? Nope.
Stephen E Arnold, March 5, 2021
An Existential Question: LinkedIn or LinkedOut?
March 5, 2021
Writer Joan Westenberg is over LinkedIn, and advises us we would all be better without it. The Next Web posts, “Delete LinkedIn—You’ll Have Zero F****ing Regrets.” After years of enduring countless messages from those who want to sell her something, she finally deleted her LinkedIn account. Not only did the platform fail to provide her any professional benefits, she was also disheartened by the superficial relationships with her hundreds of contacts. (At least this platform does not call them “friends.”)
Having had some success at sales for her business, Westenberg has observed that the way to sell to someone is to build a real relationship with them. Her favorite way to do so is to offer help with no agenda, to demonstrate her products have value. She writes:
“That is the antithesis of LinkedIn. Where people send you off-brand and clumsy sales pitches at best — or at worst, scrape your details for scalable and utterly useless outbound campaigns. They send pitch decks in the same breath that they introduce themselves for the first time. They want you to buy with no reason why. LinkedIn feels less like a platform for selling, and more like a platform for being sold to. A LinkedIn message is the 2020s equivalent of a cold sales call. You dread it. You hate it. You just don’t want to deal with it. … I would rather focus my attention on platforms where I know people have come to genuinely research, interact, learn and consume. Quora. Angel List. Dribble. Medium. Substack. And yes, Twitter. And I would rather remove the false sense of accomplishment we get from engaging on LinkedIn, where we log into a landfill of utter [excrement] several times a day and feel like we’ve done our bit of networking and growing, with no evidence to support that belief.”
Westenberg advises others to join her in ditching the platform. All we will lose, she concludes, are the vanity metrics of clicks, likes, shares, and comments, all of which provide nothing of value. Hmm. I for one have never gotten a job through the platform, but I do know someone who has. Then there are all the professional courses the platform acquired when it snapped up Lynda.com in 2015, many of which are quite helpful. I suppose each user must weigh the site’s role in their professional lives for themselves, but on this point I agree—LinkedIn is not fundamental to professional success. No one should feel they have to use it by default.
Cynthia Murrell, March 5, 2021
Google Gets Kicked Out of Wizard Class: Gibru Jibberish to Follow
March 5, 2021
I read “AI Ethics Research Conference Suspends Google Sponsorship.” Imagine, a science club type organization suspended. Assuming the “real” and ad-littered story is accurate, here’s the scoop:
The ACM Conference for Fairness, Accountability, and Transparency (FAccT) has decided to suspend its sponsorship relationship with Google, conference sponsorship co-chair and Boise State University assistant professor Michael Ekstrand confirmed today. The organizers of the AI ethics research conference came to this decision a little over a week after Google fired Ethical AI lead Margaret Mitchell and three months after the firing of Ethical AI co-lead Timnit Gebru. Google has subsequently reorganized about 100 engineers across 10 teams, including placing Ethical AI under the leadership of Google VP Marian Croak.
The Association for Computing Machinery no less. How many Googlers and Xooglers are in this ACM entity? How many Google and Xoogle papers has the ACM accepted? Now suspended. Yikes, just a high school punishment for an outfit infused with the precepts of high school science club management and behavior.
What’s interesting is the injection of the notion of “ethical.” The world’s largest education and scientific organization is not into talking, understanding the Google point of view, or finding common ground.
Disruptors, losers, and non-fitting wizards and wizardettes are not appropriate for the ethic sub group of ACM. Oh, is that ethical? Good question.
But ACM knows who writes checks. The ad besotted article states:
Putting Google sponsorship on hold doesn’t mean the end of sponsorship from Big Tech companies, or even Google itself. DeepMind, another sponsor of the FAccT conference that incurred an AI ethics controversy in January, is also a Google company. Since its founding in 2018, FAccT has sought funding from Big Tech sponsors like Google and Microsoft, along with the Ford Foundation and the MacArthur Foundation. An analysis released last year that compares Big Tech funding of AI ethics research to Big Tobacco’s history of funding health research found that nearly 60% of researchers at four prominent universities have taken money from major tech companies.
Should I raise another question about the ethics of this wallet sensitive posture? Nah. Money talks.
I find the blip on the ethical radar screen quite amusing. One learns each day what really matters in the world of computers and smart software. That’s a plus.
I am waiting for Google Gibru gibberish to explain the situation. I am all ears.
Stephen E Arnold, March 5, 2021
Cloud or Not? Fighting Words for Sure
March 5, 2021
I read “SolarWinds Hack Pits Microsoft against Dell, IBM over How Companies Store Data.” Ah, ha, a dispute with no clear resolution. The write up suggests that some big dogs in technology will be fighting over the frightened gazelles. Will the easily frightened commercial buyers take off when the word “cloud” is voiced. Or, will the sheep-inspired animals head for the perceived security of computers in the farm house?
The write up states:
[The dispute over where to put data] pits Microsoft Corp., which is urging clients to rely on cloud-computing systems, against others including Dell Technologies Inc. and International Business Machines Corp., who argue customers want to mix the cloud with the more traditional on-premise data-storage systems in a construct called hybrid-cloud.
Do you want pickle on top of a hamburger or underneath the juicy patty? Which method? Come on. Decide.
The write up reports:
Microsoft, one of the world’s biggest cloud vendors, has said cloud services offer customers the most robust data protection. A mixed approach “creates an additional seam that organizations need to secure. A consequence of this decision is that if the on-premises environment is compromised, this creates opportunities for attackers to target cloud services,” Microsoft said in a blog post on its investigation of the hack. The notion that the hybrid cloud is less secure is inaccurate, said Paul Cormier, chief executive of Red Hat, the business IBM acquired two years ago in part in a bet on the growing demand for hybrid cloud services. “Any software could get broken into. The cloud providers could get broken into as well,” he told The Wall Street Journal.
Plus the article points out:
Microsoft itself was a victim in the attack and had some of its source code used to write software downloaded. The hackers viewed software linked to Microsoft’s Azure cloud, the company said. Mr. Smith, at the Senate hearing on the hack on Tuesday, called for a “full examination of what other cloud services and networks the Russians have accessed.”
I don’t think any computer data are secure, but that’s just me. Here in Harrod’s Creek, professional etch secrets on lumps of boghead. Once the message has been read, one burns it. Good for secrecy, not so good for the environment.
Who will win this battle? The key is marketing. Security is a slippery fish particularly when the boats are owned by Dell, IBM, and Microsoft. The SolarWinds’ attack exploited the cloud and on premises devices. How does one spell “insider threat”? One can unplug computing devices. Put them in a locked room. Don’t let anyone enter the room. Is that a solution?
Stephen E Arnold, March 5, 2021
Saudi Influence in Silicon Valley
March 5, 2021
Data scientist Vicki Boykis noticed many of the cool kids in tech have something in common—Uber, WeWork, Flexport, Slack, MapBox, and DoorDash have all received financing from Vision Fund, a venture capital firm run by Japanese holding company SoftBank. The firm does not exactly advertise it, at least in the US, but one of its major contributing partners is the Kingdom of Saudi Arabia. Boykis considers the implications on her blog, Normcore Tech, in “Silicon Valley Runs on Saudi.” She ponders:
“What does it mean for relentlessly forward-looking companies like Slack, who publish effusive blog posts about diversity and collaborative leadership, to be fueled in part by money from a government that only recently allowed women to drive, and has a record of flogging bloggers who disagree with the regime? Even probably more importantly, what does it mean for members of the Saudi Public Wealth Fund to be on the boards of these companies and directly calling the shots? As I’ve written about before, being on the board, particularly if you have money, has influence.”
Few have more money than the Saudi crown prince, who has taken a personal interest in US tech companies and invested a huge chunk of personal change into the Vision Fund. The write-up shares some photos of him palling around with Silicon Valley nobility. Boukis writes:
“But of course, it doesn’t hurt that Mohammad bin Salman (MBS), the Crown Prince of Saudi Arabia and the one who made the decision to pledge $45 of the initial $92 billion into fund, is making all the right noise in Silicon Valley. MBS has been making a ton of noise lately about modernizing Saudi Arabia, including starting a long-term investment project out of oil, restricting the power of the religious police, and giving women the ability to drive. On the surface, it all looks great, particularly when he engages with Silicon Valley.”
Below the surface, however, we cannot know what decisions have been informed by Saudi values. True, the Crown Prince is a progressive—compared to others in Saudi Arabia. That is not saying much. It looks like his influence is waning a bit, we’re told, but other nations are taking an interest in Silicon Valley—like Kazakhastan. Not better. We know tech companies need funding. So go where the money is.
Cynthia Murrell, March 5, 2021
The Microsoft Supply Chain Works Even Better Going Backwards
March 4, 2021
Do you remember the character KIR-mit. He once allegedly said:
Yeah, well, I’ve got a dream too, but it’s about singing and dancing and making people happy. That’s the kind of dream that gets better the more people you share it with.
I am not talking about Jim Henson’s memorable character. That frog spelled its name Kermit. This is KIR-mit, an evil doppelgänger from another universe called Redmonium.
This KIR-mit is described in “Microsoft Is Using Known Issue Rollback (KIR) to Fix Problems Caused by Windows 10 Updates.” I learned that KIR
enables Microsoft to rollback changes introduced by problematic patches rolled out through Windows Update. KIR only applies to non-security updates.
Does the method expand the attack service for bad actors? Will weird calls to senior citizens increase with offers to assist with KIR-mit modifications? Will questionable types provide links to download KIRs which are malware? Yes, yes, and yes.
The article points out:
Known Issue Rollback is an important Windows servicing improvement to support non-security bug fixes, enabling us to quickly revert a single, targeted fix to a previously released behavior if a critical regression is discovered.
KIR is something users have said they wanted. Plus Microsoft has had this capability for a long time. I recall reading that Microsoft had a method for verifying the “digital birth certificate” of software in order to identify and deal with the SolarWinds-type of supply chain hack. I point this out in my upcoming lecture for a law enforcement entity. Will my audience find the statement and link interesting? I have a hunch the cyber officers will perk up their ears. Even the JEDI fans will catch my drift.
Just regular users may become woozy from too much KIR in the system. Plus, enterprise users will be “in charge of things.” Wonderful. Users at home are one class of customers; enterprise users are another. In between, attack surface the size of the moon.
Several questions:
- Why not improve the pre release quality checks?
- Why not adopt the type of practices spelled out by In Toto and other business method purveyors?
- Why not knock off the crazy featuritis and deliver stable software in a way that does not obfuscate, mask, and disguise what’s going on?
And the answers to these questions is, “The cloud is more secure.”
Got it. By the way a “kir” is a French cocktail. Some Microsoft customers may need a couple of these to celebrate Microsoft’s continuous improvement of its outstanding processes.
As KIR-mit said, “It’s about making people happy.” That includes bad actors, malefactors, enemies of the US, criminals, and Microsoft professionals like Eric Vernon and Vatsan Madhava, the lucky explainers of KIR-mit’s latest adventure.
Stephen E Arnold, March 4, 2021
Facebook Found Lax in Enforcement of Own Privacy Rules. Surprised?
March 4, 2021
Facebook is refining its filtering AI for app data after investigators at New York’s Department of Financial Services found the company was receiving sensitive information it should not have received. The Jakarta Post reports, “Facebook Blocks Medical Data Shared by Apps.” Facebook regularly accepts app-user information and feeds it to an analysis tool that helps developers improve their apps. It never really wanted responsibility for safeguarding medical and other sensitive data, but did little to block it until now. The write-up quotes state financial services superintendent Linda Lacewell:
“Facebook instructed app developers and websites not to share medical, financial, and other sensitive personal consumer data but took no steps to police this rule. By continuing to do business with app developers that broke the rule, Facebook put itself in a position to profit from sensitive data that it was never supposed to receive in the first place.”
Facebook is now stepping up its efforts to block sensitive information from reaching its databases. We learn:
“Facebook created a list of terms blocked by its systems and has been refining artificial intelligence to more adaptively filter sensitive data not welcomed in the analytics tool, according to the report. The block list contains more than 70,000 terms, including diseases, bodily functions, medical conditions, and real-world locations such as mental health centers, the report said.”
A spokesperson says the company is also “doing more to educate advertisers on how to set-up and use our business tools.” We shall see whether these efforts will be enough to satisfy investigators next time around.
Cynthia Murrell, March 4, 2021
Email: A Vulnerable Service
March 4, 2021
Cyber security firm Barracuda counted the number of email attacks that slipped through its clients’ enterprise-wide security measures last year. New Zealand’s SecurityBrief reveals the results in, “Millions of Email Attacks Missed by Organizations’ Cyber Security Protection.” Writer Shannon Williams reports:
“In 2020, 4550 organizations used Barracuda Email Threat Scanner to scan 2,600,531 unique mailboxes and found 2,029,413 unique attacks. On average, 512 attacks were found per organization, and one out of seven mailboxes (14%) had at least one attack currently sitting inside, even if messages were scanned by an email gateway solution, the cyber security firm says. The attacks detected fall into four email threat types: phishing, scamming, extortion, and business email compromise (BEC). Of the 2,029,413 unique attacks detected, phishing was the number one threat missed by the organizations email security solutions (59%). Scamming was the second most common (39%). Extortion, at 9%, and BEC, at 8%, were less prevalent, but cybercriminals tend to send these types of attacks in smaller volumes because they are highly personalized.”
Barracuda recommends companies adopt its inbox-based Email Threat Scanner to detect attacks that slip through any broader security measures. What a surprise! Of course, since the organizations studied were already Barracuda clients, it is entirely possible at least some of them were relying on that solution and skimping on gateway-side security. Even so, the report is a reminder to take email security seriously. One could choose a product like Barracuda’s, if desired. (Or Cyren’s, to name just one competitor.) At the least, workers should learn what to look for and actively avoid opening attack emails should they land in their inboxes. And turn off preview pane, for goodness’ sake.
Founded in 2003. The firm states over 200,000 customers around the world use its software, which some say is effective, affordable, and user-friendly.
Cynthia Murrell, March 4, 2021
-
- Subscribe to Beyond Search
Feature archive
News archive
-
