Start Up: Wikipedia’s AI effort, Krebs wins again, Uber’s harassment problem, Hololens on hold?, and more

The trouble with IBM’s Watson is that it doesn’t, by a long stretch – and that has people asking questions about its value. Photo by stewf on Flickr.

You can now sign up to receive each day’s Start Up post by email. You’ll need to click a confirmation link, so no spam.

A selection of 8 links for you. Oh well. I’m @charlesarthur on Twitter. Observations and links welcome.

Inside Wikipedia’s attempt to use artificial intelligence to combat harassment • Motherboard

Sarah Smellie:


A paper published last week on the arXiv preprint server by the Detox team offers the first look at how Wikimedia is using AI to study harassment on the platform. It suggests that abusive comments aren’t the domain of any specific group of trolls, and that diverse tactics are going to be needed to combat them on Wikipedia. 

“This is not ground-breaking machine learning research,” said Ellery Wulczyn, a Wikimedia data scientist and Detox researcher, in a telephone interview. “It’s about building something that’s fairly well known but allows us to generate this data scale to be able to better understand the issue.”

The goal at Jigsaw, an Alphabet tech incubator that began as Google Ideas, is nothing short of battling threats to human rights and global security. Their projects include a map that shows the sources and targets of global DDoS attacks in real time, and an anti-phishing extension for Chrome originally developed to protect Syrian activists from hackers.

To get their algorithm to recognize personal attacks, the Detox team needed to train them on a solid data set. They started with 100,000 comments from Wikipedia talk pages, where editors hash out their disagreements. Next, 4,000 crowdworkers evaluated the comments for personal attacks. Each comment was inspected by 10 different people.


Certainly a place to go for personal attacks. For another view, on getting humans to make Wikipedia better, there’s this: How Wikipedia is cultivating an army of fact checkers to battle fake news.
link to this extract

Men who sent swat Team, heroin to my home sentenced • Krebs on Security

Brian Krebs:


It’s been a remarkable week for cyber justice. On Thursday, a Ukrainian man who hatched a plan in 2013 to send heroin to my home and then call the cops when the drugs arrived was sentenced to 41 months in prison for unrelated cybercrime charges. Separately, a 19-year-old American who admitted to being part of a hacker group that sent a heavily-armed police force to my home in 2013 was sentenced to three years probation.


Krebs, who does remarkable work, is proof that right can win over wrong if you just persist. (And that hackers are terribly susceptible to hubris.)
link to this extract

Reflecting on one very, very strange year at Uber • Susan J. Fowler

She went to work at Uber:


After the first couple of weeks of training, I chose to join the team that worked on my area of expertise, and this is where things started getting weird. On my first official day rotating on the team, my new manager sent me a string of messages over company chat. He was in an open relationship, he said, and his girlfriend was having an easy time finding new partners but he wasn’t. He was trying to stay out of trouble at work, he said, but he couldn’t help getting in trouble, because he was looking for women to have sex with. It was clear that he was trying to get me to have sex with him, and it was so clearly out of line that I immediately took screenshots of these chat messages and reported him to HR.

Uber was a pretty good-sized company at that time, and I had pretty standard expectations of how they would handle situations like this. I expected that I would report him to HR, they would handle the situation appropriately, and then life would go on – unfortunately, things played out quite a bit differently. When I reported the situation, I was told by both HR and upper management that even though this was clearly sexual harassment and he was propositioning me, it was this man’s first offense, and that they wouldn’t feel comfortable giving him anything other than a warning and a stern talking-to. Upper management told me that he “was a high performer” (i.e. had stellar performance reviews from his superiors) and they wouldn’t feel comfortable punishing him for what was probably just an innocent mistake on his part.

I was then told that I had to make a choice: (i) I could either go and find another team and then never have to interact with this man again, or (ii) I could stay on the team, but I would have to understand that he would most likely give me a poor performance review when review time came around, and there was nothing they could do about that. I remarked that this didn’t seem like much of a choice, and that I wanted to stay on the team because I had significant expertise in the exact project that the team was struggling to complete (it was genuinely in the company’s best interest to have me on that team), but they told me the same thing again and again.


And guess what? She left the team, spoke to other women, and they had had the same kind of problem. In some cases, previously with the same man.

It also sounds like an absolute rats’ nest in there as well. And guess what? Once this had gone viral, Travis Kalanick ordered an immediate investigation.

Charitable explanation: Kalanick didn’t know how screwed up his company has become. However, changing the culture is going to be hard – if he really wants to.
link to this extract

Microsoft accelerates HoloLens V3 development, sidesteps V2 •

Paul Thurrott:


By skipping what was version two on their roadmap, the company can accelerate version three which will be closer to a generational leap and help keep Microsoft ahead of the competition. My sources are telling me that this version of Hololens will not arrive until 2019.

Yes, 2019 is a considerable amount of time away but for Microsoft, if they would have built what was known as version two, the company would not be able to get version three delivered by 2019. In short, the company is making a bet that the advancements they are investing in today for the v3 version of Hololens are significant enough and add enough value to the product that it will make sure they continue to lead the segment by getting that device to the market earlier.

Of course, it’s always possible the device arrives before then but do not expect a new device this year and likely nor will one arrive next year, based on what I have been told. I did reach out to Microsoft for comment and they provided the following statement but it’s generic and doesn’t add any new context to the information already provided:

“Mixed reality is the future of computing, and Microsoft HoloLens is the future and present of mixed reality. Our commitment requires no roadmap”.


So you’re accelerating to leave a two-year gap? 🤔 But on the other hand, I don’t think there will be huge demand for AR systems this year or next. If this is Microsoft’s timeline, it sounds sensible.
link to this extract

Apple acquires Israel firm RealFace, specializing in facial recognition • MacRumors

Tim Hardwick:


RealFace’s website is currently offline, but according to promotional material, the startup had developed a unique facial recognition technology that integrates artificial intelligence and “brings back human perception to digital processes”. RealFace’s software is said to use proprietary IP in the field of “frictionless face recognition” that allows for rapid learning from facial features.

The Israeli startup also developed a now-defunct app called Pickeez, which selected and collated a user’s best photos across various platforms using the RealFace recognition software.

According to iPhone 8 rumors, Apple may ditch Touch ID along with the physical home button, in favor of a facial recognition-capable front-facing 3D laser scanner, although with the RealFace acquisition coming at such a late time, it’s unlikely that the any of the startup’s technology will feature.

RealFace is the fourth Israel-based firm Apple is known to have acquired. In 2011 it bought flash memory maker Anobit for a reported $400m, then in November 2013 it acquired 3D sensor company PrimeSense for an estimated $345m. Most recently in 2015, Apple bought LinX for around $20m.


All the stuff about OMG LAZERS is rather overexcited, and I’m not sure “face unlock” meets the sort of security standards Apple has liked. I could see it being useful for analysing faces in photos, though.
link to this extract

15-year-old Amanda Todd’s alleged sextortionist on trial at last • Daily Beast

Nadette De Visser:


Despite the publicity that surrounded Amanda Todd’s suicide, it took two years before Coban was found and arrested, and the initial break in the case actually came before her death.

In 2012, at the same time Coban was allegedly targeting Amanda, he also was pursuing a Norwegian girl who decided to approach the police about it. They tracked his IP addresses to the Netherlands. The Norwegian police contacted the Dutch police and the predator’s IP address was then traced to the trailer park in Oisterwijk where Aydin Coban lived. But that did not lead to his arrest. It took Todd’s suicide and a report compiled by Facebook in its wake to jolt the Dutch police into further action. In Facebook’s own investigation a relationship was established between a phone number, an IP address and 86 accounts in which it appeared aliases were being used to target young girls. That information then led Dutch police to Coban’s home, where they installed spyware on both of his computers.

In all, Coban has been indicted for 72 alleged offenses related to sexual exploitation and extortion over the internet. At his place, police found a trove of more than 204,000 photos and videos on a partially encrypted hard drive, many of them involving child pornography. The Dutch police also found a drive with 5,800 bookmarked names that served as a database of potential victims and their social networks.
Much of the evidence presented in court has come from the spyware that allowed police to collect every keystroke and multiple screenshots from Coban’s computers.


Depressing. A notable point is how a single person like this can, through the network, cause misery – or worse – for hundreds. Before the internet, their reach would have been much smaller. With the good, the bad.
link to this extract

MD Anderson benches IBM Watson in setback for artificial intelligence in medicine • Forbes

Matthew Herper:


The partnership between IBM and one of the world’s top cancer research institutions is falling apart. The project is on hold, MD Anderson confirms, and has been since late last year. MD Anderson is actively requesting bids from other contractors who might replace IBM in future efforts. And a scathing report from auditors at the University of Texas says the project cost MD Anderson more than $62m and yet did not meet its goals.

“When it was appropriate to do so, the project was placed on hold,” an MD Anderson spokesperson says. “As a public institution, we decided to go out to the marketplace for competitive bids to see where the industry has progressed.”

The disclosure comes at an uncomfortable moment for IBM. Tomorrow, the company’s chief executive, Ginni Rometty, will make a presentation to a giant health information technology conference detailing the progress Watson has made in health care, and announcing the launch of new products for managing medical images and making sure hospitals deliver value for the money, as well as new partnerships with healthcare systems. The end of the MD Anderson collaboration looks bad.

But IBM defended the MD Anderson product, known as the Oncology Expert Advisor or OEA. It says the OEA’s recommendations were accurate, agreeing with experts 90% of the time.


Um. Agreeing with experts 90% of the time means it’s wrong 10% of the time (and no indication on whether that’s false positives, false negatives or both). That has the potential to be very harmful in oncology.

More broadly, I’m seeing a growing groundswell of opinion that IBM’s pushes in AI are all talk, little result. This, while the company’s revenues have been falling for years under Rometty. This story isn’t quite finished yet.
link to this extract

What do you think happened to flight MH370 passengers during its final hour? • Quora

Sy Gunson is a former pilot and worked in airline operations:


Two of [MH370 pilot] Zaharie’s sisters are friends of mine and during a face to face meeting with them in December 2016 they surprised me by revealing after listening to the ATC audio, they could identify the final three radio calls by MH370 were made by their brother and his voice was slurred. They volunteered this fact unprompted.


This is a comprehensive look at what could have happened; he also suggests that the plane broke up in the air due to G-forces as it went into a dive after running out of fuel. His answer is calm and has a lot of detail I hadn’t seen before – and I used to write articles on MH370.
link to this extract

Errata, corrigenda and ai no corrida: none notified

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.