Monthly Archives: November 2016

Welcome Devs With Open Arms

Two leaders in the field of artificial intelligence have announced that they’re open-sourcing their AI platforms.

After investing in building rich simulated environments to serve as laboratories for AI research, Google’s DeepMind Lab on Saturday said it would open the platform for the broader research community’s use.

DeepMind has been using its AI lab for some time, and it has “only barely scratched the surface of what is possible” in it, noted team members Charlie Beattie, Joel Leibo, Stig Petersen and Shane Legg in an online post.

By open-sourcing the platform, DeepMind Lab hopes to open new opportunities for developers to make significant contributions to AI.

Meanwhile, OpenAI, which is cochaired by Tesla CEO Elon Musk, on Monday invited developers to try on its Universe platform for size.

It’s hoping an influx of development talent will help it achieve its overarching mission: to create a single AI agent that can be flexible in applying its past experience within Universe to quickly master unfamiliar, difficult environments.

Now that it’s open source, the platform is available on GitHib, a home for many developers on the Web.

Developers will be able to add custom levels to its platform via GitHub, the DeepMind team explained. In addition, all DeepMind assets will be hosted on GitHub. along with code, maps and level scripts.

DeepMind hopes the GitHub community will help it shape and develop the platform going forward, said Beattie, Leibo, Petersen and Legg.

OpenAI is designed to allow an AI agent to use a computer as a human does.

OpenAI wants to train AI systems on a full range of tasks to solve, it noted in an online post.

Universe enables the training of a single agent to perform any task a human can complete with a computer, according to OpenAI.

 

Sign of Frustration?

Computer scientists for some time have been trying to make algorithms learn from patterns, but progress has been slow.

“There’s a little bit of frustration that we’re not making as much progress as we wanted to,” maintained Sorin Adam Matei, an associate professor atPurdue University.

“This attempt to get more people involved, to excite people, to get another angle to the problem is a sign of frustration,” he told LinuxInsider.

Nevertheless, opening the AI platforms to the at-large development community could be beneficial.

“If a lot of people are attracted to these tools, we’ll see new creative interesting services,” Matei pointed out. “There could be a very healthy lateral development.”

What’s more, that development may come faster and cheaper than if DeepMind and OpenAI continued to go it alone.

“Certainly, part of the motivation [to go open source] is rooted in a desire to innovate quickly and cost-effectively,” said Austin Ogilvie, CEO of Yhat.

Mysteries of the Mind Facebook

Facebook last week released six videos to educate people about artificial intelligence.

AI will bring major changes to society, and will be the backbone of many of the most innovative apps and services of the future, but it remains mysterious, noted Yann LeCun, Facebook’s director of AI research, and Joaquin Candela, the company’s director of applied machine learning in an online post.

The videos are “simple and short introductions” that will “help everyone understand how this complex field of computer science works,” they said.

The announcement immediately sparked speculation that Facebook would use its AI technology to tackle the fake news items posted on its pages.

However, tackling fake news is more a question of ethics than technology, LeCun told reporters, with questions coming into play about the trade-off between filtering and censorship versus free expression and democracy.

 

Mysteries of the Mind

Understanding the videos, which LeCun created, does require a certain amount of grounding in science.

Still, they “set a baseline — help explain what an AI can do in the near term,” said Rob Enderle, principal analyst at the Enderle Group.

Watching the videos “likely makes the topic less frightening,” he told TechNewsWorld.

Facebook “is trying to reset the perception of AI,” and from that standpoint, they’re dead-on, observed Jim McGregor, a principal analyst at Tirias Research.

“This is what the market needs,” he told TechNewsWorld, to counter “unrealistic Hollywood scenarios and sensationalism in the general press.”

Facebook’s long-term goal is to understand intelligence and build intelligent machines, according to LeCun and Candela.

Trying to understand intelligence and how to reproduce it in machines will “help us not just build intelligent machines,” they suggested, “but develop keener insight into how the mysterious human mind and brain work.”

 

Corralling Fake News

AI might help clamp down on the proliferation of fake news stories, which circulated in unprecedented numbers during the recent presidential election, and may have influenced voters.

One motive for the creation of fake news stories is money, according to one fake-news writer who told The Washington Post that he believed his made-up stories helped President-Elect Donald Trump win votes.

In separate investigations, both BuzzFeed and The Guardian earlier this year found upwards of 100 pro-Trump sites in Macedonia — many apparently run by teenagers looking to make a quick buck.

Fake news also comes from partisan news sites slanted strongly toward one candidate.

Lets Talking About Foreign Election Control

A few years back, when it was one company, HP made a huge mistake that cost a number of people their jobs and forced the replacement of many of its board members. The company suffered through some nasty litigation and several top executives almost landed in jail.

The mistake was tied back to something the board authorized, which at the time was called “pretexting.” It also went by the more common term “identity theft.” It is my belief that the board wouldn’t have authorized the effort if it had been told that what the teams planned to do was steal the identities of reporters.

Given how risk-averse boards were, and still are, HP’s directors simply would not have been willing to take the risk, in my view, and much of HP’s pain in the last decade could have been avoided.

Given that Russia is the source for much of it, I now wonder if our use of the term “fake news” as a label — as opposed to the older and more relevant term — isn’t doing us a disservice, by not highlighting the inherently evil nature of the practice.

Fake News is intentionally designed to mislead, and it should be treated like propaganda. Blocking propaganda as a matter of law would be far easier to accomplish than blocking “fake news,” because “fake news” seems more benign than “propaganda” — even though, like “pretexting” and “identity theft,” they are the same thing.

I’ll share my thoughts on that and close with my product of the week: a new Magellan Dash camera that might make a decent gift for those needing to document some of the insane drivers on the road, or catch someone messing with their car.

 

There Is a Lot of ‘Fake News’

Now much of the fake news I currently get on Facebook is simply to get me to click a link, often as part of a process to install some form of malware. Often, these stories have been about the death of a celebrity who hasn’t died, but during the election, much of the fake news surrounded things that weren’t true about Hillary Clinton but that clearly were intended to change my vote. They were attempts to change how I viewed a candidate, in order to elicit a reaction.

Given the nature of the false stories and the fact that polls showed Clinton would win anyway, my belief is that the effort was to impede her ability to govern after she won, and the anticipated disclosure of the effort was designed to do the same thing to Trump.

The sure thing for Russia wasn’t to elect Trump or Clinton, but to ensure that whoever won would have such a cloud hanging overhead that neither could really execute. In other words, Russia wasn’t going after a candidate — it was going after the country.

Beyond the idea that another country could have a material impact either on the election or on the effectiveness of the elected candidate is the frightening fact that it happened in a country that has the tools to formulate a proper response but chose not to use them.

As initial attempts go, this was a powerful one. Given the propagation of ever more intelligent tools to create increasingly more targeted messages, it means a foreign power with adequate funds — like Russia or China — could gain near-absolute control over who gets elected in the U.S. That’s troubling — particularly given that the U.S. developed the tools both to carry out and to defend against such a strategy.

 

Defending Against Foreign Election Control

Clearly, there are free speech and censorship issues with regard to the identification and elimination of fake news, but with analytics, we can identify both trends and the organized manipulation of facts that go viral.

That is why switching from the name “fake news” to the name “propaganda” when a foreign, criminal or terrorist organization is generating this “news” could go a long way toward reducing its impact.

Once it’s identified, there are tools that can explain to people that the news they are seeing isn’t fact-based, and/or source the information so people understand there may be inherent bias.

Facebook is setting up to do this, and I expect it will turn into something of an arms race, with artificial intelligence agents on both sides trying to outsmart each other. Still, if the country with Silicon Valley can not win this fight against the countries without it, we likely should just take our toys and go home.

Halts Malware Avalanche on Multinational Effort

The U.S. Department of Justice on Monday released new details about the multinational takedown of Avalanche, a multimillion-dollar malware and money-laundering network, following a four-year probe led by German police and prosecutors. Assistant Attorney General Leslie R. Caldwell, Acting U.S. Attorney Soo C. Song and Assistant Director Scott S. Smith of the FBI’s Cyber Division made the announcement in Pittsburgh.

Prosecutors and investigators in 40 countries were involved in the probe, led by the Public Prosecutor’s office in Verden, Germany, and police in Luneberg. They received assistance from the DoJ, Eurojust and Europol.

The investigation uncovered a multinational malware campaign that started in 2009, sending out more than a million infected emails with damaging links and attachments.

The cyberthieves managed to use the information to transfer funds from the accounts of unsuspecting victims after stealing their bank and email passwords. The stolen funds, estimated to be in the hundreds of millions of euros, then were redirected to other criminals through a double fast flux infrastructure.

More than 20 families of malware were used, including goznym, marcher, matsnu, urizone, xswkit and pandabaker. Highly organized networks of “mules” bought goods with stolen funds, enabling the cyberthieves to launder money they obtained through the scheme.

 

Arrests, Seizures

The takedown operation marks the largest ever use of sinkholing to combat botnet infrastructures, and is unprecedented in scale, involving more than 800,000 domains seized, sinkholed or blocked.

Five suspects were arrested, 37 premises were searched, and 39 servers were seized, officials said last week. Victims were found in 180 different countries, and 221 servers were kicked offline through abuse notifications sent to hosting providers.

U.S. District Judge Arthur Schwab late last month granted federal prosecutors a temporary restraining order allowing them to block and reroute data from the infected computers used in the Avalanche malware scheme to prevent further malicious activity, according to Margaret Philbin, spokesperson for the U.S. Attorney’s office in Pittsburgh.

The order essentially allows the illegal data to be trapped and traced over to government controlled systems that can track the illegal activity and protect victims of the scheme.

“People across the globe, including residents and companies here in western Pennsylvania, have been victimized by Avalanche and the malware distributed using its intricate infrastructure,” said Robert Johnson, special agent in charge of the FBI in Pittsburgh.

At least three companies or government entities in Pennsylvania were impacted by the attacks, based on court filings.

From Feb. to April of this year, a New Castle-based firm was targeted with seven unauthorized wire transfers that totaled more than US$243,000, based on an attack using GozNym malware. The wire transfers were stopped before any money was lost.

In January 2015, a government entity in Allegheny County was victim to a Nymaim malware attack and had to pay 6 bitcoins, or about $1,400, to get a decryption tool to rescue its files.

In April of this year a Carnegie business was victim of an ATO fraud using GozNym malware attack that resulted in $387,500 being fraudulently transferred from a Pittsburgh-based financial institution to an account in Bulgaria.

“This investigation highlights once again that through the international cooperation of law enforcement and private industry, we can be as effective investigating criminals in cyberspace as we are on the streets of our communities here at home,” Johnson pointed out.