Chess, Go, StarCraft

Photo from MIT Technology Review – Professional StarCraft player Byun Hyun Woo playing in the 2016 StarCraft II World Championship Series, which he won.

Scientists continue to train AI to compete professionally in classical strategic games like Chess and Go as a sort of basic Turing Test. Now that AI have shown their ability to out-maneuver humans in the latter examples, some consider StarCraft – a strategic multi-player game where players can compete to dominate the map as an alien race – to be AI’s next challenge.

“When you play StarCraft, you have to respond very quickly to lots of uncertainties and variables, but I’ve noticed that AI like AlphaGo isn’t that good at reacting to unexpected scenarios,” Byun says.

Hot

A StarCraft victory for an AI trained via reinforcement-learning would be proof that its intelligence is capable of executing both long and short-term decisions on the fly – and would bring AI one step closer to human-like decision making.

Full article on MIT Technology Review

Who programs the AI? (Not women or people of color)

You might assume that technology and AI are neutral forces in this world. The truth is, our technology is biased and created in the image of its creators – as Melinda Gates and Fei-Fei Li argue in this interview, these are “guys with hoodies.”

Have you ever?

  • Tried on an Oculus Rift to find that the hardware does not fit your facial profile?
  • Had face tracking software totally fail because it wasn’t programmed to register your traits (standard human features such as eyes, a nose, a mouth)?
  • Had voice assistants / voice recognition not understand you due to your accent or dialect? Perhaps the voice assistant straight up doesn’t speak your native language.

Consider: Her and Ex Machina, two recent and popular representations of AI in cinema, both of which represent AI, and its characters’ interactions with AI, from the point of view of male psychology and desire.

As Gates points out:
“If we don’t get women and people of color at the table — real technologists doing the real work — we will bias systems. Trying to reverse that a decade or two from now will be so much more difficult, if not close to impossible.”

The entire interview is worth a read

Together, Gates and Li are launching a national non-profit called AI4ALL, aimed at increasing the diversity of voices behind AI, and getting people of color and women educated in a field where they are highly underrepresented.

Why it’s hot:
AI has the potential to redefine our future. Where is the diversity of minds necessary to make it a future for ALL?

Samsung’s app for users who can no longer read, write or speak

Aphasia is a disorder usually caused by brain injury or disease whose effects can include losing the ability to read, write and speak. Aphasic patients can still comprehend visual cues like symbols, gestures and facial expressions, and written language aside, are able to fully function and communicate.

By mapping series of emojis to everyday human actions and needs, Samsung’s Wemogee app provides a visual interface for aphasic patients to communicate with friends and family. Aphasic users enter emojis into Wemogee, the app translates the visual entries into sentences for their friends and family, and vice versa, helping bridge their gaps in expression and communication.

It is still all too common for websites and apps to consider accessibility and special needs as an afterthought. It’s especially refreshing to see an app created to help a specific group of humans who would normally be ignored in these spaces.

Original article

Computers that recognize hate speech

“Based on text posted on forums and social media, a new machine learning method has been developed to detect antisocial behaviours such as hate speech or indications of violence with high accuracy.”

Link to article

What this can be used for:

  • Identifying clusters and patterns of hateful speech on social media platforms
  • Preventing hate crimes (“In extreme cases, perpetrators of school shootings or other acts of terror post angry or boastful messages to niche forums before they act.”)
  • Big brother

On a related note, remember this:
Twitter taught Microsoft’s AI chatbot to be a racist asshole in less than a day

“Smart” Cities can be hacked

Last Friday at 11:40PM, Dallas’s city hurricane warning system sounded: 156 emergency sirens all at once. The alarms a total of 15 times, with each burst lasting 90 seconds, until the alarms fell silent around 1:20AM on Saturday morning.

“But as the New York Times reports, there was no hurricane coming—the sounds were triggered by a hacker who’d penetrated the system’s security measures. Few details have emerged about the hack, save for the fact that it’s thought to have been carried out locally and was very effective (technicians couldn’t stop the hacker, so they had to shut down the entire system to quiet the alarms).”

As cities and government entities rapidly adopt technologies and networks into day-to-day life and infrastructures, have they overlooked the potential shortfalls of a ubiquitous digital infrastructure?

In the same way that new buildings must meet baseline architectural requirements, perhaps the same minimums should be demanded of tech- and cyber- security.

Source: https://www.technologyreview.com/s/604124/smart-cities-could-be-crippled-by-dumb-security/

Voluntary employee microchipping (parties)

And X-Ray of an Epicenter employee’s microchip implants reveals its location between the index finger and the thumb. Image Source: Pinterest

Employees and renters at Epicenter, a Stockholm-based co-working space have been opting-in to microchip implants that allow them to open office doors, operate printers, or put a smoothie purchase on their company tab with a wave of their augmented hands.

The chips are injected between volunteers’ thumb and index fingers and use Near Field Communication (NFC), the same tech used for mobile payments such as Apple Pay (and yes, yours pets’ microchips).

150 of the company’s 2000 workers have received the implants, and they have become so popular that workers at Epicenter hold parties for those willing to get implanted.

The technology isn’t new, but its implementation – within the human body, raises privacy concerns.

From the AP article:
“While biologically safe, the data generated by the chips can show how often an employee comes to work or what they buy. Unlike company swipe cards or smartphones, which can generate the same data, a person cannot easily separate themselves from the chip.”

With Elon Musk’s recent announcement of Neuralink, his latest enterprise focused on brain-computer interfaces, one starts to think:

  • When it comes to adoption of new tech, consumers continue to show a willingness to trade Privacy for Convenience.
  • Could these body-mod developments signal the first wave of consumer-level bionics, heralding a realization of the cyborg future envisioned and popularized by sci-fi writers of the 70s and 80s?