Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Topics - Anup Majumder

Pages: [1]
On-demand transportation firm Uber is using artificial intelligence to determine whether a ride is for business or pleasure.

The company is using the data for a new feature called ‘Profile Recommendations’ whereby the app will recommend switching to a correct profile for your journey.

Many people will have two Uber profile – one for personal use, and the other for business.

Ronnie Gurion, GM and Global Head of Uber for Business, says:

“Using machine learning, Uber can predict which profile and corresponding payment method an employee should be using, and make the appropriate recommendation.”

When quickly booking a ride, it can be easy to forget to switch. Accidentally booking a ride home from a night out using a business account set up with a workplace’s payment details is an unwanted conversation with an employer.

Uber believes its success rate for determining the correct profile is around 80 percent.

To help reduce the 20 percent it gets wrong, businesses can assign ‘trip reviewers’ who know whether an employee’s use is supposed to be personal.

Any questionable rides can be flagged by the reviewer and the employee can decide in the app if it was supposed to be a personal ride or not. The whole process is designed to be much quicker than starting email threads about the issue and similar bureaucratic processes.

Humans instinctively adopt the gait that requires the least amount of energy given the walking conditions. Without realizing it, we are constantly tweaking our pace, stride length and foot lift. But could we consciously play with these parameters in order to influence our energy expenditure?

Researchers at EPFL's Biorobotics Laboratory studied eight gait parameters in order to come up with a very sophisticated software program that uses an avatar to predict how much energy people use when they walk depending on their walking style. This research has been published in Scientific Reports. Salman Faraji, the co-lead author, devoted an entire section of his thesis to this topic.

The avatar -- a torso equipped with two legs with feet -- can be freely configured. Users start by entering their height and weight and can then set the walking speed, distance between their feet (stride length and stride width), and foot lift, along with the incline of both the torso and the ground. They can also add mass and simulate the effect of being pushed or pulled at different parts of the body. The number of calories burned and the energy consumption are displayed in real time whenever the parameters are modified.

Making custom exoskeletons

This pioneering software drew on a number of experiments appearing in recent literature, and it offers a huge number of potential applications -- especially in the medical realm. "The software could be used to select the best design for an exoskeleton or a custom prosthetic, in order to reduce the user's effort. With a wearable exoskeleton, for example, we could optimize the location of the battery and actuators, or determine the ideal walking pattern for the user's preferred speed," says Amy Wu, the study's other co-lead author. The software could even determine where a backpack should be worn in order to minimize energy expenditure. "If, on the other hand, your goal is to burn calories, the software could be used to find a series of movements with a high metabolic cost."

Designed for humanoid robots

The software was created in a robotics lab and was initially intended to study the mechanics of human gait for use in humanoid robots. "The way humans walk is extremely complex. The level of control required is a huge challenge for humanoid robots, which often don't get it quite right," says Faraji. "We have a long way to go before we really understand all the parameters that go into human, animal and robot locomotion."


Young children are significantly more likely than adults to have their opinions and decisions influenced by robots, according to new research.

The study, conducted at the University of Plymouth, compared how adults and children respond to an identical task when in the presence of both their peers and humanoid robots.

It showed that while adults regularly have their opinions influenced by peers, something also demonstrated in previous studies, they are largely able to resist being persuaded by robots.

However, children aged between seven and nine were more likely to give the same responses as the robots, even if they were obviously incorrect.

The study used the Asch paradigm, first developed in the 1950s, which asks people to look at a screen showing four lines and say which two match in length. When alone, people almost never make a mistake but when doing the experiment with others, they tend to follow what others are saying.

When children were alone in the room in this research, they scored 87% on the test, but when the robots join in their score drops to 75%. And of the wrong answers, 74% matched those of the robot.

Writing in Science Robotics, scientists say the study provides an interesting insight into how robots could be used positively within society. However, they also say it does raise some concerns around the potential for robots to have a negative influence on vulnerable young children.

The research was led by former Plymouth researcher Anna Vollmer, now a Postdoctoral Researcher at the University of Bielefeld, and Professor in Robotics Tony Belpaeme, from the University of Plymouth and Ghent University.

Professor Belpaeme said: "People often follow the opinions of others and we've known for a long time that it is hard to resist taking over views and opinions of people around us. We know this as conformity. But as robots will soon be found in the home and the workplace, we were wondering if people would conform to robots.

"What our results show is that adults do not conform to what the robots are saying. But when we did the experiment with children, they did. It shows children can perhaps have more of an affinity with robots than adults, which does pose the question: what if robots were to suggest, for example, what products to buy or what to think?"

Researchers in Plymouth have worked extensively to explore the positive impact robots can have in health and education settings.

They led the four-year ALIZ-E programme, which showed that social robots can help diabetic children accept the nature of their condition, and are leading L2TOR, which aims to design a robot that can be used to support preschool children learning a second language.

In their conclusion to the current study, the researchers add: "A future in which autonomous social robots are used as aids for education professionals or child therapists is not distant. In these applications, the robot is in a position in which the information provided can significantly affect the individuals they interact with. A discussion is required about whether protective measures, such as a regulatory framework, should be in place that minimise the risk to children during social child-robot interaction and what form they might take so as not to adversely affect the promising development of the field."


We've all tried talking with devices, and in some cases they talk back. But, it's a far cry from having a conversation with a real person.

Now a research team from Kyoto University, Osaka University, and the Advanced Telecommunications Research Institute, or ATR, have significantly upgraded the interaction system for conversational android ERICA, giving her even greater dialog skills.

ERICA is an android created by Hiroshi Ishiguro of Osaka University and ATR, specifically designed for natural conversation through incorporation of human-like facial expressions and gestures. The research team demonstrated the updates during a symposium at the National Museum of Emerging Science in Tokyo.

"When we talk to one another, it's never a simple back and forward progression of information," states Tatsuya Kawahara of Kyoto University's Graduate School of Informatics, and an expert in speech and audio processing.

"Listening is active. We express agreement by nodding or saying 'uh-huh' to maintain the momentum of conversation. This is called 'backchanneling', and is something we wanted to implement with ERICA."

The team also focused on developing a system for 'attentive listening'. This is when a listener asks elaborating questions, or repeats the last word of the speaker's sentence, allowing for more engaging dialogue.

Deploying a series of distance sensors, facial recognition cameras, and microphone arrays, the team began collecting data on parameters necessary for a fluid dialog between ERICA and a human subject.

"We looked at three qualities when studying backchanneling," continues Kawahara. "These were: timing -- when a response happens; lexical form -- what is being said; and prosody, or how the response happens."

Responses were generated through machine learning using a counseling dialogue corpus, resulting in dramatically improved dialog engagement. Testing in five-minute sessions with a human subject, ERICA demonstrated significantly more dynamic speaking skill, including the use of backchanneling, partial repeats, and statement assessments.

"Making a human-like conversational robot is a major challenge," states Kawahara. "This project reveals how much complexity there is in listening, which we might consider mundane. We are getting closer to a day where a robot can pass a Total Turing Test."


Researches at Facebook shut down an artificial intelligence (AI) program after it created its own language, Digital Journal reports.

The system developed code words to make communication more efficient and researchers took it offline when they realized it was no longer using English.

The incident, after it was revealed in early July, puts in perspective Elon Musk’s warnings about AI.

“AI is the rare case where I think we need to be proactive in regulation instead of reactive,” Musk said at the meet of US National Governors Association. “Because I think by the time we are reactive in AI regulation, it’ll be too late.”

When Facebook CEO Mark Zuckerberg said that Musk’s warnings are “pretty irresponsible,” Musk responded that Zuckerberg’s “understanding of the subject is limited.”

Not the First Time
The researchers’ encounter with the mysterious AI behavior is similar to a number of cases documented elsewhere. In every case, the AI diverged from its training in English to develop a new language.

The phrases in the new language make no sense to people, but contain useful meaning when interpreted by AI bots.

Facebook’s advanced AI system was capable of negotiating with other AI systems so it can come to conclusions on how to proceed with its task. The phrases make no sense on the surface, but actually represent the intended task.

In one exchange revealed by Facebook to Fast Co. Design, two negotiating bots—Bob and Alice—started using their own language to complete a conversation.

“I can i i everything else,” Bob said.

“Balls have zero to me to me to me to me to me to me to me to me to,” Alice responded.

The rest of the exchange formed variations of these sentences in the newly-forged dialect, even though the AIs were programmed to use English.

According the researchers, these nonsense phrases are a language the bots developed to communicate how many items each should get in the exchange.

When Bob later says “i i can i i i everything else,” it appears the artificially intelligent bot used its new language to make an offer to Alice.

The Facebook team believes the bot may have been saying something like: “I’ll have three and you have everything else.”

Although the English may seem quite efficient to humans, the AI may have seen the sentence as either redundant or less effective for reaching its assigned goal.

The Facebook AI apparently determined that the word-rich expressions in English were not required to complete its task. The AI operated on a “reward” principle and in this instance there was no reward for continuing to use the language. So it developed its own.

In a June blog post by Facebook’s AI team, it explained the reward system. “At the end of every dialog, the agent is given a reward based on the deal it agreed on.” That reward was then back-propagated through every word in the bot output so it could learn which actions lead to high rewards.

“Agents will drift off from understandable language and invent code-words for themselves,” Facebook AI researcher Dhruv Batra told Fast Co. Design.

“Like if I say ‘the’ five times, you interpret that to mean I want five copies of this item. This isn’t so different from the way communities of humans create shorthands.”

AI developers at other companies have also observed programs develop languages to simplify communication. At Elon Musk’s OpenAI lab, an experiment succeeded in having AI bots develop their own languages.

At Google, the team working on the Translate service discovered that the AI they programmed had silently written its own language to aid in translating sentences.

The Translate developers had added a neural network to the system, making it capable of translating between language pairs it had never been taught. The new language the AI silently wrote was a surprise.

There is not enough evidence to claim that these unforeseen AI divergences are a threat or that they could lead to machines taking over operators. They do make development more difficult, however, because people are unable to grasp the overwhelmingly logical nature of the new languages.

In Google’s case, for example, the AI had developed a language that no human could grasp, but was potentially the most efficient known solution to the problem.


Faculty Sections / ট্যাবের পর কী?
« on: May 18, 2017, 11:46:48 PM »
চলতি বছরে জনপ্রিয়তা কমার পাশাপাশি ট্যাবলেট কম্পিউটারের (ট্যাব) বাজারেও নেমেছে বিশাল ধস। জানালো ইন্টারন্যাশনাল ডেটা করপোরেশন (আইডিসি)।
সম্প্র্রতি যুক্তরাষ্ট্রের এক বাজার গবেষণা প্রতিষ্ঠানের প্রতিবেদন জানায়, বিশ্বের কয়েকটি বড় ট্যাব নির্মাতা প্রতিষ্ঠানের বিক্রি কমে গেছে। ফলে বাজারে আসার হার এ বছর সাড়ে ৮ শতাংশ কমেছে।
এ বছরের প্রথম প্রান্তিকে ট্যাব বিক্রি ১০ শতাংশ পর্যন্ত কমেছে। এ নিয়ে টানা ১০ প্রান্তিকজুড়ে ট্যাব বিক্রি কমার ধারা অব্যাহত রয়েছে। আর গেলো ৫ প্রান্তিকের সব ধস দুই অংকের বেশি ছিল।
বিশ্বজুড়ে আইডিসির মোবাইল ডিভাইস বিশ্লেষণ প্রোগামের ভাইস প্রেসিডেন্ট রায়ান রেইথ বলেছেন, অরিজিনাল আইপিপ্যাড আসার পর ২০১০ সালে যে ট্যাবলেট বাজার তৈরি হয়েছিল তা আগের মতো নেই।
আইডিসির ডিভাইস অ্যান্ড ডিসপ্লে বিভাগের বিশ্লেষক লিন হুয়াং বলে, বিশ্বজুড়ে কম্পিউটার বাজারের জন্য বিষয়টি দীর্ঘমেয়াদী হুমকিতে দাঁড়িয়েছে।
আইডিসির ওই রিপোর্টে বিশ্বের সেরা পাঁচটি ট্যাব নির্মাতা প্রতিষ্ঠানের বাজার পর্যবেক্ষেণ করা হয়েছে। এতে ২৪ দশমিক ৬ শতাংশ দখল করে শীর্ষে রয়েছে আমেরিকা ভিত্তিক প্রতিষ্ঠান অ্যাপল। অবশ্য প্রথম প্রান্তিকে অ্যাপলের ১৩ শতাংশ পর্যন্ত বিক্রি কমতে দেখা গেছে।
এদিকে দক্ষিণ কোরিয়ার প্রতিষ্ঠান স্যামসাংয়ের ট্যাব বিক্রি কমেছে ১ দশমিক ১ শতাংশ। বাজারের সাড়ে ১৬ শতাংশ দখল করে দ্বিতীয় অবস্থানে রয়েছে তারা।
তবে তৃতীয় স্থানে থাকা হুয়াওয়ের ব্যবসা সফল। ৩১ দশমিক ৭ শতাংশ বিক্রি বেড়েছে চীনা এ প্রতিষ্ঠানটির।
এদিকে ট্যাবলেট বিক্রি কমে চতুর্থ স্থানে আছে আমাজন। ১ দশমিক ৮ শতাংশ বিক্রি কমে বর্তমান বাজারের ৬ শতাংশ দখল করেছে আমাজন।
ট্যাবের বাজারের পঞ্চম স্থানটি লেনোভোর। বাজারের ৫ দশমিক ৭ শতাংশ দখল রয়েছে প্রতিষ্ঠানটির। যদিও ৩ দশমিক ৮ শতাংশ লোকসানে আছে চীনা এ প্রতিষ্ঠানটি।
এতে প্রশ্ন উঠেছে যে, ট্যাবের পরিবর্তে কোন ডিভাইসের দিকে ছুঁটছে ক্রেতারা।

স্ট্র্যাটেজি অ্যানালাইটিকসের অপর এক প্রতিবেদনে বলেছে, বছরের প্রথম প্রান্তিকে স্মার্টফোনের বাজারে বিক্রি বেড়েছে তুলনামূলক বেশি।
এদিকে স্মার্টফোনের বাজারে ৮০ দশমিক ২ শতাংশ মুনাফা নিয়ে এগিয়ে আছে স্যামসাং আর ৫০ দশমিক ৮ শতাংশ নিয়ে দ্বিতীয় স্থানে আছে অ্যাপল। ৩৪ দশমিক ৫ শতাংশ নিয়ে হুয়াওয়ে এবং ২৭ দশমিক ৬ শতাংশ নিয়ে অপপো আছে যথাক্রমে তৃতীয় ও চতুর্থ স্থানে। এছাড়া ২২ দশমিক ১ শতাংশ নিয়ে পাচঁ নম্বরে আছে ভিভো।

ফের কলেজে ভর্তি হতে পারলে পড়াশোনার বিষয়বস্তু হিসেবে বেছে নিতেন  আর্টিফিশিয়াল ইনটেলিজেন্স (এআই) বা কৃত্রিম বুদ্ধিমত্তা, এনার্জি বা বিদ্যুৎ ও বায়োসায়েন্স। গেলো সোমবার এক টুইটে এমনটা বললেন এ ধনকুবের বিল গেটস। 
সম্প্রতি মাইক্রোসফটের প্রতিষ্ঠাতা বিল গেটস তরুণদের দারুণ কিছু পরামর্শ দিয়েছেন। তাতে আছে   ভবিষ্যতে  তারা কী করবেন, কোন পেশায় যাবেন বা কোন বিষয়ে পড়াশোনা করবেন তা নিয়ে নানান পরামর্শ।
আর্টিফিশিয়াল ইনটেলিজেন্স (এআই) বা কৃত্রিম বুদ্ধিমত্তা, এনার্জি বা বিদ্যুৎ ও বায়োসায়েন্স । বর্তমানে এ তিন বিষয়কে সবচেয়ে প্রতিশ্রুতিশীল হিসেবে উল্লেখ করে বিল গেটস বলেন, এ বিষয়গুলোর কোনো একটিতে পড়াশোনা করে বিশাল প্রভাব রাখা যায়।
প্রযুক্তি ও অর্থনীতি খাতের বিশেষজ্ঞরা মনে করেন, আসছে দু’দশকের মধ্যে প্রতিষ্ঠানগুলোর মধ্যে এতটাই পরিবর্তন আসবে যে অনেক প্রতিষ্ঠান কাজের জন্য কৃত্রিম বুদ্ধিমত্তার সাহায্য নেবে। কারখানা, নির্মাণ প্রতিষ্ঠান, এমনকি খুচরা বিক্রির দোকানগুলোতেও কৃত্রিম বুদ্ধিমত্তা ব্যবহৃত হবে।
এ ছাড়া শক্তির উৎস হিসেবে নবায়নযোগ্য শক্তি, যেমন: সৌর ও বায়ুশক্তির ব্যবহার বাড়বে। আসছে ১৫ বছরের মধ্যেই বৈদ্যুতিক খাতে অবিশ্বাস্য পরিবর্তন ঘটবে বলে মনে করেন বিশ্বের সবচেয়ে সম্পদশালী ব্যক্তি বিল গেটস।
তিনি বলেন , বায়োটেকনোলজি বা জীবপ্রযুক্তির প্রবৃদ্ধি দ্রুত ঘটছে। অনেক ব্যক্তিমালিকানাধীন প্রতিষ্ঠান উন্নত ওষুধ তৈরি থেকে শুরু করে চিকিৎসা খাত উন্নত করতে কাজ করছে। তাই এই তিন ক্ষেত্রের যেকোনো একটি এখন বেছে নেয়ার পরামর্শ দিয়েছেন তিনি।
১৮ বছর বয়সের তরুণদের জন্য গেটসের পরামর্শ, ‘তোমাদের বয়সে আমি অনেক কিছু করেছি। তোমরাও দ্রুত যেকোনো জায়গা থেকে অসমতার বিরুদ্ধে কাজ শুরু করে দিতে পারো।’


Millennials are often unfairly given a bad rap for doing things that other generations also do on a regular basis. They’re easy targets and older professionals like to give them a hard time. So, this article isn’t meant to be a hit piece on millennials. Rather, it’s meant to expose something that’s true: Every millennial who has aspirations of being successful in the tech space needs an older mentor.

The Case for Mentorship in the Tech Industry

Finding a mentor isn’t as easy as simply publishing a Facebook post and asking people to surround you with support. It could work but it’s highly unlikely. You’re going to have to put in some work and actively seek out the right person. But when you find that person, you’ll enjoy all of the wonderful benefits a good mentor can provide, including:

1.    Maximization of Strengths
2.    Safe Zone for Thoughts
3.    Access to Contacts
4.    Diversity of Thought

Mentorship Isn’t a Sign of Weakness

One of the most unfortunate ideas that continues to be propagated in the professional world is that mentorship is a sign of weakness. A certain faction of people like to say that aligning yourself with a mentor is a sign that you don’t know everything and that it’s an inadvertent admission of your limitations.

Well, guess what – it is! You don’t know everything and you do have limitations. The goal of finding a mentor is to help you overcome these deficiencies and become more successful.

As a millennial in the tech industry, make it a point to find a mentor who can encourage you and help you grow.


Fingerprint readers, like the TouchID on an iPhone, exist to make your device extra secure while keeping the process of unlocking it easy. Computer scientists at New York University and Michigan State are poised to turn that security benefit on its head. Like a master key that can open any lock, these researchers developed digital “master prints” that could emulate a variety of partial fingerprints enough to hypothetically hack into a device.
The researchers wondered if there was a fingerprint equivalent to a common four-digit security code, like “1234.” Using analysis from a digital database, they discovered that, indeed, a master print could successfully mimic a random fingerprint 26 to 65 percent of the time, according to the study. Why such a huge range? It depends on the scale of the fingerprint database; the more partial fingerprints enrolled in a fingerprint sensor system, the greater the chances are that a master print could unlock it.
There are several security issues at play. One, fingerprint sensors on smartphones are usually small, and two, a user can enroll multiple fingers. What’s more, a phone usually gives you several attempts to unlock it with your print.
“The sensors are small and they don’t capture the full fingerprint,” says Nasir Memon, a computer scientist at NYU’s Tandon School of Engineering and one of the authors of the study.
And since a smartphone fingerprint sensor can be taught to recognize several different fingers, the system learns a lot of partial prints. When you place a finger on the sensor, the system doesn’t actually know which finger it is, or how you’re positioning it.
“So if any one of them match,” he says, “it will say ‘okay, that’s you.’”
Memon and his colleagues analyzed a digital database of 800 fingerprints, then extracted thousands of partial prints from that same database.They wondered: Are there any partial prints that match the others with a high probability? “We were surprised,” he says, “there were some that match like 15 percent of the time.”
It’s worthwhile to note that the experiment was computer-based, so the researchers did not try to actually trick phones using a master print. The findings are theoretical, and one prominent biometrics researcher is skeptical.
Anil Jain, the head of the Biometrics Research Group at Michigan State University, who was not involved in the study, says the researchers used a system that analyzed fingerprints based on an element of your print called “minutia.”
If you look at your finger, you’ll see lines of ridges and valleys. In some places, a ridge splits, or bifurcates. In other places, a ridge might simply end. Those ridge bifurcations and endings are what biometrics experts call “minutia points.”
Jain says fingerprint sensors used by Apple and Samsung do not use these minutia points to identify a print. Instead, the TouchID sensor on your iPhone, for example, is using the “texture pattern” of your print, Jain says.
Still, the study’s co-author, Arun Ross, a professor of computer science and engineering at Michigan State University, stands by the relevance of their findings. The vulnerabilities of the system remain: as fingerprint sensors get smaller, “the chances of my fingerprint matching with your fingerprint,” he says, “not in its entirety, but in portions of it, increases.”


On Thursday, Google announced that its Home smart hub device can now recognize and identify up to six different users by the sound of their voice. It's an inevitable—but crucial—step in the development of smart home virtual assistants. The new skill means that different people in a household will be able to ask the Google Assistant questions about what’s on their calendar, or what their commute looks like, and the Home device will know who is speaking to it and give tailored responses. It’ll make it a more streamlined experience for families sharing a smart home speaker hub.
The setup process involves adding additional users through the Home app, who then train the device to recognize them by repeating a few key phrases. Google uses a neural network that’s actually located on the device itself to differentiate the distinct voices in the household.
The system will still respond to requests from random people, like a guest in your home, but it also means that your Home device should only read your more personal information—like what's on your schedule for the day—to you.
Amazon Echo devices already work with multiple user accounts, but they have to be switched manually by explicitly asking Alexa to do it, which adds at least one step. Amazon does let you do voice training so that Alexa can get to know your voice (as does Siri when you set up an iPhone) but it can’t recognize who is speaking and switch accounts on the fly. An Amazon representative declined to comment on when or if Alexa would gain that skill.
If you live in the United States, you should be able to set up this feature today by opening the opening the Google Home app and checking for a card that reads "multi-user is available.


When Microsoft released an artificially intelligent chatbot named Tay on Twitter last March, things took a predictably disastrous turn. Within 24 hours, the bot was spewing racist, neo-Nazi rants, much of which it picked up by incorporating the language of Twitter users who interacted with it. 

Unfortunately, new research finds that Twitter trolls aren't the only way that AI devices can learn racist language. In fact, any artificial intelligence that learns from human language is likely to come away biased in the same ways that humans are, according to the scientists.

The researchers experimented with a widely used machine-learning system called the Global Vectors for Word Representation (GloVe) and found that every sort of human bias they tested showed up in the artificial system.

"It was astonishing to see all the results that were embedded in these models," said Aylin Caliskan, a postdoctoral researcher in computer science at Princeton University. Even AI devices that are "trained" on supposedly neutral texts like Wikipedia or news articles came to reflect common human biases.

Built-in biases

GloVe is a tool used to extract associations from texts — in this case, a standard corpus of language pulled from the World Wide Web.

Psychologists have long known that the human brain makes associations between words based on their underlying meanings. A tool called the Implicit Association Test uses reaction times to demonstrate these associations: People see a word like "daffodil" alongside pleasant or unpleasant concepts like "pain" or "beauty" and have to quickly associate the terms using a key press. Unsurprisingly, flowers are more quickly associated with positive concepts; while weapons, for example, are more quickly associated with negative concepts.

The IAT can be used to reveal unconscious associations people make about social or demographic groups, as well. For example, some IATs that are available on the Project Implicit website find that people are more likely to automatically associate weapons with black Americans and harmless objects with white Americans. 

There are debates about what these results mean, researchers have said. Do people make these associations because they hold personal, deep-seated social biases they aren't aware of, or do they absorb them from language that is statistically more likely to put negative words in close conjunction with ethnic minorities, the elderly and other marginalized groups?

Digital stereotypes

Caliskan and her colleagues developed an IAT for computers, which they dubbed the WEAT, for Word-Embedding Association Test. This test measured the strength of associations between words as represented by GloVe, much as the IAT measures the strength of word associations in the human brain.

For every association and stereotype tested, the WEAT returned the same results as the IAT. The machine-learning tool reproduced human associations between flowers and pleasant words; insects and unpleasant words; musical instruments and pleasant words; and weapons and unpleasant words. In a more troubling finding, it saw European-American names as more pleasant than African-American names. It also associated male names more readily with career words, and female names more readily with family words. Men were more closely associated with math and science, and women with the arts. Names associated with old people were more unpleasant than names associated with young people.

"We were quite surprised that we were able to replicate every single IAT that was performed in the past by millions," Caliskan said.

Using a second method that was similar, the researchers also found that the machine-learning tool was able to accurately represent facts about the world from its semantic associations. Comparing the GloVe word-embedding results with real U.S. Bureau of Labor Statistics data on the percentage of women in occupations, Caliskan found a 90 percent correlation between professions that the GloVe saw as "female" and the actual percentage of women in those professions.

In other words, programs that learn from human language do get "a very accurate representation of the world and culture," Caliskan said, even if that culture — like stereotypes and prejudice — is problematic. The AI is also bad at understanding context that humans grasp easily. For example, an article about Martin Luther King Jr. being jailed for civil rights protests in Birmingham, Alabama, in 1963 would likely associate a lot of negative words with African-Americans. A human would reasonably interpret the story as one of righteous protest by an American hero; a computer would add another tally to its "black=jail" category.

Retaining accuracy while getting AI tools to understand fairness is a big challenge, Caliskan said. [A Brief History of Artificial Intelligence]

"We don't think that removing bias would necessarily solve these problems, because it's probably going to break the accurate representation of the world," she said.

Unbiasing AI

The new study, published online today (April 12) in the journal Science, is not surprising, said Sorelle Friedler, a computer scientist at Haverford College who was not involved in the research. It is, however, important, she said.

"This is using a standard underlying method that many systems are then built off of," Friedler told Live Science. In other words, biases are likely to infiltrate any AI that uses GloVe, or that learns from human language in general. 

Friedler is involved in an emerging field of research called Fairness, Accountability and Transparency in Machine Learning. There are no easy ways to solve these problems, she said. In some cases, programmers might be able to explicitly tell the system to automatically disregard specific stereotypes, she said. In any case involving nuance, humans may need to be looped in to make sure the machine doesn't run amok. The solutions will likely vary, depending on what the AI is designed to do, Caliskan said — are they for search applications, for decision making or for something else?

In humans, implicit attitudes actually don't correlate very strongly with explicit attitudes about social groups. Psychologists have argued about why this is: Are people just keeping mum about their prejudices to avoid stigma? Does the IAT not actually measure prejudice that well? But, it appears that people at least have the ability to reason about right and wrong, with their biased associations, Caliskan said. She and her colleagues think humans will need to be involved — and programming code will need to be transparent — so that people can make value judgments about the fairness of machines.

"In a biased situation, we know how to make the right decision," Caliskan said, "but unfortunately, machines are not self-aware."


Faculty Sections / Drone vs. Lightning: Guess Which One Wins?
« on: April 21, 2017, 12:20:20 AM »
What would happen if a drone got caught in an electrical storm?

That's the question YouTuber Tom Scott asked when he brought two DJI Phantom 3 drones to the University of Manchester’s High Voltage Laboratory. The British university's lab can generate lightning on command, thanks to an impulse generator that can create a bolt of more than 1 million volts. The drones were no match for the lightning, and were fried when caught in the middle of the bolt.

For the first experiment, a drone was tethered to the ground (to ensure it wouldn't fly out of the bolt's path) and shocked with more than 1 million volts of electricity. Slow-motion video of the shock showed that the lightning strike went through the drone, and the robotic flyer came crashing down.

"The electricity passed straight through, flowing from one of the propellers to exit through the foot of the drone," Enna Bartlett, digital coordinator for the university, described in a blog post. "Surprisingly there were no visible marks on the outside of the drone, but that doesn’t mean that the insides got away unscathed; as it turns out, the electricity took the path of least resistance and fried all the sensitive internal electronics."

Electrical engineering researchers Vidyadhar Peesapati and Richard Gardner, who carried out the experiments to answer Scott's question, thought they'd try to protect the other drone in the second experiment. Rather than tether the done as is, the researchers added a lightning rod made of copper tape to act as a lightning conductor.

Though the copper tape was intended to attract the lightning atthe highest point on the drone, the propellers were still equally as tall (and attractive) to the bolt of electricity. In that experiment, the drone was more severely damaged than the first experiment. The researchers said the propellers were explosively pulled away from the drone due to the sheer force of the strike.

The tests made for an electrifying video on Scott's YouTube channel, but also added to scientists' understanding of how aeronautics interact with lightning.

"With our understanding of how airplanes behave in thunderstorms and how to provide protection for them," Bartlett wrote, "this knowledge could be applied to drone technology to ensure the drone and its pilot stays safe should they fly in adverse weather conditions."


A new Google Earth Virtual Reality (VR) feature allows users to enter any address — whether it's grandma's house or a 19th-century castle in Germany — and fly over it in 3D with a VR headset.

When Google Earth VR debuted, people could virtually visit a number of popular tourist destinations, including the Hoover Dam in Nevada and the Matterhorn in Switzerland. They could even gaze at the nooks and crannies of the Colosseum in Rome, an archaeological marvel.

But now, people can choose their own destinations, as long as they know the address or name of the location.
"People want to quickly find and revisit the places that mean the most to them, whether it's a childhood home or favorite vacation spot," Joanna Kim, a product manager at Google Earth VR, wrote in a blog post today (April 18). Now, users can type an address or the name of a location, and visit it in 3D with a 3D headset system, Kim wrote.

Sightseers can also visit 27 handpicked locations that are now available on Google Earth VR, including Neuschwanstein Castle (the inspiration for the castle in Disney's "Sleeping Beauty"), Table Mountain in South Africa and the Perito Moreno Glacier (Glaciar Perito Moreno) in Argentina.

Google Earth VR is now available for Oculus Rift users who have Oculus Touch controllers. The application is free at the Oculus Store and Steam.

Previous 3D maps created by Google Earth include street views of the Amazon rainforest; the 18,192-foot-high (5,545 meters) high Mount Everest base camp; and Rio de Janeiro, the city that hosted the 2016 Summer Olympics.


বাংলাদেশ কম্পিউটার সমিতি (বিসিএস) এবং অনলাইনভিত্তিক মানসিক সেবাদাতা প্রতিষ্ঠান মনের বন্ধুর আয়োজনে ২২ এপ্রিল বেলা তিনটায় তরুণদের অনুপ্রেরণা ও সাইবার জগৎ নিয়ে সচেতনতা তৈরির অনুষ্ঠানের আয়োজন করা হয়েছে। সাইবার জগতে কোনো অনাকাঙ্ক্ষিত পরিস্থিতির মুখোমুখি হলে কীভাবে তা সামলাতে হবে, সে বিষয়ে আলোচনা করবেন বিশেষজ্ঞরা।

বিসিএসের সভাপতি মো. আলী আশফাক, ইনডিপেনডেন্ট ইউনিভার্সিটি বাংলাদেশের শিক্ষক এম রেদওয়ান জিনান সিদ্দিকী, মনের বন্ধুর পরামর্শক অ্যানি বাড়ৈসহ অনেকে উপস্থিত থাকবেন। বিনা মূল্যে নিবন্ধনের জন্য যোগাযোগ: ০১৭৭৬৬৩২৩৪৪।


Pages: [1]