Connect with us

Tech News

What can we do about bias in AI?

(TECH NEWS) Each AI system is a reflection of their interaction with the humans who designed, programmed, trained, or used them – and they are reflecting out own deep-rooted biases.

Published

on

ai

Skewed samples

The machines we make reflect our own biases, according to Kristian Hammond, who wrote an article breaking down the different ways an artificially intelligent system can be biased. Each system is a reflection of their interaction with the humans who designed, programmed, trained, or used them.

bar
Data-driven bias results when a system learns from a skewed sample. Often we think this won’t be a problem because of the sheer volume of examples given to facilitate machine learning, but that isn’t the case. In fact a viral video demonstrated that HP motion trackers don’t track faces with non-white skin tones, which might be just such a problem.

Your system reflects your bubble

Bias through interaction comes when smart systems learn from interacting with humans. Never has a more cautionary tale of this kind of bias come as quickly as the day-long life of Microsoft’s Tay. Tay was a twitter chat bot designed to learn based on its interactions with human tweeters. As anyone who has ever spent one day in junior high could have predicted, human users bombarded Tay with offensive statements. With this as a model, the chat bot became an aggressive racist and misogynist and had to be shut down.

Emergent bias and similarity are a little more complicated, but have to do with systems aimed at personalization.

Similar to the problem of social news bubbles, systems trained to show you what you want to see will become more biased the longer they do that thing.

Systems can also have conflicting goals, where they are told to perform a specific purpose, and the interactions with users push them towards a different one.

Identification is key

Hammond notes we view AI and machines as cold and indiscriminating. Whether or not we think this is a design flaw or a great accomplishment, it it reinforces a misconception that the smart machines we make are objective.

After looking specifically at each way things can go terribly wrong, I have to agree with Hammond: we need to identify our own biases. At every level, from engineering to product use, we need to think of how we are making, training, and using these systems in order to prevent our own flaws from getting a Terminator-style upgrade.

#AIBias

Felix is a writer, online-dating consultant, professor, and BBQ enthusiast. She lives in Austin with two warrior-princess-ninja-superheros and some other wild animals. You can read more of her musings, emo poetry, and weird fiction on her website.

Tech News

AI’s reading abilities catch up with humans – scary or awesome?

(TECH NEWS) Measured by Stanford University’s reading comprehension test, AI achieves a major milestone and some people fear being replaced, others admire the innovation.

Published

on

ai robot not human

Going into 2018, I decided not so much to set major resolutions off the bat, but rather set mini goals along the way. One of those mini goals started in November, when I challenged myself to watch 100 new (to me) movies in 365 days.

A few days ago, I hit number 35 which was the award-winning Joaquin Phoenix film, “Her.” Without giving too much away, the plot is set in the near future where a man develops a relationship with a life-like operating system (OS).

During one of their first conversations, the OS states that she read an entire book in under a second. While watching it, I couldn’t help but think that we’re not too far away from that.

Turns out my thought wasn’t far off, as artificial intelligence (AI) developed by Alibaba and Microsoft have out read humans on a Stanford University reading comprehension test.

“This is the first time that a machine has outperformed humans on such a test,” Alibaba said in a statement.

Experts in the field of artificial intelligence at Stanford developed this test to measure and assess the growing reading abilities of computers. The test creates comprehension questions based off of a selection of Wikipedia articles.

Alibaba scored an 82.44, beating humans by a hair as they scored 82.304. The next day, Microsoft’s AI software also beat humans with the score of 82.650.

While this is a monumental feat for technology, it will pose a problem for humans as more jobs will be at risk. This is a continued issue for humans as robots are designed and trained for their jobs (i.e. Amazon).

The technology has already been put to work for the Singles Days shopping extravaganza, as a large numbers of computers were used for customer service queries. It’s suggested that the technology can be used for other customer service outlets, as well as having the ability to give museum tours.

Brands such as Facebook, Tencent, and Samsung have submitted AI models for the Stanford test in the past. “These kinds of tests are certainly useful benchmarks for how far along the AI journey we may be,” said Andrew Pickup, a spokesman for Microsoft. “However, the real benefit of AI is when it is used in harmony with humans,” he added.

AI continues to grow and shows no sign of stopping, and some humans are concerned what this will mean for the future – some fear being replaced, others laud the ability to innovate our world. What do you think?

Continue Reading

Tech News

Practice for coding interviews, live with another human

(CAREER) Looking to practice coding interviews with a real human? Earn yourself a sweet win-win situation with this online solution to acing technical interviews.

Published

on

technical coding interview

Practicing for interviews with a friend helps prepare you for the real thing. When the time comes, unless y’all seriously slacked off or you got the company totally wrong, you should at least feel comfortable interviewing since you practiced with a pal.

Just about anyone can help you prep for standard interview questions like listing your strengths, weaknesses, and talking about a five-year plan. But when it comes to coding interviews, your friends may not be able to offer as much help if they’re not in the field.

Technical interviews typically include demonstrating not only solid communication, but also CS knowledge, coding, and demonstrable problem solving.

Well, count your blessings and say hello to Pramp, an online platform that helps you practice technical interviews for software engineering roles. Named for the acronym version of “practice makes perfect,” Pramp pairs you with a peer for a live HD video interview practice session.

Personal matching considers experience, education, target companies, practice topics you’d like to cover, preferred programming language, and availability.

You and your peer take turns being the interviewer and interviewee for 30 minutes each. Pramp provides the questions to ask 24 hours prior to the interview, and gives answers as well. There’s even a collaborative real-time code editor to share your work.

Pramp supports twelve programming languages, including C++, Java, and Python. You select multiple programming languages for your interviews.

Once the interviews are over, you provide feedback and rate each other’s performance.

There’s no limit to how many interviews you can participate in, and you can choose your level of anonymity. Pramp will only share what you allow, but if you go the public route, they’ve got a pretty sweet perk…

After a few coding interviews, Pramp assesses your strengths to match you with relevant opportunities. From there, you can schedule a phone screen directly on Pramp with a hiring manager. You don’t even need a resume.

Passing the phone screen gets you an onsite interview with the hiring manager. All that practice sure paid off, huh? You don’t need to install anything to start using Pramp, and mega bonus: it’s is free to use. Check them out now if you’re a prospective employee or employer to get connected with talent.

Continue Reading

Tech News

Upload a pic, this site sizes it for 7 social media sites at once

(SOCIAL MEDIA) Creating the correctly-sized images for each social media platform can take a lot of time and patience. A new app helps to make this process simple.

Published

on

A goal that most all of us have is to always be increasing efficiency and productivity. The more that we complete repeated tasks, the more we are able to cut out unnecessary steps and get to the end result faster.

I have found this to be true in terms of social media management. Each platform has their own rules, their own means of posting, and it takes time and attentiveness to get your message across each individual site.

One of the most time consuming aspects of this is the visual component. It’s not only time consuming, but also crucial to have as part of your post as that is what draws an audience’s attention.

The issue with this is that there are different settings for each platform, forcing us to have specially-sized photos for each social media site. This was a problem I ran into a few months back when I was attempting to create a logo for my company to be used for each platform. What looked good on Facebook, wouldn’t translate to LinkedIn, and so on.

Now, I’ve learned of a one-stop-shop to create sizes for each social media site. This was found in the form of Landscape.

Landscape describes itself as “streamlined image resizing for social media.” The app lets you prepare images that meet the aspects for each social media site.

The free app was created by Sprout Social and it’s easy to use. First, you select an image and upload it to Landscape. Then, you choose the social networks that you want the image to be sized for. Finally, you crop the images to their respective sizes and upload them to your pages.

“Landscape is a powerful image resizing tool designed to help social media marketers, content creators and business owners develop a standout presence in an increasingly visual social world,” says Sprout Social.

“Our tool offers social media professionals an efficient way to produce multiple image sizes optimized for social media profiles, messages and campaigns – ultimately giving them more time to focus on what matters most: fostering engagement and authentic conversations through social.”

It’s as simple as that and helps save you time and aggravation.

Continue Reading
Advertisement

The
American Genius
News neatly in your inbox

Join thousands of AG fans and SUBSCRIBE to get business and tech news updates, breaking stories, and MORE!

Emerging Stories