Issue 206: CognitionX Data Science, AI and Machine Learning Briefing

 


Ikea + AI. Awesome AI tools for personal use. Numerai founder discusses company. Tinder outrage. https://cognitionx.com/news-briefing/

AI is not magically deployed. It takes hard work- from choosing the right product (should we use IBM Watson? Or maybe Synthesys?), to trialling it, to training the AI to do the job right, it is often easier said than done. If you haven’t seen it already, definitely check out the recent NYT article on training AI, which talks about how ordinary employees are teaching the AI to do their job.

If you have any stories, either positive or negative, about training AI, I’d love to hear it. Tweet @Cognition_X to start the discussion.

Best,

Tabitha UntiltheBotsTakeOver Goldstaub

What You Might Have Missed From Last Week

Meet the people who train the robots (to do their own jobs)

Daisuke Wakabayashi, from the New York Times, put out a great piece on the oft-neglected side of AI deployment: training it to do your job. They spoke with five people — a travel agent, a robotics expert, an engineer, a customer-service representative and a scriptwriter, of sorts — who have been put in this remarkable position. More than most, they understand the strengths (and weaknesses) of artificial intelligence and how the technology is changing the nature of work.

The trainers had mixed emotions about their tasks, with some being inspired by the task, while others were frustrated. It’s an important read- check it out here.

Business Impact of AI

IKEA dives into world of Artificial Intelligence

It’s known for flat-pack wood, but Ikea’s made no secret of its digital future. The Swedish business was quick to embed wireless charging into its furniture and has already launched its own affordable line of smart lights you can control from your phone.

Now, the company’s Copenhagen-based innovation lab is testing out new digital waters: it’s collecting research on public perception of artificial intelligence (AI). They have put out a survey which asks questions that refer to how AI might find purpose in your home like “should your AI fulfil your needs before you ask?” and “should your AI prevent you from making mistakes?” The lab also asks about contentious issues, such as if and when it’s okay to collect user data, and whether AI should be programmed to mirror your views.

Inspiration

The future of technology and innovation

Richard Muirhead, General Partner of OpenOcean and Co-Founding Chairman of Firestartr, has written a thought-provoking essay on the rise and scalability of technology, for better or for worse.

He also touches upon “the post-truth world and fake news and the inequality and relative stagnation that has stirred the kraken of populism”, but writes that there are also positive repercussions. He says that what we need now is a step towards Yuval Harari’s ‘Homo Deus’: primed for continuous learning, open to internationalisation and geared for innovation.

[OpenOcean held its inaugural Annual CEO Summit on 4 April 2017 in London, you can find the opening slides here]

Products We Love

A list of AI tools you can use today — for personal use

This is Part 1 of 3 in a series of posts that looks at the landscape of the artificial intelligence industry and the companies and institutes developing products that are moving the needle of knowledge of machine intelligence and consciousness forward for humanity.

Liam Hänel, creator of Lyra, trawled through literally thousands of websites (6,000+ links) over a few of weeks to bring you a comprehensive list of the best AI products and most promising companies in the field. He lists AI for personal use, work, social, education, health, agents, entertainment, travel, and more.

Future of Health

Deep learning offers promising results in health

Paul Hsieh (Forbes) is quite impressed by the progress we have been making in health, thanks to deep learning. For example, deep learning algorithms have been able to diagnose the presence or absence of tuberculosis (TB) in chest x-ray images with astonishing accuracy. Similar deep learning algorithms have shown encouraging successes in other branches of medicine such as pathology, ophthalmology (eye diseases), cancer detection, and cardiology.

However, he is most intrigued by the possibility of AIs detecting new associations not yet detected by humans. For instance, UK researchers gave data on 295,000 patients to machine learning algorithms, to allow them to correlate medical history with rates of heart attacks. Then the algorithms were given records from another 82,000 patients and asked to predict which ones would have heart attacks. He writes that “in time, AIs will likely displace many practitioners in many branches of medicine, including my own specialty of radiology. But for all of us, the potential benefits outweigh the short-term costs. I, for one, welcome our future AI medical experts.”

Podcasts We Love

Founder Richard Craib talks about his AI hedge fund Numerai

There’s a new way to make stock market predictions. Numerai, a hedge fund based in San Francisco, has gained a following as the first hedge fund that gives stock market data to machine learning data scientists using structure-preserving encryption to prevent them from mimicking the fund’s trades themselves. Several thousand anonymous data scientists compete to create the best trading algorithms—and win bitcoin for their efforts.

At the same time, the company carefully organizes this
encrypted data in a way that allows the data scientists to build models that are potentially able to make better trades. Founder Richard Craib believes that Numerai can become even more successful if it can align the incentives of everyone involved. His hope is that his new kind of currency, Numeraire, will turn online competition into a collaboration—and turn Wall Street on its head in the process.

Products We’re Not Sure About

Tinder users outraged as machine learning app scrapes and exposes 40,000 profile photos

Your Tinder profile may be getting some unwelcome attention, since a machine learning platform recently uploaded facial data from roughly 40,000 Tinder profiles. The data is available on GitHub and has been uploaded under a CC0: Public Domain License.

The data set “People of Tinder” was released with six downloadable zip file, four of which contained around 10,000 profile images apiece and two files with sample sets of 500 images per gender. Stuart Colianni, the developer of the data set, stated that he released a “simple script to scrape Tinder profile photos for the purpose of creating a facial dataset.” Tinder offers access to thousands of potential profiles and Colianni hopes to use the dataset to create a convolutional neural network that will be able to distinguish between men and women.

Dates for Your Diary

I’ve been making some changes based on Feedback. Would love to hear from more of you. Please do click to share your thoughts!


0
Looking

 

 

 

Published in
0 Comments

Leave a reply

Processing...
Thank you! Your subscription has been confirmed. You'll hear from us soon.
ErrorHere

Log in with your credentials

or    

Forgot your details?

Create Account