Issue 194: CognitionX Data Science, AI and Machine Learning Briefing


Online Customer Experience. Driverless cars. Filtering out unfairness. Payments chatbots.

Just last night, the British Retail Consortium released UK retail sales figures that showed that the rise in inflation is already starting to bite, with non-food high street sales suffering the worst fall in nearly six years.

In this environment, retailers are increasingly looking to AI to better understand and target their customers. For example, using IBM Watson’s expert shopper, North Face can help you find the perfect jacket for your next adventure. On the L’Oreal app, you can upload a selfie to virtually try on makeup. What are the other innovations coming on the horizon? What does this mean for us as customers?

Find out more at our event tomorrow looking at How AI is Transforming the Online Customer Experience or follow us on Twitter to get the highlights.


Tabitha UntiltheBotsTakeOver Goldstaub

Business Impact of AI

Quarter of miles driven in U.S. by 2030 will be driverless

BCG projects that as many as 1.5 trillion km traveled in the U.S. in 2030 will be in shared, self-driving electric cars. The shift will begin gradually in the early 2020s and could happen even faster if innovations in technology and breakthroughs in pricing accelerate. A typical Chicago consumer could cut his commuting cost by $7,000 a year by ditching a personal car in favor a shared autonomous electric vehicle. Auto companies will face wrenching changes as more than 5 million conventional cars are replaced by an estimated 4.7 million autonomous electric vehicles by 2030.

Audio for Your Journey Home 

The Social Impact of Robotics: A Conversation with Kate Darling

Kate Darling is a leading figure in the emerging field of robot ethics that she helped and is currently helping to define at the MIT Media-Lab. In this podcast Kate discusses the emotional connection between robots and human and the ramifications that we have not planned for that are just around the corner.


New research explores how to filter out unfairness from machine learning

Past data used to train machine learning algorithms is often biased against certain subpopulations. Hence it is important for machine learning predictors to account for this to avoid perpetuating discriminatory practices.

A team of researchers at The Alan Turing Institute proposed an approach methods to analyse the fairness of such algorithms and see whether they contain hidden discrimination. Their approach called “counterfactual fairness” aims to address fairness through causal relationships rather than relying entirely on correlations, like the spurious one between race and arrest for marijuana possession.

To illustrate this, the team analysed a similar policing example in the paper using stop and frisk data from New York City.

+ Hear more expert opinions at our CogX Innovation Exchange held in association with The Alan Turing Institute.

Products We Love

Google’s AutoDraw Turns Doodles to Clipart Masterpieces

Building upon Quick, Draw! (another of
Google’s AI Experiments) Google have designed AutoDraw. Instead of telling you what you’ve drawn (Quick Draw!) it offers a suggestions of a user generate image that… well… looked better than my doodles every time. While not perfect it still offers some moments of genuine “wow” but could easily eat up a few hours.

Another Example of Robots Taking Human’s Jobs 

Tesla’s “Advanced Automation” Division Gears Up for Model 3 Production

Last year, Tesla announced plans to acquire German engineering firm Grohmann Engineering and reconstitute it as “Tesla Advanced Automation Germany.” Now, the carmaker is using all of its new possession’s resources to gear up for Model 3 production. CEO Elon Musk wants the company to be building 500,000 cars a year next year, the more affordable Model 3 will make up the majority of that total. Tesla already has hundreds of thousands of reservations for the car, but the company has also missed every one of its deadlines for new-car launches so far.

Dinner Talk

Machines learning evolves, and hackers stand to gain

Can Machine Learning be used as a cyber security tool? As government agencies are beginning to turn over security to automated systems that can teach themselves, the idea that hackers can sneakily influence those systems is becoming the latest (and perhaps the greatest) new concern for cybersecurity professionals.


Did you catch that? Robot’s speed of light communication could protect you from danger

Terrorism and security are constant news topics nowadays. Cornell University researchers are developing a system to enable teams of robots to share information as they move around, and if necessary, interpret what they see. This would allow the robots to conduct surveillance as a single entity with many eyes. What does this mean for the future?

Chat Bots, Yadda Yadda Yadda

Payments is the ‘killer app’ for chatbots

Integrating payments capabilities into messaging apps could be the key to making chatbots the next big platform, according to Kik CEO Ted Livingston. Without integrated payments chatbots provide some useful and engaging experiences, but it is difficult for brands to drive revenue. If chat apps can monetize chatbot interactions the way Apple did for the App Store, it could one day generate up to $32 billion in revenue for the platforms, according to an earlier report by BI Intelligence.

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!




Published in

Leave a reply

Log in with your credentials


Forgot your details?

Create Account