Data Science Insider: August 26th, 2022

In This Week’s SuperDataScience Newsletter: AI Model Can Detect Parkinson’s from Nighttime Breathing Patterns. FN Meka Dropped by Record Label. Facial Recognition for Turtles. AI Startup Makes Call Centre Workers Sound ‘Whiter.’ Viral LinkedIn Post Generator. Cheers, - The SuperDataScience Team P.S. Have friends and colleagues who could benefit from these weekly updates? Send them to [this link]( to subscribe to the Data Science Insider. --------------------------------------------------------------- [AI Can Detect Parkinson’s from Nighttime Breathing]( brief: Researchers from MIT have developed an AI model that can detect Parkinson’s, at an early stage, just from simply reading a person’s breathing patterns. Principal Investigator, Dina Katabi, said: “Parkinson's disease diagnosis relies on motor symptoms like tremors and stiffness, however, those symptoms tend to appear several years after the onset of the disease. But no physician today can detect Parkinson's or assess its severity merely from breathing. AI can help doctors and medical professionals extract new insights from standard physiological signals like breathing, heart rhythms, electrical activity, gait or walking patterns. This is because AI, or neural networks, can detect complex patterns that may be hard for humans to see in this kind of data." The researchers tested the model on 7,671 individuals using data from several public datasets and hospitals in the United States. The AI learned to predict each subject's quantitative electroencephalogram (qEEG) from nocturnal breathing. Why this is important: Parkison’s Disease is the fastest-growing neurological disease worldwide. This means that there is an urgent need for novel diagnostic biomarkers that can detect the disease at an early stage. This advance in detection by MIT could be the breakthrough we need. [Click here to learn more!]( [FN Meka Dropped by Record Label]( brief: Last week in the SuperDataScience weekly newsletter, we followed the story of how virtual rapper FN Meka had been signed to Capitol Records, who billed the rapper as “the world’s first AR [augmented reality] artist to sign with a major label.” However, what was a rather lighthearted story has taken a dark turn with the label being forced to drop him over accusations of racial stereotyping, including numerous uses of the N-word and stimulating being beaten by a police officer. A statement from Capitol Records said: “We offer our deepest apologies to the Black community for our insensitivity in signing this project without asking enough questions about equity and the creative process behind it. We thank those who have reached out to us with constructive feedback in the past couple of days – your input was invaluable as we came to the decision to end our association with the project.” Why this is important: We’ve covered stories numerous times about how the data used for AI isn’t adequately checked for bias. This story is an extreme case of how AI, when left unchecked, can perpetrate harmful stereotypes. [Click here to read on!]( [Facial Recognition for Turtles]( In brief: In November 2021, marine conservation organizations came together with Zindi, the largest professional network for data scientists in Africa and DeepMind, the British AI subsidiary of Alphabet Inc. and research laboratory, in order to design a challenge that would aid conservation efforts by using ML to distinguish between turtles of the same species. Keeping track of the number of sea turtles is incredibly difficult but important for conservation efforts as they are regarded by biologists as an indicator species or class of organisms whose behaviour helps scientists understand the underlying welfare of their ecosystem. As such, a competition, dubbed Turtle Recall, was created to build a model that would identify individual sea turtle faces. The ML tool was successful as it used a dataset already created by Zindi and the fact that the pattern of scales on a turtle's face is unique and remains the same over their lifespan. Why this is important: Here at SuperDataScience we’re used to stories where facial recognition technology is coming under fire for being biased or inaccurate so makes a great change to read a story where the positives of the technology are apparent. [Click here to discover more!]( [AI Startup Makes Call Centre Workers Sound ‘Whiter’]( In brief: Silicon Valley start-up, Sanas, has caused criticism this week by unveiling technology they’ve developed, which can change the accents of call centre workers in real-time. The AI is designed to make speakers sound like white Americans, which the company claims could help to overcome accent-based prejudice and reduce racist abuse faced by staff. However, critics have been vocal in claiming that it is a move in the wrong direction and language diversity should be celebrated instead. Sanas has targeted Asia, where a large number of call centres are based, to implement the technology- with around 1,000 call centre workers in the Philippines and India already using the software daily. They make the changes to workers’ voices by gathering data about the sounds of different accents and how they correspond to each other, their AI engine then transforms a speaker’s accent into what passes for another one. Why this is important: Language neutralization training was already big business in this part of the world. Some may see this as a tool which simply assists in this task, but to many, it's another indication that being anything other than a white westerner is wrong and can be corrected. . [Click here to see the full picture!]( [Viral LinkedIn Post Generator]( In brief: This comical article by Jennifer Wong, in the Guardian, examines the AI that has been making waves on social media this week. Examples of the work of AI bot ViralPostGenerator have, well gone viral after Twitter users discovered the cringey posts that could be created and shared on the employment-oriented online service, LinkedIn. The tool was created by Tom Orbach with the goal of allowing you to” use AI to write the perfect LinkedIn post." In order to create the post, you simply fill two boxes appropriately titled: What did you do today? And inspirational advice. You can also select the cringe meter to your desired level. Orbach created the tool by tasking the AI with analyzing more than 100,000 posts that had gone viral on LinkedIn. The launch was so successful that within a week the ViralPostGenerator was acquired by Taplio, an advertising agency specializing in LinkedIn content. Why this is important: This story is rather funny and the creativity of Twitter users when let loose with the technology offers a fun trip down the rabbit hole. However, the technology does highlight how AI is leading us towards more template-driven content which is designed to go viral, rather than offering any real value. [Click here to find out more!]( [Super Data Science podcast]( this week's [Super Data Science Podcast](, Christina Stathopoulos, Analytical Lead for Waze and Adjunct Professor at IE Business School, joins us for a captivating talk that covers geospatial data and the hard and soft skills required to build a thriving career in data science." --------------------------------------------------------------- What is the Data Science Insider? This email is a briefing of the week's most disruptive, interesting, and useful resources curated by the SuperDataScience team for Data Scientists who want to take their careers to the next level. Want to take your data science skills to the next level? Check out the [SuperDataScience platform]( and sign up for membership today! Know someone who would benefit from getting The Data Science Insider? Send them [this link to sign up.]( # # If you wish to stop receiving our emails or change your subscription options, please [Manage Your Subscription]( SuperDataScience Pty Ltd (ABN 91 617 928 131), 15 Macleay Crescent, Pacific Paradise, QLD 4564, Australia