Faulty AI emotion recognition and its dangers, China uses DNA to map faces / Humans + Tech - Issue #7
+ New procedure to relieve tremors in Parkinson's patients, AI can detect ADHD better than humans, and Exoskeletons are helping Japan's elderly.
Hi there!
Only two more weeks left in 2019. I hope you have fun plans for the Christmas and New Year holidays.
I have a small favour to ask you — if you’re enjoying these newsletters, please forward them to a friend or two who may enjoy them as well :) Thank you!
Here are the articles for this issue:
🧠Minimally invasive procedure relieves tremors in Parkinson's patients
🔎This AI can detect ADHD better than humans
🇨🇳China uses DNA to map faces, with help from the West
😧AI Now's 2019 Report highlights the dangers of implementing emotion recognition technology
👴Elderly in Japan are wearing exoskeletons to keep working past retirement age
1. Minimally invasive procedure relieves tremors in Parkinson's patients
From ScienceDaily:
A procedure that applies pulses of focused ultrasound to the brain is safe and effective for reducing tremors and improving quality of life in people with essential tremor (ET) or Parkinson’s disease (PD) tremor, according to a new study being presented next week at the annual meeting of the Radiological Society of North America (RSNA).
A more recently available option is magnetic resonance-guided focused ultrasound (MRgFUS) thalamotomy, an incisionless interventional radiology procedure in which focused beams of sound energy are used to heat and destroy a small part of a structure in the brain called the thalamus. The procedure gives relief to the opposite side of the body, meaning that treatment to the right side of the brain would relieve tremors on the left side of the body, and vice versa.
The results are encouraging, with 35 out of 37 patients in the study having a substantial reduction of tremors. This procedure also has the advantage of being a non-invasive procedure, unlike deep brain stimulation which is currently used.
2. This AI can detect ADHD better than humans
Victor Tangermann, writing for Futurism:
A team of researchers used a type of artificial intelligence to predict attention deficit hyperactivity disorder (ADHD) in patients by having it analyze magnetic resonance imaging (MRI) scans. According to a new paper published in the journal Radiology: Artificial Intelligence, their technique could also be used to spot other neurological conditions.
The researchers developed a method using a deep learning model that can analyze multiple connectome maps from different regions of the brain. Their model “improved ADHD detection performance considerably” while analyzing a data set of 973 participants according to the paper.
AI is increasingly getting better than humans at medical diagnosis as reported here, here, and here. This is mostly positive, as it allows doctors to more time to focus on effective treatments. But, I hope they look at the diagnosis from the AI systems critically and not trust them to be right every time.
3. China uses DNA to map faces, with help from the West
Sui-Lee Wee and Paul Mozur, writing for The New York Times:
With a million or more ethnic Uighurs and others from predominantly Muslim minority groups swept up in detentions across Xinjiang, officials in Tumxuk have gathered blood samples from hundreds of Uighurs — part of a mass DNA collection effort dogged by questions about consent and how the data will be used.
In Tumxuk, at least, there is a partial answer: Chinese scientists are trying to find a way to use a DNA sample to create an image of a person’s face.
The technology, which is also being developed in the United States and elsewhere, is in the early stages of development and can produce rough pictures good enough only to narrow a manhunt or perhaps eliminate suspects. But given the crackdown in Xinjiang, experts on ethics in science worry that China is building a tool that could be used to justify and intensify racial profiling and other state discrimination against Uighurs.
Like any technology, this can be used for both bad and good.
Already, China is exploring using facial recognition technology to sort people by ethnicity. It is also researching how to use DNA to tell if a person is a Uighur. Research on the genetics behind the faces of Tumxuk’s men could help bridge the two.
The Chinese government is building “essentially technologies used for hunting people,” said Mark Munsterhjelm, an assistant professor at the University of Windsor in Ontario who tracks Chinese interest in the technology.
Using it to discriminate and persecute people based on race, ethnicity or beliefs definitely falls under bad use of the technology.
4. AI Now's 2019 Report highlights the dangers of implementing emotion recognition technology
From the AI Now Institute’s 2019 Report:
Affect recognition is an AI-driven technology that claims to be able to detect an individual’s emotional state based on the use of computer-vision algorithms to analyze their facial microexpressions, tone of voice, or even their gait. It is rapidly being commercialized for a wide range of purposes—from attempts to identify the perfect employee to assessing patient pain to tracking which students are being attentive in class.
For example, the company Kairos is marketing video-analytics cameras that claim to detect faces and then classify them as feeling anger, fear, and sadness, along with collecting customer identity and demographic data. Kairos sells these products to casinos, restaurants, retail merchants, real estate brokers, and the hospitality industry, all with the promise that they will help those businesses see inside the emotional landscape of their patrons.
Employment has also experienced a surge in the use of affect recognition, with companies like HireVue and VCV offering to screen job candidates for qualities like “grit” and to track how often they smile.
… while Boston-based company BrainCo is creating headbands that purport to detect and quantify students’ attention levels through brain-activity detection, despite studies that outline significant risks associated with the deployment of emotional AI in the classroom.
There remains little to no evidence that these new affect-recognition products have any scientific validity. In February, researchers at Berkeley found that in order to detect emotions with accuracy and high agreement requires context beyond the face and body.
Given the high-stakes contexts in which affect-recognition systems are being used and their rapid proliferation over the past several years, their scientific validity is an area in particular need of research and policy attention—especially when current scientific evidence suggests that claims being made about their efficacy don’t hold up. In short, we need to scrutinize why entities are using faulty technology to make assessments about character on the basis of physical appearance in the first place. This is particularly concerning in contexts such as employment, education, and criminal justice.
This is quite troubling. I wonder how many people are being subjected to these faulty emotion recognition techniques without their knowledge.
Should it be mandated that any use of emotion recognition be disclosed to people before being subject to it? I think so.
Would you go into a retail store that posts a sign outside that they use emotion recognition cameras within their store to help serve you better? Or will they eventually become like the cookie notices that we blindly accept on almost every website we visit?
This is just one of the areas of AI in the report. It’s worth reading the whole report to get an idea of all the different areas in which AI is being deployed and the current state of the AI industry.
5. Elderly in Japan are wearing exoskeletons to keep working past retirement age
Fabienne Lang, writing for Interesting Engineering:
Exoskeletons aren’t merely useful for paralyzed people, super-strong soldiers, or for those assembling automotives — they’re now also being used to give support to the elderly.
NewScientist reported that the older generation in Japan is using exoskeletons to complete regular everyday tasks that allow them to keep working later in life.
With a high rate of aging citizens, Japan currently has one of the oldest populations in the world. In order to counteract this, companies are creating exoskeletons for the elderly to wear to keep working.
26% of Japan’s population is over the age of 65 and the government is contemplating raising the retirement age from 60 to 70.
These exoskeletons will allow the elderly to continue working in jobs where there is a requirement to carry heavy objects as well as provide assistance in day-to-day living.
🎭Time to start wearing masks to bypass all the facial and emotion recognition tech being built around us. Maybe the fashion industry should start adding masks to their new fashion lines from next year. Something tells me they will be very popular :)
Here’s a quote to ponder over this week:
Success in creating AI would be the biggest event in human history. Unfortunately, it might also be the last, unless we learn how to avoid the risks.
—Stephen Hawking
Wish you a brilliant day!
Neeraj