Issue #33: When AI makes sense of emotions
Hello there!
This week I’ve been thinking a lot about the data humans are creating — without even noticing it. How they look, speak, or even how fast their heart beats can indicate emotions like stress, anger, or excitement. AI can learn to analyze these patterns, yet I wonder if this isn’t a little too private — and if employers and business leaders should take advantage of software that could explore the emotions of workers or clients. I’d love to read your comments on that!
Take care
Alice
Headlines you shouldn’t miss
THE VERGE DeepMind says its new AI coding engine is as good as an average human programmer: The coding engine AlphaCode is supposed to compete with human programming skills, and first tests indicate that it’s ranging in the top 54% of human coders. Yet, coding algorithms won’t replace humans — but they might become assistants. Although there have been massive steps forward, coding algorithms are still prone to produce buggy code or have security vulnerabilities.
FORTUNE The labor shortage is bringing Blade Runner to bartending as A.I. and robots start taking your drink order: Estonian company Yanu.ai has developed a robot bartender — an AI-driven bar with robot arms that is not only capable of mixing up to 100 drinks per hour but as well talk to customers. Yanu’s founder, CEO Alan Adojaan, faces opposition from some trade unions fearing the impact of automation. Yet, there is a drastic labor shortage pushing busy venues with a bar towards automated solutions.
VENTURE BEAT How AI will drive the hybrid work environment: AI can improve the work-from-home experience by optimizing task flows or communication processes in organizations. However, companies need to establish a proper cloud infrastructure first — an investment many are struggling with. Once implemented, AI can analyze individual work patterns and find the best tasks for each worker.
THE HILL No, China is not winning the AI race: Law professor James Cooper and AI advisor Kashyap Kompella argue that the US has the edge over China in the global AI competition. While many see the high manpower in STEM fields and the increasing number of Chinese AI patents as proof of (upcoming) global dominance, both experts argue that the US is in a favorable position as it attracts worldwide talent and allows for more free experimentation due to the liberal university ecosystems.
BROOKINGS The EU and U.S. are starting to align on AI regulation: The Biden administration signals with new hires and policy measures that AI regulation is becoming increasingly important. This approach brings the US closer to the EU, which has been working on a legal, competitive and fair framework for the development and application of AI. With these new prospects, the US and EU could collaborate to establish a shared understanding and definition of AI.
When AI is making sense of emotions
💗 Is human emotion quantifiable? Since the dawn of the earliest technology, philosophers, ethics experts, and tech gurus have been debating this question. Attempts to make sense of facial expression, choice of words, speaking pace, intonation, and even the blinking of the eye have fueled a wide array of AI apps. With strong pattern recognition capabilities, AI could identify emotional cues which might be challenging to explain or what some people might simply describe as “intuitive knowledge.” While emotion is still difficult to grasp for both humans and machines, there is a large number of AI applications maneuvering emotions.
💬 A recent Harvard Business Review article recommends various AI-driven services to business leaders to increase their “people skills” — particularly communication and empathy. For instance, Cyrano.AI has developed a patent that analyzes previous communication patterns with customers and helps business leaders adjust their presentation style according to the customer's personality type. The service Cresta on the other hand analyzes speech in real-time and gives consultants immediate prompts and tips to sound more empathetic.
👨👩👧 Another domain where AI is being increasingly applied is the field of mental health and counseling. AI chatbots are supposed to understand if somebody is in a poor mental state by analyzing their choice of words. Algorithms are being developed to analyze conflict potential in families and relationships, and apps remind couples to express love and affection. Data from smartwatches could interpret heart rate, breathing, and other vital markers to analyze if somebody is experiencing stress and subsequently guide them through a relaxation exercise.
❓ Clearly, there are many ways to analyze behavior and emotions. Algorithms are picking up on the cues humans are sending. While there can be positive results like improving personal stress management, AI that picks up on emotions at the workplace poses various challenges and caveats. First of all, it is uncertain if a human emotional expression is universal. With AI reproducing racial and gendered bias, it wouldn’t be surprising if there are significant cultural, regional, and even age-related differences in the way people express emotions. Depending on the type of learning set, AI could learn specific patterns for one group of people, while neglecting the other. Secondly, in my opinion, “emotional data” is private data no employer should touch. Intonation, the blinking of the eyes, or vital symptoms are aspects that give away information, yet they are deeply personal and can hardly be controlled. It’s an ethical question to decide who is allowed to have access to which “emotional data”.
Video of the week: Bloomberg explains recruiting AI
In the United States, AI-assisted recruiting programs are helping to screen resumes, optimize job openings, and analyze the personality types and skills of applicants. The technology has the potential to increase efficiency in HR and is already widespread in large corporations. Applicants and workers rights’ advocates are concerned. Watch Bloomberg’s video or click on the link: How AI is Deciding Who Gets Hired - YouTube