Forgive my impatience, but I feel strongly that the pace of affective computing adoption remains frustratingly slow. This is a field of study combining insights from computer science, psychology and cognitive science to build machines that can better understand and support their human counterparts – concepts that are now timelier than ever!
Extraordinary technologies are coming onstream that are bristling with human-like capabilities such as reasoning, planning and imagination. Yet still we interface them just like we did with dumb systems. A definite case of dumb and dumber in my book – especially with all the promise of affective computing. Economic prosperity, social wellbeing and everyday convenience are being held back by hesitance, ethical debates and, yes, outdated interface models.
Sentient Machines
The recently announced partnership between Sir Jony Ive and Sam Altman of OpenAI underlines my point. They described how computers are now seeing, thinking and understanding – yet despite that, humanity’s experience remains shaped by traditional products and interfaces.
For me, affective computing is a key enabling component for the future of human-machine interaction. It’s a future based on intelligent systems that partner with human cognition through human-machine understanding (HMU). HMU draws on insights from AI, cognitive science, neuroscience and psychology to enable machines to infer goals, interpret behaviour and adapt their support based on intent, workload, stress and experience – just as a good teammate would do.
Applying HMU in the right way will unlock new modes of collaboration, where humans and machines co-adapt and prosper in real time. This can transform sectors such as healthcare, industry and consumer. Some early adoptions of affective computing innovation have already resulted in use cases in wellbeing, media analytics, automotive and fraud detection.
Market Impact
The market for affective computing technologies is expected to grow significantly in the coming years. Estimates are plentiful – and here are a couple of examples for perspective. Verified Market Research projects the market will grow at a compound annual growth rate (CAGR) of 36.5 per cent between 2026 and 2032, while LinkedIn cites a CAGR of 29.72 per cent from 2024 to 2030.
Healthcare Tech
Healthcare is just one area of huge opportunity. Affective computing technologies can be integrated into healthcare tech to provide more empathetic and personalised experiences.
This involves AI systems capable of recognising patient mental states, which allows for tailored support in areas such as mental health and wellbeing. Specifically, these technologies support patient care and diagnosis, with applications in socially assistive robotics to aid patient care, especially for mental health conditions. In remote patient monitoring affective computing can be used to detect signs of distress or changes in patient states, which ensures timely intervention and enhances the overall quality of care and life.
Media Analytics
In media analytics meanwhile, affective computing technologies are providing advanced systems that analyse real-time emotional responses to content. These systems utilise various inputs, including facial expressions, user attention, tone of voice and physiological signals, to provide deeper insights into audience engagement. This capability enables content creators and advertisers to optimise their material for maximum impact and tailor content to specific user profiles, enhancing the effectiveness of media and advertising.
Automotive Applications
Affective computing technologies are enhancing the automotive sector by creating safer and more comfortable driving experiences. These technologies monitor driver states and attentiveness to detect and prevent dangerous situations. By detecting driver distraction, drowsiness, stress or anger, affective computing can provide timely warnings or adjust vehicle settings to mitigate potential risks.
Crucially, these technologies facilitate more natural and intuitive human-machine interfaces within vehicles, improving communication and control by responding to the driver states. This becomes increasingly important as we move towards autonomous vehicles, and the role of the driver in the cabin evolves.
Fraud Detection
In fraud detection, affective computing can analyse subtle cues during financial transactions and online interactions. These systems primarily rely on voice and face ID. While there is interest in monitoring facial expressions and micro-expressions as indicators of fraudulent behaviour, this remains highly controversial and unreliable. Lightweight facial verification techniques use affective computing to detect faked user behaviour through bots, and to tackle user fraud better than typical CAPTCHA (setting a test to check if you’re human) approaches.
Stepping Up
My examples above illustrate the benefits of integrating just a part of what affective computing can achieve by augmenting existing systems. They are great benefits for sure – but they fall short of the step change that true Human-Machine Understanding can create. This brings me back to my opening message about urgency. We need a mind-shift across industry – a fundamental and collaborative rethink of how we work with machines.
Affective computing technologies introduce several ethical and regulatory challenges because they operate at a very personal level. Key concerns include data privacy, the potential for emotional dependency, and the impact of cultural biases in AI algorithms.
Legislation and Regulation
Regulations, such as the EU Artificial Intelligence Act, are evolving to address these challenges, imposing restrictions on certain AI applications like emotion recognition in workplaces and education. Regulators are faced with controversy and concerns by researchers and practitioners – but I urge stakeholders to proceed with speed as well as care.
Inter-institution Collaboration
A further point on collaboration. It’s disappointing to see that the majority of this potential and promise remains limited to the research lab. I think it’s time for greater synergy between industry and academia. The former can play its part by being more proactive in identifying commercial opportunities and ensuring that affective computing technologies are scalable and integrable into existing systems. Academia, meanwhile, can push harder on the boundaries of knowledge and technology, unconstrained as it is by immediate commercial or regulatory pressures.
Initiatives like joint research projects can lead to the development of robust tools and methodologies that are not only innovative but also practical for real-world applications. This is the way to ensuring that transformative affective computing technologies are both cutting-edge and viable for the market.
Ali Shafti, Head of Human-Machine Understanding at Cambridge Consultants, part of Capgemini Invent
Apprenticeship Funding Cuts to Cost Employers £214m
I had no idea the taxpayer was funding such an advanced level of education. Certainly graduates and post-graduates don´t receive comparable financial...