Imagine you are in a job interview. As you answer the recruiter’s questions, an artificial intelligence (AI) system scans your face , scoring you for nervousness , empathy and dependability . It may sound like science fiction, but these systems are increasingly used, often without people’s knowledge or consent.

Emotion recognition technology (ERT) is in fact a burgeoning multi-billion-dollar industry that aims to use AI to detect emotions from facial expressions. Yet the science behind emotion recognition systems is controversial : there are biases built into the systems.

Many companies use ERT to test customer reactions to their products, from cereal to video games. But it can also be used in situations with much higher stakes , such as in hiring , by airport security to flag faces as revealing deception or fear, in border control , in policing to identify “dangerous people” or in education to monitor students’ engagement with their homework.

Shaky scientific ground

Fortunately, facial recognition technology is receiving public attention. The award-winning film Coded Bias, recently released on Netflix, documents the discovery that many facial recognition technologies do not accurately detect darker-skinned faces. And the research team managing ImageNet, one of the largest and most important datasets used to train facial recognition, was recently forced to blur 1.5 million images in response to privacy concerns .

Revelations about algorithmic bias and discriminatory datasets in facial recognition technology have led large technology companies, including Microsoft, Amazon and IBM, to halt sales. And the technology faces legal challenges regarding its use in policing in the UK. In the EU, a coalition of more than 40 civil society organisations have called for a ban on facial recognition technology entirely.

Like other forms of facial recognition, ERT raises questions about bias, privacy and mass surveillance. But ERT raises another concern: the science of emotion behind it is controversial. Most ERT is based on the theory of “basic emotions” which holds that emotions are biologically hard-wired and expressed in the same way by people everywhere.

This is increasingly being challenged, however. Research in anthropology shows that emotions are expressed differently across cultures and societies. In 2019, the Association for Psychological Science conducted a review of the evidence, concluding that there is no scientific support for the common assumption that a person’s emotional state can be readily inferred from their facial movements. In short, ERT is built on shaky scientific ground.

Also, like other forms of facial recognition technology, ERT is encoded with racial bias. A study has shown that systems consistently read black people’s faces as angrier than white people’s faces, regardless of the person’s expression. Although the study of racial bias in ERT is small, racial bias in other forms of facial recognition is well-documented.

There are two ways that this technology can hurt people, says AI researcher Deborah Raji in an interview with MIT Technology Review : “One way is by not working: by virtue of having higher error rates for people of color, it puts them at greater risk. The second situation is when it does work — where you have the perfect facial recognition system, but it’s easily weaponized against communities to harass them.”

So even if facial recognition technology can be de-biased and accurate for all people, it still may not be fair or just. We see these disparate effects when facial recognition technology is used in policing and judicial systems that are already discriminatory and harmful to people of colour. Technologies can be dangerous when they don’t work as they should. And they can also be dangerous when they work perfectly in an imperfect world.

The challenges raised by facial recognition technologies – including ERT – do not have easy or clear answers. Solving the problems presented by ERT requires moving from AI ethics centred on abstract principles to AI ethics centred on practice and effects on people’s lives.

When it comes to ERT, we need to collectively examine the controversial science of emotion built into these systems and analyse their potential for racial bias. And we need to ask ourselves: even if ERT could be engineered to accurately read everyone’s inner feelings, do we want such intimate surveillance in our lives? These are questions that require everyone’s deliberation, input and action.

Citizen science project

ERT has the potential to affect the lives of millions of people, yet there has been little public deliberation about how – and if – it should be used . This is why we have developed a citizen science project .

On our interactive website (which works best on a laptop, not a phone) you can try out a private and secure ERT for yourself, to see how it scans your face and interprets your emotions. You can also play games comparing human versus AI skills in emotion recognition and learn about the controversial science of emotion behind ERT.

Most importantly, you can contribute your perspectives and ideas to generate new knowledge about the potential impacts of ERT. As the computer scientist and digital activist Joy Buolamwini says : “If you have a face, you have a place in the conversation.”

This article by Alexa Hagerty , Research Associate of Anthropology, University of Cambridge and Alexandra Albert , Research Fellow in Citizen Social Science, UCL , is republished from The Conversation under a Creative Commons license. Read the original article .

Reinforcement learning could be the link between AI and human-level intelligence

Last week, I wrote an analysis of “Reward Is Enough,” a paper by scientists at DeepMind. As the title suggests, the researchers hypothesize that the right reward is all you need to create the abilities associated with intelligence , such as perception, motor functions, and language.

This is in contrast with AI systems that try to replicate specific functions of natural intelligence such as classifying images, navigating physical environments, or completing sentences.

The researchers go as far as suggesting that with well-defined reward, a complex environment, and the right reinforcement learning algorithm, we will be able to reach artificial general intelligence, the kind of problem-solving and cognitive abilities found in humans and, to a lesser degree, in animals.

The article and the paper triggered a heated debate on social media, with reactions going from full support of the idea to outright rejection. Of course, both sides make valid claims. But the truth lies somewhere in the middle. Natural evolution is proof that the reward hypothesis is scientifically valid. But implementing the pure reward approach to reach human-level intelligence has some very hefty requirements.

In this post, I’ll try to disambiguate in simple terms where the line between theory and practice stands.

Natural selection

Credit: George Desipris

In their paper, the DeepMind scientists present the following hypothesis: “Intelligence, and its associated abilities, can be understood as subserving the maximisation of reward by an agent acting in its environment.”

Scientific evidence supports this claim.

Humans and animals owe their intelligence to a very simple law: natural selection. I’m not an expert on the topic, but I suggest reading The Blind Watchmaker by biologist Richard Dawkins, which provides a very accessible account of how evolution has led to all forms of life and intelligence on out planet.

In a nutshell, nature gives preference to lifeforms that are better fit to survive in their environments. Those that can withstand challenges posed by the environment (weather, scarcity of food, etc.) and other lifeforms (predators, viruses, etc.) will survive, reproduce, and pass on their genes to the next generation. Those that don’t get eliminated.

According to Dawkins, “In nature, the usual selecting agent is direct, stark and simple. It is the grim reaper. Of course, the reasons for survival are anything but simple — that is why natural selection can build up animals and plants of such formidable complexity. But there is something very crude and simple about death itself. And nonrandom death is all it takes to select phenotypes, and hence the genes that they contain, in nature.”

But how do different lifeforms emerge? Every newly born organism inherits the genes of its parent(s). But unlike the digital world, copying in organic life is not an exact thing. Therefore, offspring often undergo mutations, small changes to their genes that can have a huge impact across generations. These mutations can have a simple effect, such as a small change in muscle texture or skin color. But they can also become the core for developing new organs (e.g., lungs, kidneys, eyes), or shedding old ones (e.g., tail, gills).

If these mutations help improve the chances of the organism’s survival (e.g., better camouflage or faster speed), they will be preserved and passed on to future generations, where further mutations might reinforce them. For example, the first organism that developed the ability to parse light information had an enormous advantage over all the others that didn’t, even though its ability to see was not comparable to that of animals and humans today. This advantage enabled it to better survive and reproduce. As its descendants reproduced, those whose mutations improved their sight outmatched and outlived their peers. Through thousands (or millions) of generations, these changes resulted in a complex organ such as the eye.

The simple mechanisms of mutation and natural selection has been enough to give rise to all the different lifeforms that we see on Earth, from bacteria to plants, fish, birds, amphibians, and mammals.

The same self-reinforcing mechanism has also created the brain and its associated wonders. In her book Conscience: The Origin of Moral Intuition , scientist Patricia Churchland explores how natural selection led to the development of the cortex, the main part of the brain that gives mammals the ability to learn from their environment. The evolution of the cortex has enabled mammals to develop social behavior and learn to live in herds, prides, troops, and tribes. In humans, the evolution of the cortex has given rise to complex cognitive faculties, the capacity to develop rich languages, and the ability to establish social norms.

Therefore, if you consider survival as the ultimate reward, the main hypothesis that DeepMind’s scientists make is scientifically sound. However, when it comes to implementing this rule, things get very complicated.

Reinforcement learning and artificial general intelligence

In their paper, DeepMind’s scientists make the claim that the reward hypothesis can be implemented with reinforcement learning algorithms , a branch of AI in which an agent gradually develops its behavior by interacting with its environment. A reinforcement learning agent starts by making random actions. Based on how those actions align with the goals it is trying to achieve, the agent receives rewards. Across many episodes, the agent learns to develop sequences of actions that maximize its reward in its environment.

According to the DeepMind scientists, “A sufficiently powerful and general reinforcement learning agent may ultimately give rise to intelligence and its associated abilities. In other words, if an agent can continually adjust its behavior so as to improve its cumulative reward, then any abilities that are repeatedly demanded by its environment must ultimately be produced in the agent’s behavior.”

In an online debate in December , computer scientist Richard Sutton, one of the paper’s co-authors, said, “Reinforcement learning is the first computational theory of intelligence… In reinforcement learning, the goal is to maximize an arbitrary reward signal.”

DeepMind has a lot of experience to prove this claim. They have already developed reinforcement learning agents that can outmatch humans in Go, chess, Atari, StarCraft, and other games. They have also developed reinforcement learning models to make progress in some of the most complex problems of science .

The scientists further wrote in their paper, “According to our hypothesis, general intelligence can instead be understood as, and implemented by, maximizing a singular reward in a single, complex environment [emphasis mine].”

This is where hypothesis separates from practice. The keyword here is “complex.” The environments that DeepMind (and its quasi-rival OpenAI ) have so far explored with reinforcement learning are not nearly as complex as the physical world. And they still required the financial backing and vast computational resources of very wealthy tech companies . In some cases, they still had to dumb down the environments to speed up the training of their reinforcement learning models and cut down the costs. In others, they had to redesign the reward to make sure the RL agents did not get stuck the wrong local optimum.

(It is worth noting that the scientists do acknowledge in their paper that they can’t offer “theoretical guarantee on the sample efficiency of reinforcement learning agents.”)

Now, imagine what it would take to use reinforcement learning to replicate evolution and reach human-level intelligence. First, you would need a simulation of the world. But at what level would you simulate the world? My guess is that anything short of quantum scale would be inaccurate. And we don’t have a fraction of the compute power needed to create quantum-scale simulations of the world.

Let’s say we did have the compute power to create such a simulation. We could start at around 4 billion years ago, when the first life-forms emerged. You would need to have an exact representation of the state of Earth at the time. We would need to know the initial state of the environment at the time. And we still don’t have a definite theory on that.

An alternative would be to create a shortcut and start from, say, 8 million years ago, when our monkey ancestors still lived on earth. This would cut down the time of training, but we would have a much more complex initial state to start from. At that time, there were millions of different lifeforms on Earth, and they were closely interrelated. They evolved together. Taking any of them out of the equation could have a huge impact on the course of the simulation.

Therefore, you basically have two key problems: compute power and initial state. The further you go back in time, the more compute power you’ll need to run the simulation. On the other hand, the further you move forward, the more complex your initial state will be. And evolution has created all sorts of intelligent and non-intelligent life-forms and making sure that we could reproduce the exact steps that led to human intelligence without any guidance and only through reward is a hard bet.

Many will say that you don’t need an exact simulation of the world and you only need to approximate the problem space in which your reinforcement learning agent wants to operate in.

For example, in their paper, the scientists mention the example of a house-cleaning robot: “In order for a kitchen robot to maximize cleanliness, it must presumably have abilities of perception (to differentiate clean and dirty utensils), knowledge (to understand utensils), motor control (to manipulate utensils), memory (to recall locations of utensils), language (to predict future mess from dialogue), and social intelligence (to encourage young children to make less mess). A behavior that maximises cleanliness must therefore yield all these abilities in service of that singular goal.”

This statement is true, but downplays the complexities of the environment. Kitchens were created by humans. For instance, the shape of drawer handles, doorknobs, floors, cupboards, walls, tables, and everything you see in a kitchen has been optimized for the sensorimotor functions of humans. Therefore, a robot that would want to work in such an environment would need to develop sensorimotor skills that are similar to those of humans. You can create shortcuts, such as avoiding the complexities of bipedal walking or hands with fingers and joints. But then, there would be incongruencies between the robot and the humans who will be using the kitchens. Many scenarios that would be easy to handle for a human (walking over an overturned chair) would become prohibitive for the robot.

Also, other skills, such as language, would require even more similar infrastructure between the robot and the humans who would share the environment. Intelligent agents must be able to develop abstract mental models of each other to cooperate or compete in a shared environment. Language omits many important details, such as sensory experience, goals, needs. We fill in the gaps with our intuitive and conscious knowledge of our interlocutor’s mental state. We might make wrong assumptions, but those are the exceptions, not the norm.

And finally, developing a notion of “cleanliness” as a reward is very complicated because it is very tightly linked to human knowledge, life, and goals. For example, removing every piece of food from the kitchen would certainly make it cleaner, but would the humans using the kitchen be happy about it?

A robot that has been optimized for “cleanliness” would have a hard time co-existing and cooperating with living beings that have been optimized for survival.

Here, you can take shortcuts again by creating hierarchical goals, equipping the robot and its reinforcement learning models with prior knowledge, and using human feedback to steer it in the right direction. This would help a lot in making it easier for the robot to understand and interact with humans and human-designed environments. But then you would be cheating on the reward-only approach. And the mere fact that your robot agent starts with predesigned limbs and image-capturing and sound-emitting devices is itself the integration of prior knowledge.

In theory, reward only is enough for any kind of intelligence. But in practice, there’s a trade off between environment complexity, reward design, and agent design.

In the future, we might be able to achieve a level of compute power that will make it possible to reach general intelligence through pure reward and reinforcement learning. But for the time being, what works is hybrid approaches that involve learning and complex engineering of rewards and AI agent architectures.

This article was originally published by Ben Dickson on TechTalks , a publication that examines trends in technology, how they affect the way we live and do business, and the problems they solve. But we also discuss the evil side of technology, the darker implications of new tech, and what we need to look out for. You can read the original article here .

AI discovered an old arthritis drug that can help elderly people survive COVID-19

An arthritis drug discovered by AI could reduce the risk of elderly people dying from COVID-19, according to a new study .

The once-daily medicine, called baricitinib, was first identified as a potential treatment for the virus by BenevolentAI, a startup based in London.

The company used its AI software to comb through scientific literature for drugs that might block the infection process. The most promising treatment it discovered was baricitinib, which it predicted could stop the infection from entering lung cells.

The new study has now shown the drug’s safety and efficacy in the largest group of patients published thus far.

Scientists from Imperial College London and Sweden’s Karolinska Institutet gave baricitinib to 83 COVID-19 patients with a median age of 81 who were being treated in hospitals across Italy and Spain.

They found that patients receiving baricitinib had a 71% reduction in mortality compared t o the 83 people in the control group who had not taken the drug.

In addition, only 17% of the patients given baricitinib died or required a ventilator, compared to 35% of the patients who received only standard care.

The study suggests the drug could both reduce organ damage caused by inflammation and block the virus from entering human cells.

“We urgently need to find more effective treatments for COVID-19 while we wait for a vaccine to become widely available,” said Imperial’s Professor Justin Stebbing, the study‘s co-lead author. “This is one of the first COVID-19 treatments to go from computer to clinic and laboratory.”

The findings are now being followed up with large-scale clinical trials. You can read the research paper in the journal Science Advances.