On show at the Lethaby Gallery to 20 April, MA Material Futures and DeepMind present Model Machines: featuring twelve student projects that use AI to offer solutions to complex challenges and contribute to fair and sustainable futures.
Stretching human capabilities in practical and thought-provoking ways, these ideas include Mindwave, a device to ease people’s social phobias; PERL_AI, a menopause heirloom to accompany generations of women through this profound transition; and Airdrop AI, which combines AI and 3D printing to drop personalised prostheses simply and quickly to adults and children in war zones. In this conversation, the students share some of the ethical perspectives guiding their ideas for a positive future AI.
MindWave
Lucy Gao and Suryanka Bagri
This device uses AI-enabled technologies to tap into the unconscious behaviours of the human mind that are often beyond a person’s control in order to identify, relieve and overcome cognitive phobia triggers. Emotion Recognition (ER) and Machine Learning (ML) technologies can measure and analyse the skin conductivity, heartbeat and brain activity of a user suffering from a phobia. The device then transmits electrical frequencies through earphones to help stabilise the user’s condition, offering a cognitive distraction that, over time, helps alleviate and potentially overcome the phobia itself.
What kind of ethical considerations did you deal with in the course of your project?
At the start of the project after figuring out what MindWave would become, we knew there could be some ethical risks that would be asked about the project. MindWave’s devices are based on combining EEG technologies and artificial intelligence to detect the user’s brainwaves and emotions to help ease the person’s phobias. It is understandable that there would be ethical concerns especially when the user is providing information about their most intimate factors and reactions of their mind, which is the most private personal information for humans.
To provide less of a risk, the user’s data can only be downloaded by themselves, or a doctor that has recommended the MindWave devices to their patients and given approval by the user to access and download their personal results onto their laptops.
What were the challenges of working with machine intelligence on human-centred projects?
The main challenges working with machine intelligence on human centred projects was to find the right balance between both worlds. MindWave is based on trying to ease people who have severe social phobias, so for this project we wanted to create an alternative method for patients that are usually recommended medication that comes with heavy side effects. If the project had a working artificial machine intelligence within the devices; to make it practical the computing systems would require to be compatible for both passive and active emotions that are connected within one’s phobias. The machine would be able to understand deep within what the user’s triggers to their phobias, how the human is feeling emotionally and detect which brainwave is rising abnormally to then provide the correct sound frequency track to calm the user’s nerves down.
What can creative perspectives bring to AI?
I believe creativity is an ability that can be imaginative or perceiving the world in innovative ways. Especially with artificial intelligence, machine can dream up an image in various intricate ways that could be perceived different to human creativity. AI’s creative perspectives can bring new beginnings to where we as humans cannot think of anything new and different to what we have done in the past. Creativity within artificial intelligence can either be a really great thing or a negative impact, as I think if we rely too much on artificial intelligence such as ChatGPT, humans could also lose creativity and imagination of their own.
The exhibition shows that AI can be used in very practical ways and for public good. Why is this important?
Artificial intelligence offers advantages that can improve and help move businesses across many sectors across the world. The use of artificial intelligence has increased over the years as it is seen to be more consistent and can make swift but efficient automatic decisions. It has been noticed that artificial intelligence has been applied to good for people through diagnosing cancer, helping blind people navigate their surroundings, agriculture farming and education. I think the use of artificial intelligence should be applied to various types of personal devices for the public, whether it is through to help with language translation, tracking different types of health problems to creating awareness of one’s mental health.
PERL_AI
Adriana Tabuenca Jimenez and Julia Labis
Many women can feel isolated and completely alone during menopause with those around them not really understanding how difficult and challenging this transition can be. This smart AI device is intended to not only gather much-needed personal data that could radically improve current research within the field, but also encapsulate this data in a future ‘pearl’. Supporting medical research, the pearl can also act as a personal heirloom, which women can pass down to their daughters who will most likely experience the same menopausal symptoms as their biological mothers.
What kind of ethical considerations did you deal with in the course of your project?
The fact that our project is based on menopause is a statement by itself and involves risks regarding confidentiality. After exhaustive research we found there was a lack of research and conversations about menopause, let alone technological devices that target this area - a lot of that due to complexity of the symptoms and how intimate they can be.
This made us question the reasons why society does not pay attention at all to a biological process that 50% of population will experience at some point in their lives. While gathering such delicate data we realised the importance of privacy in a world where everything is shared and uploaded to a network where the user's agency is sometimes arguable.
We were particularly cautious when looking for individual cases of menopause. Medical data is personal and confidential so we made sure none of that research crossed any boundaries or was controversial in approach.
What were the challenges for you of working with machine intelligence on human-centred projects?
One of the main challenges was to fully understand how Artificial Intelligence works and the consequences it could have. Coming from an artistic background it is not easy to comprehend what a algorithm is capable of. While designing our project, we had to imagine future scenarios with our product in mind and consider its impacts, both positive and negative, for our potential users - menopausal women.
During the project we were stuck for a while trying to find a solution to how best design for such a broad, yet specific group of women. The lack of general research and knowledge about menopause, that fact that it´s still a taboo, increased the difficulty of our design process. We do think that menopause is both a societal and personal discussion that needs to be brough up to a daily life.
What can creative perspectives bring to Ai?
The simplest answer is - a lot. AI is a very useful tool because it can recognise patterns and analyse data much faster than humans can, but because of its complex and intangible nature, it seems like an unknown concept or something out of a sci-fi movie.
Creative perspectives on Ai can bring it closer to humans and help us find new practical and beneficial applications for it. Ones that seem more approachable by an every-day person and stand outside of the mainstream AI uses, like chatbots and apps.
The mix of creativity and Ai also allows those, who are currently designing Ai - such as DeepMind, to think about the end user of their products and what kind of futures are they envisioning through their concepts. Creative perspectives put Ai into a more tangible and less unknown context - they envision cutting-edge technologies entering our every-day lives.
The exhibition shows that AI can be used in very practical ways and for public good. Why is this important?
It’s important because Ai is a tool, and as all tools it can be either used for the good or for the bad. At the beginning of our brief, one of our tutors compared Ai to a hammer. A hammer it’s a tool that can be used to both construct and destruct. It’s practical, but it can be dangerous if used badly or for malicious purposes. The same analogy can be used for Ai, and the sooner a positive, approachable image of Ai is created the better, as it would mean more people will see it as a tool to create positive changes around us and contribute to it. It’s important to see new technologies as something that can benefit our existence, rather than threaten it, because they really can help us improve our lives when used ethically.
Airdrop AI
Ali Hossain and Alonso Hernandez
Airdrop AI uses AI to produce prostheses for war zones. A 3D printer that uses SAI and scanning software found on any smart phone can quickly analyse images, videos and scans of a patient’s limbs to create simple protheses for adults and children who have lost limbs. It can be air dropped into a war zone without critical infrastructure – roads, hospitals etc.
What kind of ethical considerations did you deal with in the course of your project?/ What, if any, are the risks involved?
Our project designs emergency prosthetics that fit the wound, and once the crisis is over, the patient will require a medical examination. So, one of the primary concerns is privacy. Gathering and interpreting medical data and requesting to be scanned implies giving access to sensitive personal information. All these data will be stored and processed so they must be kept confidential and used only for their intended purpose. Also, they must comply with data protection regulations, such as the Protected Health Information or the General Data Protection Regulation to secure patient data against unprecedented cyber threats. So basically it’s to protect customer data from unethical use and distribution without the proper consent. Additionally, it’s to verify any biases against certain patients or groups.
Another consideration was the safety and effectiveness of the prosthetic. Ideally, before launching the project rigorous testing must be completed to ensure that the prosthetic devices are safe and effective and do not harm patients in any way, including the materials. Likewise, there are ISO with which artificial intelligence can utilise to design and print, like ISO 10328:2016 (Structural testing of lower-limb prostheses). In summary, we faced several ethical considerations in the development and implementation of AI-powered prosthetic emergency devices; patient privacy, patient autonomy, safety, and efficacy.
What were the challenges for you of working with machine intelligence on human-centred projects?
For this project, it was complicated to determine where artificial intelligence was going to be applied. In the beginning, we did not know about AI, so the first weeks were about research and understanding the whole learning process of AI. I can say the main challenge was to balance the use of AI technology with the human touch; at the end of the day, not everyone will trust a "robot" to provide a prosthesis.
Time after choosing that, was to determine what information the artificial intelligence receives, how it processes it, and how it uses that data for human benefit. Also, we must ensure that our use of AI aligns with ethical considerations. But, first and foremost, it was to understand how anyone would trust AI and that the final piece should be intuitive, the user would know what was supposed to happen and how to collect the prosthesis.
What can creative perspectives bring to AI?
Generally speaking, I believe creative perspectives can propose new alternatives, not existing applications of AI. This project brought new ways to apply prosthetics and medical data. In the future, it can improve the design of prosthetics, search for new materials and improve the user experience - if we can call it that way- a person must go through to get a prosthesis. But, most importantly, it allows us to explain and showcase in an accessible way. So people can have a better approach to a trendy topic and maybe afford to imagine new scenarios; other than those we had in media (books, series, or movies)
The exhibition shows that AI can be used in very practical ways and for public good. Why is this important?
This exhibition allows people to clarify some misconceptions about AI since it's a trending topic. By showcasing speculative applications of AI that serve the public good, such as in healthcare, the exhibition can help to understand the applications of AI and how it can be used as a tool that might benefit some aspects of our daily life. Ultimately, AI has the potential to solve many complex problems, but it is important to ensure it is utilised with responsibility. Only by exposing this topic, we generate a conversion around it.