The robot in a doctor's coat
Good morning, how can I help you?
My name is Headfield, your medical AI assistant.
I sigh and enter all the data from the consultation just now. Given the towering shortages in healthcare, it seemed such a good idea that us doctors would get AI support, but, if I’m honest, I sometimes feel like I’m the AI’s assistant rather than the other way around. An assistant who talks to patients and enters data, that’s it.
I press a button and a diagnosis comes rolling out of the system. All I have to do is sign it.
After that, I get up to call in the next patient from the waiting room; a pale hollow-eyed boy who sits hunched up between two concerned looking parents. ‘Good morning young man, my name is Dr Henton. Who are you here with today?’
Back in the consulting room, I listen to their story. I examine the boy's inexplicable bruises. I notice his listlessness and look deep into the eyes of the worried parents.
When they are gone, I enter all the data. Headfield returns with a preliminary diagnosis: leukaemia.
Additional lab tests are requested. Waiting time: two weeks.
Only later that night, when I’m at the dinner table with my family, I think to myself: leukaemia? Could this really be leukaemia? A terrible feeling floats around inside me somewhere, a thought of doubt that says: could it really not be something else?
I run through all the information again. I envision the parents sitting beside him, so caring. I reconsider the child’s symptoms and the possible alternative causes. And I realize that, for this child, every day of waiting could be one too many if the diagnosis is wrong.
‘Hey mum,’ my daughter says. ‘Where’s your head at?’
That night, I dreamed that I was watching the sunset. Someone is standing next to me. Someone in a doctor’s coat, someone who is also watching the deeply purple-coloured sky. His face looks angular and cold, the tin face of a robot. His doctor’s coat bears the name Headfield.
‘Am I seeing a contemplative look?’ I ask.
‘Could be. Do you prefer a different facial expression while you are talking to me?’
‘No, not at all, contemplate all you want. Were you actually contemplating something?’
‘No, robots don’t contemplate.’
‘Are you watching the sunset?’
‘Yes. People do it too. They stand and they watch. So I stand and I watch.’
All of a sudden feeling angry, I turn toward the robot.
‘You airhead!’
The terrible feeling that had gripped me during the day turns into rage in my dream. The robot just keeps looking at the horizon.
He cannot do anything but stick to his plan: he is unable to adjust.
So rude.
‘With all due respect ma’am, robots are better at some specific tasks than people. It might perhaps be more correct to say that I haven’t expanded my skills that broadly.’
‘Ugh, whatever. You are lacking humanity, that’s the problem! You would do well to ponder over the case of that little boy with the pale face and the bruises again. Come, I’ll teach you.’
I grab the robot by the hand and want to pull him along, back to the GP’s practice and back to the little boy’s file, but the robot doggedly stays where he is. He doesn’t understand the first thing about human etiquette! He’s not even polite enough to engage with an angry person.
‘Alright, at least explain to me why you think it’s leukaemia. Which symptoms add up to this diagnosis? And have you considered any alternatives?
‘It’s the outcome based on the parameters you entered, ma’am.’
‘Can’t you bloody understand that a child can also get sick from tension and stress? That he can have bruises because he’s being abused?’
I face the robot, grab him by the collar of his doctor’s coat, and pull him toward me until our faces are just inches apart.
‘Are you absolutely certain that this isn’t abuse, can you rule it out? If there’s only a single doubt on our minds, we have to keep looking, keep thinking! We owe it to the child and his parents. Otherwise, that child is going to be waiting for some tests for two weeks while he has to get out of that house NOW!’
Gripping the collar more tightly, I violently shake the robot but he doesn’t respond.
I feel immensely powerless.
The moment freezes; my grip, this powerlessness — it all stays the same.
Beep beep, good morning. It’s a sunny day today with a few clouds here and there but no rain. Rise and shine!
The friendly voice of my AI alarm clock startles me awake. I can still feel the tension of my grip on the robot in my body and the rage still slumbers within me. On my bike to work, the discussion with the robot continues.
‘Yes, the parents seemed loving and concerned, but they could be faking it. Or another relative could be involved, an uncle or something. I don’t know! I just have this feeling, a feeling!’
But I know very well how the robot would have replied:
‘With all due respect ma’am, AI can help one to remain objective.’
When I arrive at work, I know what I have to do. I make some phone calls and finally call the parents of the young patient.
‘Hello, Dr Henton speaking. How is Alain doing today? Yes, that’s what I’m calling about. I don’t have any definitive results for you. However, we would like to admit Alain to hospital for a week for further examination. Yes, that is mainly necessary to rule out a number of suspicions. I will write a referral letter for you and make haste with it. I will immediately call the hospital to let them know it’s an emergency.’
When I hang up the phone a few moments later, I think to myself: There, a week away from home. If we put him in a room with a child who is not extremely sick, we will know soon enough whether this is a question of domestic abuse or not. If my suspicion is right, he will liven up.
I click away the pop-up of Headfield the AI assistant with a satisfied smile.
