This was the advert from The IHM to get us to go to this health chat!
“Rashik Parmar is a Fellow of IBM, the leader of IBM’s European technical community and an IBM Distinguished Engineer. He was previously President of IBM’s Academy of Technology and has spent his whole career immersed in technology and artificial intelligence”.
OK – yes, we hoped for insight. And we did get it.
(And if you want to view the recording of this chat – click here – to see if you agree with my take on it!)
Could he perhaps just give us a bit of a steer on and IT Road map for the NHS? That’s what the IT people usually say. Road map. How pleasantly old fashioned! But what about all that new fanged stuff? Machine Learning and Artificial Intelligence? Always just AI now of course…
Backstory is always the fascinating parts of these health chats with Roy Lilley. And Rashik was no exception. Born in Africa, Mombassa, parents from India. Terrorised at ‘a level you can’t imagine’ as Indians in a foreign land. Carried cash bundles always to pay off threatening locals. Pushed out and ended up in Leeds, following an Aunty from a year prior. Family always held that belief that you had to become a doctor – everything else was a failure. But his careers teacher said he wasn’t bright enough for that route. 1978 was early to decide to go into computing. But that he did, and managed to get A level course in Computer Science at a night school, sourced by a teacher at school. Hope he or she remembers! Decided on Imperial over Cambridge (the only choice then?), because the 3 year course did everything from Silicon to make the chopping, to Coding itself. And he always wanted to understand the A to Z of what he was talking about. Stood him in good stead when he got summer jobs at IBM in Leeds. But an aptitude test and 4 interviews for a summer job says a lot about IBM searching for talent early – or just arrogant??! 1984 now (not prophetically appropriate) – still pre Internet of course.Desk tops had only just arrived. The 20 young visitors we had from Windsor Girls School must have wondered what this ancient history story was all about? More of them later.
What’s the next big thing, asks Roy. We get smaller and faster – but that’s just progress not ground shaking? What about AI and machine Learning? Rashik did point out that you can do very little on a day to day basis without Technology intervening. Even turning on a tap involves many interfaces further down the line to open the right bits of the reservoir and piping and to look after the water quality. We don’t see it most of the time unless it goes wrong of course. He bemoaned SciFi writers for making us believes ruff and systems were already here, not just a figment of imagination.
He told a story that became a good focus for us all. 80 year old picked up from her care home to go for routine out patients appointment. Saw doc for 5 minutes. But had 3 hour wait first. Then was taken to the wrong pick up point for her ambulance home. Not looked at or questioned for 10 hours. No food, no interaction, no one cared.
And there’s the rub. We want the shiny new tech to cure cancer. But let’s be honest. Tech can’t do the full brain and emotion yet. And maybe never will. But it can a really shine in the mundane. Routine work is what it will help with most, in all types of organisation, not just the NHS. The business case in the old lady’s case should be iPad or similar all talking to each other all along the patient pathways. The best part of £600 had been spent on that 5 minute out patients appointment. Only the cleaner asking what she was doing still waiting fixed the situation. Amelia Tickell representing the next generation and her school friends from Windsor wondered if AI at the mass produced robotic level could have helped her. Not today, Rashik said. But companion bots are being introduced in Japan – driven by a total lack of Carers workforce. People though talked and confided more in a cat robot than humanoid…fascinating! She also asked about what should people study the STEM subjects. His advice was fab. “Do what matters to you. Being good is far more important than followed the trend.
Roy did get into population health and peoples data. And big brother watching and admonishing you for that bag of crisps…but we don’t want to be told, we want to be informed. Now I felt we were getting there….
- Should we regulate engineers? There probably does need to be an Ethics and Conduct code.
- Algorithms are always biased to serve the needs and prejudices of the community they are serving.
- 11 to 15 % of the data we have is plain wrong. It needs cleaning first. (And sad we don’t know exact figure!)
- The NHS has a huge bank of talent and is very lucky to have them
- Only 3% of clinicians are comfortable using AI
Let’s expand that last one. Training – e learning – helps when you need to get people tuned into cyber security. It will take huge efforts but hey, perhaps we can do the same with leading edge GPs? Gotta start somewhere!
Machine learning process was fascinating. 3 equations:
Data+ analysis –> information
Information + context –> insight
Insight + action –> Outcome
And to put this into action? You need different skill sets:
Squirrels – to capture and understand data
Owls – to use the data and provide its context
Foxes – the mathematicians, to create the machine learning algorithms.
Weaver Birds – (tiny birds who build the biggest nest to court their mate) these are your data engineers
Hawks – to oversee ethics and legality
No one person can run the project. All skill sets need to be covered.
My final view? No, I agree we can’t abdicate to AI. But we can force it to help with mundane, so we can have time to do the people bits better. The spectrum of skills is a fascinating area. Can we do it? Do we have a choice? As was said more than once, the future comes a lot faster than you think…
There was so much more to this chat. Go watch it. Click here. Fascinating. The future is now. We just need to centralise the project and do it.