Technical Analyst at the UK National Institute for Health and Care Excellence (NICE), interests: #digitalhealth, #AI, #ethicsinSTEM. Not official work account.
Whenever we design a system we impose a power structure. We grant some rights and options and take away others through the choices we make. Advocate for the people not in the room.
It occurs to me one disadvantage of #MachineLearning is it's difficult to react to unexpected developments. Eg with #Coronavirus, it's easy to tell human clinicians to adjust their thinking/protocols, but much harder to quickly change a system that only learns from data.
It's an extraordinary coincidence. But for me, the most powerful bit of the original tv prog was this:
This is our 2018 simulation of the difference that everyone washing their hands 5-10 times a day might make to the spread of a flu-like virus.
See how much time it buys us?
12-18 months ago I suggested a publicly funded initiative to use AI to find a new antibiotic - obviously that didn't go anywhere - but I'm hugely relieved someone with some influence has done it!
BBC News - Scientists discover powerful antibiotic using AI https://t.co/404YHYWmFp
When a colleague casually asks what's the difference between satsumas, tangerines and clementines... Totally worth 15 mins of research. I missed a trick not making flavour intensity = bubble size though.
A computer scientist invented some new big data prediction techniques. Did a stint in industry.
Warned about the potential for algorithmic discrimination and data-driven monopoly power.
Sounds very 2020s.
But it's Charles Babbage, in the 1820s.
https://t.co/FCohWqVb6W
A new paper has been making the rounds with the intriguing claim that YouTube has a *de-radicalizing* influence. https://t.co/TTtWR0uBgi
Having read the paper, I wanted to call it wrong, but that would give the paper too much credit, because it is not even wrong. Let me explain.
'Recycle mode' that irreversibly deactivates a working product so it has to be traded in for a newer model, causing more waste than re-use.
#ethicsinSTEM
@JakeBerry If you selected 23 people at random from the UK population, the chance of drawing 23 white males is something like one in a hundred million (0.45^23). Just saying.
@bengoldacre Excellent work, thank you. Great to see this innovative, open approach allowing orgs of all sizes to compare their data to peers for NICE prescribing guidance.
HAPPY XMAS THE NHS
Under the tree we've made you tens of thousands of free dashboards for every GP, CCG, PCN and more, showing adherence to NICE guidance. If we can get more data, we will ship more tools!
https://t.co/en8MFPQLTM
1) In 2008 a salmon (yes the fish) was put through an MRI scanner. The salmon was shown a series of photographs of humans in social situations, and asked to determine the emotion of the person in the photo (yes, really).