Il Regno Unito sarà il primo paese a utilizzare il sistema sanitario AI per prevenire futuri scandali
https://www.lbc.co.uk/news/health/uk-to-be-first-country-to-use-ai-healthcare-system-to-prevent-future-scandals/
di tylerthe-theatre
Il Regno Unito sarà il primo paese a utilizzare il sistema sanitario AI per prevenire futuri scandali
https://www.lbc.co.uk/news/health/uk-to-be-first-country-to-use-ai-healthcare-system-to-prevent-future-scandals/
di tylerthe-theatre
28 commenti
As predicted, the UK will roll out AI and once integrated into back end systems will start using them for diagnostics. Then the real scandals will begin.
Healthcare run by fancy autocorrect to prevent scandals? Can’t see any issues there.
Notice there isn’t any information about who created the AI, how it was trained, what data it was trained on, or any information that would provide transparency to satisfy the public that the software they were introuducing is itself not potentially adding biases that could lead to discrimination or detrimental care.
AI in it’s current state is no better or worse than a human, it’s just as flawed and prone to dangerous bouts of stupidity, and to unleash it without publishing details of any guardrails will only end in tears.
More like “The UK will be the first country to have a scandal caused by using by AI.”
Instead of properly funding the nhs someone’s mate is going to get a suitcase of cash.
UK first country to be embroiled in scandal encountered by AI healthcare.
Like it does really feel like labours top brass are just impressionable middle managers who get swindled.
The use case they are giving here seems fine “flag up if theres an unusual amount of cases in X ward”, so its not like AI is replacing doctors; but uh, do you need presumably multi-million pound AI outsorced contract for that? or do you need a spreadsheet that you check once a week.
>”Technology will always have a role to play, but having the right number of staff on the front line of care is the place to start the investment to make patients safe.”
And this is is: the solution is more bodies, not burning cash on “trendy” systems
Next…a scandal related to AI making bad/wrong decisions…
I work on AI implementation in the NHS.
Of all the things we could do (and are doing and need to get done) this is the one that Streeting highlights. The ‘control’ system aimed at preventing bad news stories for our politicians more than anything.
That said, it hints at finally getting a national data-warehouse (or lake/lakehouse for the tech geeks here) for health data. And that is more than welcome, in fact, it is a must if we are to meet the ambition of using digital in healthcare to eradicate waiting lists and increase productivity.
NHS systems are poorly, if at all, integrated adding dodgy AI on top won’t do anything other than waste time and money that could be spent creating an integrated health system. AI is not needed for that.
Scandal centred around use of AI coming up in 3… 2… 1…
So, if anyone’s interested, here’s how this will play out.
1) huge amounts of money go in.
2) It’s doomed from the start: the leaders managing it don’t understand the science and hard problems with making it work, and those delivering it are disincentivised from raising red flags.
3) It will transpire that the AI doesn’t work well and saves no resources on existing processes.
4) Data quality will be blamed – the clinicians aren’t capturing all the data needed for the algorithm to work well at spotting issues, and so a comprehensive clinical audit style data collection will be needed.
5) Capturing the granular clinical data needed requires too much extra resource in the form of additional clinical time and investment in data platforms.
6) No money for this, all already spent on AI consultants and managers.
7) Whole thing is torpedoed but a note is made that investment in detailed and comprehensive classic clinical audit would be helpful in the future, if only there were money.
Yeah but AI bad so I am outraged at this, yes it may save lives, but AI bad, so I am not happy
the days of old fashioned scandals are over. make way for the days of newer, shinier, AI scandals
It’s almost a lose lose for them because this AI could diagnose 10,000 people correctly at 10x the speed of a doctor, and with a lower error rate, but it only has to get 1 wrong and people will lose their mind and want it banned.
No…no… fairly certain this will create more scandals.
This is a terrible idea, AI uses information which has already been created, the issue is there is still a massive gender gap in healthcare when men and women have different symptoms for the same illness because their bodies work different, based on the current information male bodies are considered the standard with there being decades more studies and researched done based on them in comparison to women.
Using AI and letting it be trained by the information already out there will further widen the gender gap in healthcare and will lead women to dying because they had a different symptom to their male counterparts. Before this is implemented the amount of research for women needs to increase by amount 200%
“Prevent future scandals” opens up a whole new branch of potential scandals
If the government didn’t have AI they’d have no intelligence
Fiver says the “AI” is just a load of IF statements.
This is the kind of stuff AI should be used for, for looking at data and spotting patterns. Should still be done with human supervision though.
AI intergration shall continue until morale improves
The last nation on earth whose government I would want to see spraffing about such and idea is the UK. They have such a great record on IT mega infrastructure project, over priced , outdated by deployment and sketchy functionality and security. But i’m sure a few will get wealthy out of it , with data breaches see everyone else’s personal medical data disappear into the ether of the scammerverse. This constant use of AI will make everything better is such a political sound bite, just like everyone use to add .com to the end of everything to sound tech savvy. AI might be a wonderful concept to those best positioned to make money from it, for the majority of the government advisors adding it to their day talks, I wouldn’t trust them to use alexa to place a correct amazon order.
Feel like its only a matter of time before we have the first Alexa Shipman case
Hhmm wonder if Palantir are involved at all? They have a data platform for NHS as far as I’m aware but have been/are some concerns about them given their history and origins..
Is this why loads of NHS admin staff were made redundant on Friday?
I guess the question is whether we’re more concerned over people’s health outcomes, or over any resulting political shitstorm.
I’d be happy for them to use AI to make people healthier. I’m less impressed if they used AI to make politicians look better.
Bet it won’t be AI and will be data analysis with AI being used as a buzz word.
AI should be defined as analysis of unstructured data. This sounds like an analysis of data stored in a database.
People need to stop thinking ChatGPT, it’s not that. It’s machine learning. Machine learning is already used in bio imaging (and everywhere else, frankly) and is a useful technology.