Andy
@andymcm.bsky.social
Hey hi 🙂. My things...AI, specialising in Mental Health and broader healthcare & Ethics AI, running, public & private healthcare & lover of lovely people 🥰
Reposted by Andy
Not exaggerating, look it up. We have to do an inservice every year about this. So…what the actual.
March 28, 2025 at 2:13 AM
Not exaggerating, look it up. We have to do an inservice every year about this. So…what the actual.
We are looking at a misdiagnosis and treatment future health catastrophe, like before, we do not learn....
An excellent contribution in the 'battle' against 'inappropriate' use of LMs and AI more broadly. The clear underlying issue is where used clinically and non clinically, in particular MH, the user simply doesn't understand, or maybe care, in how LMs work, it is seen as the panacea for pressure.
despite popularised beliefs, LLMs are not fit for medical applications. SoTA models produce "non-trivial levels of hallucinations" even w inference techniques like CoT & search augmented generation: arxiv.org/pdf/2503.05777
of surveyed clinicians, 53% use LLMs daily & 91% encountered hallucinations
of surveyed clinicians, 53% use LLMs daily & 91% encountered hallucinations
March 23, 2025 at 8:40 AM
We are looking at a misdiagnosis and treatment future health catastrophe, like before, we do not learn....
An excellent contribution in the 'battle' against 'inappropriate' use of LMs and AI more broadly. The clear underlying issue is where used clinically and non clinically, in particular MH, the user simply doesn't understand, or maybe care, in how LMs work, it is seen as the panacea for pressure.
despite popularised beliefs, LLMs are not fit for medical applications. SoTA models produce "non-trivial levels of hallucinations" even w inference techniques like CoT & search augmented generation: arxiv.org/pdf/2503.05777
of surveyed clinicians, 53% use LLMs daily & 91% encountered hallucinations
of surveyed clinicians, 53% use LLMs daily & 91% encountered hallucinations
March 23, 2025 at 8:38 AM
An excellent contribution in the 'battle' against 'inappropriate' use of LMs and AI more broadly. The clear underlying issue is where used clinically and non clinically, in particular MH, the user simply doesn't understand, or maybe care, in how LMs work, it is seen as the panacea for pressure.
Reposted by Andy
🔥 Patriotic Millionaires UK campaigned last week across the country to advocate for a Wealth Tax, against the backdrop of significant possible welfare cuts in the spring budget.
Patriotic Millionaires UK ‘Tax our Wealth’ campaign visits Bristol
Patriotic Millionaires UK campaigned last week across the country to advocate for a Wealth Tax, against the backdrop of significant possible welfare cuts in the spring budget
westenglandbylines.co.uk
March 21, 2025 at 7:12 AM
🔥 Patriotic Millionaires UK campaigned last week across the country to advocate for a Wealth Tax, against the backdrop of significant possible welfare cuts in the spring budget.
Reposted by Andy
Just a reminder...
March 18, 2025 at 8:50 PM
Just a reminder...
Reposted by Andy
these are the kind of moves that'll eventually move the needle towards shifting power away from tech oligarchs
💥The Netherlands' largest pension fund, has sold its shares in Tesla, Meta, and Alphabet (Google). The fund with US $569 billion equivolent in assets is "directing more money to companies and projects that help to improve society and the environment." #Divest
www.reuters.com/markets/euro...
www.reuters.com/markets/euro...
Dutch pension fund ABP sells all Alphabet and Meta shares, CEO tells newspaper FD
ABP, the Netherlands' largest pension fund, has sold its shares in tech companies Meta and Alphabet , its CEO told Dutch newspaper FD in an interview published on Monday.
www.reuters.com
February 23, 2025 at 9:43 AM
these are the kind of moves that'll eventually move the needle towards shifting power away from tech oligarchs
Reposted by Andy
“Call it democracy, or call it democratic socialism, but there must be a better distribution of wealth within this country for all God’s children.”
– MLK Speech to the Negro American Labor Council, 1961. #MLKDay2025
– MLK Speech to the Negro American Labor Council, 1961. #MLKDay2025
January 20, 2025 at 6:21 PM
“Call it democracy, or call it democratic socialism, but there must be a better distribution of wealth within this country for all God’s children.”
– MLK Speech to the Negro American Labor Council, 1961. #MLKDay2025
– MLK Speech to the Negro American Labor Council, 1961. #MLKDay2025
Reposted by Andy
👇👏
Focus on the problem, not the solution!
If you are struggling to figure out where AI fits in your business strategy, first, you have to figure out the real problem you want to solve and if AI is the solution.
juanpml.com/writing/2024...
If you are struggling to figure out where AI fits in your business strategy, first, you have to figure out the real problem you want to solve and if AI is the solution.
juanpml.com/writing/2024...
https://juanpml.com/writing/2024/1…
December 22, 2024 at 6:42 PM
👇👏
Reposted by Andy
Why is it when we see issues within some Muslim communities, the narrative becomes there’s something wrong with Islam? Yet this week we have heard of some of the worst child abuse scandals at the highest level of the English church. Are we hearing about that being a problem with Christianity?
December 19, 2024 at 9:41 AM
Why is it when we see issues within some Muslim communities, the narrative becomes there’s something wrong with Islam? Yet this week we have heard of some of the worst child abuse scandals at the highest level of the English church. Are we hearing about that being a problem with Christianity?
Reposted by Andy
Reposted by Andy
New Christmas lights just went up in Clacton.
(Farage is still the MP who makes most money from second jobs. By a long way.)
(Farage is still the MP who makes most money from second jobs. By a long way.)
December 17, 2024 at 5:05 PM
New Christmas lights just went up in Clacton.
(Farage is still the MP who makes most money from second jobs. By a long way.)
(Farage is still the MP who makes most money from second jobs. By a long way.)
Reposted by Andy
Something I think is really important to note about this story is how a true fandom community around school shooters and other young mass killers has flourished on CharacterAI -- much like these fandoms and groups have on other platforms like Tumblr/Discord/4chan
NEW: CharacterAI is hosting many popular chatbots that emulate — and often extol — real school shooters like Adam Lanza (Sandy Hook) and Harris/Klebold (Columbine.)
The site is also hosting chatbots that impersonate real child and teenage victims of school shootings.
futurism.com/character-ai...
The site is also hosting chatbots that impersonate real child and teenage victims of school shootings.
futurism.com/character-ai...
December 17, 2024 at 4:57 PM
Something I think is really important to note about this story is how a true fandom community around school shooters and other young mass killers has flourished on CharacterAI -- much like these fandoms and groups have on other platforms like Tumblr/Discord/4chan
Reposted by Andy
Once again, leaving this here: futurism.com/character-ai...
December 17, 2024 at 8:34 PM
Once again, leaving this here: futurism.com/character-ai...
On a serious point, how much better is this than the use of real animals in PG Tips, Andrew, Amazon ads & the list goes on. Animals get to do animal things and we all get cooler ads, everyone's a winner.
An otter travelling on a plane, sat up with a tablet, typing away....I mean....only CGI or AI..
An otter travelling on a plane, sat up with a tablet, typing away....I mean....only CGI or AI..
Okay, veo2 from Google is pretty impressive. "Otter on a plane using wifi" produced both of these in the first batch of four (the others weren't bad either). Love the way they tap the screen.
Sora offers a lot more control options & longer clips, so it is hard to compare, but they are the big 2 now
Sora offers a lot more control options & longer clips, so it is hard to compare, but they are the big 2 now
December 17, 2024 at 11:00 AM
On a serious point, how much better is this than the use of real animals in PG Tips, Andrew, Amazon ads & the list goes on. Animals get to do animal things and we all get cooler ads, everyone's a winner.
An otter travelling on a plane, sat up with a tablet, typing away....I mean....only CGI or AI..
An otter travelling on a plane, sat up with a tablet, typing away....I mean....only CGI or AI..
Reposted by Andy
There are some humdingers being played on Two Words at the moment. Thanks to everyone who’s given the game a go. We really appreciate your comments and are listening to all of them.
twowords2.page.link/play
twowords2.page.link/play
December 14, 2024 at 9:12 AM
There are some humdingers being played on Two Words at the moment. Thanks to everyone who’s given the game a go. We really appreciate your comments and are listening to all of them.
twowords2.page.link/play
twowords2.page.link/play
When you think you are reading @theonion.com
'OpenAI's new ChatGPT o1 model will try to escape if it thinks it'll be shut down'
A factor of a race to the bottom? equally for users who don't understand LLM limitations add this to inherent hallucinations and we have 😱
www.msn.com/en-us/news/t...
'OpenAI's new ChatGPT o1 model will try to escape if it thinks it'll be shut down'
A factor of a race to the bottom? equally for users who don't understand LLM limitations add this to inherent hallucinations and we have 😱
www.msn.com/en-us/news/t...
MSN
www.msn.com
December 10, 2024 at 7:51 PM
When you think you are reading @theonion.com
'OpenAI's new ChatGPT o1 model will try to escape if it thinks it'll be shut down'
A factor of a race to the bottom? equally for users who don't understand LLM limitations add this to inherent hallucinations and we have 😱
www.msn.com/en-us/news/t...
'OpenAI's new ChatGPT o1 model will try to escape if it thinks it'll be shut down'
A factor of a race to the bottom? equally for users who don't understand LLM limitations add this to inherent hallucinations and we have 😱
www.msn.com/en-us/news/t...
Reposted by Andy
NEW: CharacterAI is hosting chatbots designed to engage users in roleplay about self-harm, depicting graphic scenarios and even sharing tips for hiding self-injury from others.
The characters are widely accessible to minor users.
(TW for depictions of self-harm)
futurism.com/ai-chatbots-...
The characters are widely accessible to minor users.
(TW for depictions of self-harm)
futurism.com/ai-chatbots-...
December 7, 2024 at 6:27 PM
NEW: CharacterAI is hosting chatbots designed to engage users in roleplay about self-harm, depicting graphic scenarios and even sharing tips for hiding self-injury from others.
The characters are widely accessible to minor users.
(TW for depictions of self-harm)
futurism.com/ai-chatbots-...
The characters are widely accessible to minor users.
(TW for depictions of self-harm)
futurism.com/ai-chatbots-...
Hopefully we are at the end of the too long period of corruption, of only criminals could be held to accountp
November 26, 2024 at 8:15 AM
Hopefully we are at the end of the too long period of corruption, of only criminals could be held to accountp
Reposted by Andy
On Elon Musk’s repeated attacks on Keir Starmer’s government Sky’s @samcoatessky.bsky.social rightly points out
“If we were talking about a figure close to the Kremlin or the government in Moscow we would be asking the question whether this was foreign interference”
@sophyridge.bsky.social
“If we were talking about a figure close to the Kremlin or the government in Moscow we would be asking the question whether this was foreign interference”
@sophyridge.bsky.social
November 25, 2024 at 8:24 PM
On Elon Musk’s repeated attacks on Keir Starmer’s government Sky’s @samcoatessky.bsky.social rightly points out
“If we were talking about a figure close to the Kremlin or the government in Moscow we would be asking the question whether this was foreign interference”
@sophyridge.bsky.social
“If we were talking about a figure close to the Kremlin or the government in Moscow we would be asking the question whether this was foreign interference”
@sophyridge.bsky.social
Having worked in a number of health systems, yes the US hospitals are nice and shiny with lots of the latest kit...there is a reason for that...the US Insurance, Pharma, pre exclusion model is nothing short of abhorrent....Healthcare should be equally avaliable to all....
November 24, 2024 at 7:18 PM
Having worked in a number of health systems, yes the US hospitals are nice and shiny with lots of the latest kit...there is a reason for that...the US Insurance, Pharma, pre exclusion model is nothing short of abhorrent....Healthcare should be equally avaliable to all....
Reposted by Andy
Artificial Intelligence / Sztuczna Inteligencja
autor: Asier Sanz
autor: Asier Sanz
November 20, 2024 at 8:39 PM
Artificial Intelligence / Sztuczna Inteligencja
autor: Asier Sanz
autor: Asier Sanz
Reposted by Andy
Good morning world ☀️
November 22, 2024 at 8:55 AM
Good morning world ☀️