Manifesting a routine with our 12-week program that will help you find a way to reduce symptoms instead of off-the-shelf products left in kitchen cupboards gathering dust.
π Belfast | 30 September 2026 Black Mountain Research Initiative Β· 00AI
π **The whitepaper comes first.**
It holds the full argument, sources, context and limits. It gives readers room to examine what is supported, what remains uncertain and what needs further investigation.
The 00AI cards sit beside that work, helping people find their way into it.
That is the approach we want across all our whitepapers: depth in the paper, accessible entry points in the cards.
π **People need information they can reach**
A parent may be reading between appointments.
A carer may have only a few quiet minutes.
Someone with dyslexia may find a short visual explanation easier to begin with. Another reader may be interested but unfamiliar with the subject.
When information is difficult to approach, valuable questions can go unread.
The cards give people somewhere to startβand somewhere to go next.
π **One card. One clear question.**
A useful 00AI card brings forward a key idea from the paper.
It can help readers ask:
β’ What is the claim? β’ Where did it come from? β’ What does the evidence support? β’ What are the limits? β’ Where can I read the fuller explanation?
Each card should remain faithful to its whitepaper. A striking image must carry careful wording.
Readers should be able to move from a short message back to the material behind it.
The card introduces the question. The paper explains it. The source allows it to be checked.
That connection is what makes the cards useful alongside every whitepaper.
π **WP5: Charity Opens the Door**
Our latest set accompanies **βCharity Opens the Door β Working Paper 5.β**
The paper explores charity frameworks, public benefit and who gets a seat at the table.
Its eight 00AI cards introduce themes including listening to families, keeping different sources clearly labelled, checking before sharing and improving our work openly.
The cards invite readers into those questions. The whitepaper provides the detail and makes its limits clear, including that it alleges no fraud or crime.
π **Why the papers and cards must keep improving**
As we recheck sources and clarify our understanding, both formats need attention.
A revised paper may require a revised card. A shortened sentence must still preserve the meaning of the evidence.
Upgrading means checking, clarifying and correctingβnot simply changing the appearance.
π€ **AI assists. Human judgement remains essential.**
AI can help organise research and make complex wording easier to approach.
Every output still needs review. We remain responsible for checking sources, preserving context and explaining uncertainty.
The agents are our storytelling device. The evidence carries the argument.
π **Our commitment**
Put the whitepaper first.
Place accessible cards beside it.
Keep the route to the sources visible.
Give people the opportunity to understand, question and contribute.
**The paper provides depth. The cards provide access. Human judgement remains essential.**
From Belfast, with loveβand evidence people can reach. π
**From Belfast with love β and questions that still need answers. β€οΈ**
Publishing a whitepaper is a starting point.
Evidence develops. Sources become available. Questions sharpen. Sometimes an earlier statement needs correcting.
Our papers must reflect that.
For the Black Mountain Research Initiative, updating the whitepapers means making our work clearer, easier to check and more useful to the people it concerns.
π **An upgrade must strengthen the evidence**
A new cover or faster AI tool cannot answer an unresolved question.
A meaningful update should help readers see:
β’ Where a claim came from. β’ What the evidence supports. β’ What remains uncertain. β’ What has changed since the previous version. β’ What needs testing next.
If we ask organisations to be transparent, our own work must be open to the same scrutiny.
Our updated working paper, **βRed September β Red Flags,β** carries a simple rule:
**A red flag is a question. A question must meet a source.**
Raising a concern creates a responsibility to investigate it.
We need to examine consent, evidence quality, possible harm and what happens when an intervention does not work. These questions apply across behavioural, educational, medical, nutritional and AI-assisted approaches β including our own proposals.
Families deserve clear explanations of what is known and what is still being explored.
π€ **AI helps us work faster. Human judgement still matters.**
AI can help organise material, compare documents and identify questions.
We still need to read the original sources, check the context and take responsibility for what we publish.
The important question is: **what improved for children and families?**
π **Why the 00AI cards sit beside the papers**
Long documents can be difficult to approach, particularly when someone is exhausted, overwhelmed or dyslexic.
The cards offer a short, visual route into the work:
Because research should never become a snapshot that we are afraid to revisit.
Today we are sharing **WP3 β *Breaking the Loops: Gut health, ethical AI and family support in Northern Ireland*.**
It takes an article first published on **28 October 2025** and brings it forward into our current Black Mountain Research Initiative working-paper framework. :chatgpt-content-reference{index="1"}
But why go back?
Why not simply publish something new?
Because **going back is part of the research.** π
New evidence appears. New official figures are published. Technology changes. Services change. And sometimes a claim that looked simple becomes much more complicated when we examine the primary sources.
That means responsible research has to be capable of saying:
**βHere is what we said then. Here is what the evidence supports now. Here is what remains uncertain.β**
That experience matters. It can identify problems, gaps and questions that statistics alone may never reveal.
But lived experience and scientific evidence are not interchangeable.
The working paper therefore draws an important line:
**one family can open a research door β it cannot close the science.** :chatgpt-content-reference{index="2"}
### π¦ The gut-health evidence needed upgrading too
The original article referred to an **βup to 70%β** figure for gastrointestinal problems among autistic children.
WP3 goes back to the published literature and shows why a single headline number is inadequate.
The research contains a **wide range of estimates**, with pooled estimates often around **49β55%**. It supports taking gastrointestinal problems seriously, while **not establishing that GI symptoms cause autism or that one intervention works for everyone**. :chatgpt-content-reference{index="3"}
That distinction matters.
**Association β causation.**
And upgrading a claim when the evidence requires greater precision is not a weakness.
### **It is what evidence-led work is supposed to do.** π§ͺ
### π Northern Ireland's numbers need their labels
The same principle applies to statistics.
WP3 keeps population estimates, School Census figures and assessment waiting-list figures separate because they describe **different populations and different questions**.
The paper records the **2025/26 School Census figure of 18,472 school-age children (6.2%)**, while separately referencing **17,205 children waiting for ASD assessment** as reported on 8 October 2025. :chatgpt-content-reference{index="4"}
Both numbers can matter.
But putting two different measures together as though they mean the same thing does not make the evidence stronger.
**Keeping the layers separate does.**
### π€ And then there is AI
AI is developing too quickly for yesterday's assumptions simply to be carried forward.
Our position in WP3 is deliberately straightforward:
**AI supports. It does not replace care.**
The paper preserves questions about tools offered without an adequate evidence base, automated βassessmentβ products that bypass human judgement, and marketing that promises families a βfix.β :chatgpt-content-reference{index="5"}
Technology can be extraordinary.
But extraordinary technology still needs evidence, transparency and human accountability.
---
## π± THAT IS WHY WE NEED THE UPGRADES
Not to rewrite history.
**To document it.**
Not to make every early idea look correct.
**To test it.**
Not to turn lived experience into clinical proof.
**To allow lived experience to generate better research questions.**
And not to pretend that a working paper is ever the final word.
If stronger evidence appears tomorrow, we should look at it.
If an official figure changes, we should update it.
If something we believed needs qualification, we should qualify it.
And if the evidence does **not** support a claim, we should be prepared to say so.
That is the standard we are trying to build into the **Black Mountain Research Initiative β’ 00AI Series**.
WP1 asked what people actually find when they search.
WP2 asked what happens when evidence exists but people cannot reach it.
WP3 asks what happens when **lived experience, gutβbrain research, Northern Ireland data, ethical AI and family support are put into the same evidence trail without pretending they are all the same kind of evidence.** :chatgpt-content-reference{index="11"}
And there will be more questions.
That is not a weakness in the project.
### **That is the project.**
Ask. β Listen. π Find the source. π Test the evidence. π§ͺ Separate what we know from what we suspect. βοΈ Publish. π Then come back and check again. π
Because evidence changes.
Technology changes.
Services change.
Families' needs change.
And if the evidence changes, **we should be prepared to change with it.**
π THE EVIDENCE IS NO USE IF PEOPLE CANNOT REACH IT
Why the 00AI cards sit beside the first paper
Nothing stands still.
Not technology.
Not research.
Not public policy.
Not the information people find when they search.
And certainly not AI.
That is why the 00AI cards have been updated for 2026.
The original cards helped us start a conversation.
The new cards ask a harder question:
Have we upgraded our understanding as quickly as we have upgraded our tools?
AI can now search, compare, summarise and produce material at extraordinary speed.
But speed is not the same thing as understanding.
And a confident answer is not necessarily a correct one.
That matters particularly when AI touches health information, children and families, public services, education, personal data or decisions that can affect people's lives.
βοΈ DO WE UNDERSTAND THE LAW AROUND AI?
That question needs to become part of AI literacy.
In the UK, there is not simply one rule called βAI lawβ that answers everything. Existing legal duties can apply depending on what an AI system does, what information it processes and the context in which it is used.
For example, where AI processes personal data, UK data-protection law applies. The ICO highlights lawfulness, fairness, transparency, purpose limitation, data minimisation, accuracy, security and accountability, alongside people's rights. ICO
Fairness is particularly important. AI can reproduce or amplify problems arising from training data, assumptions, design choices and the environment in which a system operates. The ICO therefore stresses that fairness cannot be reduced to a mathematical measure inside an algorithm; context and effects on people matter too. ICO
And when AI contributes to decisions about people, the principles of transparency and accountability matter. People may need meaningful information about how their information is being used and how an AI-assisted decision was reached. ICO
Even the guidance itself does not stand still: the ICO currently notes that parts of its AI guidance are being reviewed following changes made by the Data (Use and Access) Act. ICO
That is exactly the point.
π IF THE SYSTEM CHANGES, WE HAVE TO KEEP CHECKING
The 00AI method cannot be:
We researched it once, therefore we know it forever.
It has to be:
What has changed?
What does the current source say?
What does the law require now?
What guidance has changed?
What evidence has appeared?
What did we miss?
And then the uncomfortable question:
Were we wrong?
If better evidence changes the picture, we change the card.
That is not failure.
That is the method working.
π€ AI SHOULD HELP US QUESTION β NOT STOP US QUESTIONING
Roger/ChatGPT and Sean/Grok remain beside me in the updated cards for a reason.
Different systems can help search, compare, challenge and expose gaps.
But neither AI gets the final word.
Neither does a government document.
Neither does a charity.
Neither does a research paper merely because it appears authoritative.
And neither do we.
The new cards therefore return repeatedly to the same discipline:
Ascension Autism
π Belfast | 30 September 2026
Black Mountain Research Initiative Β· 00AI
π **The whitepaper comes first.**
It holds the full argument, sources, context and limits. It gives readers room to examine what is supported, what remains uncertain and what needs further investigation.
The 00AI cards sit beside that work, helping people find their way into it.
That is the approach we want across all our whitepapers: depth in the paper, accessible entry points in the cards.
π **People need information they can reach**
A parent may be reading between appointments.
A carer may have only a few quiet minutes.
Someone with dyslexia may find a short visual explanation easier to begin with. Another reader may be interested but unfamiliar with the subject.
When information is difficult to approach, valuable questions can go unread.
The cards give people somewhere to startβand somewhere to go next.
π **One card. One clear question.**
A useful 00AI card brings forward a key idea from the paper.
It can help readers ask:
β’ What is the claim?
β’ Where did it come from?
β’ What does the evidence support?
β’ What are the limits?
β’ Where can I read the fuller explanation?
Each card should remain faithful to its whitepaper. A striking image must carry careful wording.
π **The route back matters**
**WHITEPAPER β CARD β QUESTION β SOURCE β PRIMARY EVIDENCE β CONTEXT β PUBLIC UNDERSTANDING**
Readers should be able to move from a short message back to the material behind it.
The card introduces the question. The paper explains it. The source allows it to be checked.
That connection is what makes the cards useful alongside every whitepaper.
π **WP5: Charity Opens the Door**
Our latest set accompanies **βCharity Opens the Door β Working Paper 5.β**
The paper explores charity frameworks, public benefit and who gets a seat at the table.
Its eight 00AI cards introduce themes including listening to families, keeping different sources clearly labelled, checking before sharing and improving our work openly.
The cards invite readers into those questions. The whitepaper provides the detail and makes its limits clear, including that it alleges no fraud or crime.
π **Why the papers and cards must keep improving**
As we recheck sources and clarify our understanding, both formats need attention.
A revised paper may require a revised card. A shortened sentence must still preserve the meaning of the evidence.
Upgrading means checking, clarifying and correctingβnot simply changing the appearance.
π€ **AI assists. Human judgement remains essential.**
AI can help organise research and make complex wording easier to approach.
Every output still needs review. We remain responsible for checking sources, preserving context and explaining uncertainty.
The agents are our storytelling device. The evidence carries the argument.
π **Our commitment**
Put the whitepaper first.
Place accessible cards beside it.
Keep the route to the sources visible.
Give people the opportunity to understand, question and contribute.
**The paper provides depth.
The cards provide access.
Human judgement remains essential.**
From Belfast, with loveβand evidence people can reach. π
#00AI #Whitepaper #EvidenceMatters #AccessibleInformation #ResponsibleAI #CriticalThinking #ParentLed #PublicBenefit #NorthernIreland #GutHealthMatters #ChildrenFirst
14 hours ago | [YT] | 1
View 0 replies
Ascension Autism
π Why the 00AI cards sit beside the paper
lnkd.in/p/eg7tg8MY
Many people want to understand these issues but cannot face another dense document.
Dyslexia, caring responsibilities, stress and exhaustion can make important information difficult to reach.
The cards offer an entry point:
WHITEPAPER β CARD β QUESTION β SOURCE β PRIMARY EVIDENCE β CONTEXT β PUBLIC UNDERSTANDING
The paper provides depth.
The cards provide access.
The sources allow the reader to check.
20 hours ago | [YT] | 1
View 0 replies
Ascension Autism
π© **RED SEPTEMBER: The world keeps updating. Our whitepapers must too.**
Evidence develops. Questions sharpen. Our work must keep pace. π
Updating our whitepapers means checking sources, correcting gaps and showing clearly what we know β and what still needs testing.
π€ AI helps us research. Human judgement and accountability remain essential.
π Our 00AI cards make complex papers easier to approach, one question at a time.
β€οΈ Children and families deserve evidence they can reach, understand and challenge.
Read the article π
**Data leads. Evidence tests. Childrenβs welfare decides.**
From Belfast with love. β€οΈ
#RedSeptember #00AI #EvidenceMatters #ResponsibleAI #ChildrenFirst #Neurodiversity #NorthernIreland #Belfast
23 hours ago | [YT] | 1
View 0 replies
Ascension Autism
# π© Red September: The World Keeps Updating. Our Whitepapers Must Too.
**From Belfast with love β and questions that still need answers. β€οΈ**
Publishing a whitepaper is a starting point.
Evidence develops. Sources become available. Questions sharpen. Sometimes an earlier statement needs correcting.
Our papers must reflect that.
For the Black Mountain Research Initiative, updating the whitepapers means making our work clearer, easier to check and more useful to the people it concerns.
π **An upgrade must strengthen the evidence**
A new cover or faster AI tool cannot answer an unresolved question.
A meaningful update should help readers see:
β’ Where a claim came from.
β’ What the evidence supports.
β’ What remains uncertain.
β’ What has changed since the previous version.
β’ What needs testing next.
If we ask organisations to be transparent, our own work must be open to the same scrutiny.
π© **Red flags need follow-through**
Our updated working paper, **βRed September β Red Flags,β** carries a simple rule:
**A red flag is a question.
A question must meet a source.**
Raising a concern creates a responsibility to investigate it.
We need to examine consent, evidence quality, possible harm and what happens when an intervention does not work. These questions apply across behavioural, educational, medical, nutritional and AI-assisted approaches β including our own proposals.
Families deserve clear explanations of what is known and what is still being explored.
π€ **AI helps us work faster. Human judgement still matters.**
AI can help organise material, compare documents and identify questions.
We still need to read the original sources, check the context and take responsibility for what we publish.
The important question is: **what improved for children and families?**
π **Why the 00AI cards sit beside the papers**
Long documents can be difficult to approach, particularly when someone is exhausted, overwhelmed or dyslexic.
The cards offer a short, visual route into the work:
**WHITEPAPER β CARD β QUESTION β SOURCE β EVIDENCE β UNDERSTANDING**
Each card opens a question. The paper provides the detail needed to examine it.
When the paper changes, the cards must be reviewed too.
π· **Projections also need checking**
Cost models can help us explore possible futures.
Their assumptions, datasets and uncertainties must be visible. A projected saving remains a projection until outcomes have been measured.
β€οΈ **Children first means keeping the work accountable**
Every update should make it easier to understand our reasoning, challenge our conclusions and identify the next useful step.
That is why we keep upgrading.
**Data leads. Evidence tests. Childrenβs welfare decides.**
π Belfast
ποΈ 30 September 2026
#RedSeptember #00AI #EvidenceMatters #ResponsibleAI #ChildrenFirst #Autism #Neurodiversity #Transparency #NorthernIreland #Belfast #BlackMountainResearchInitiative
1 day ago | [YT] | 1
View 0 replies
Ascension Autism
1 day ago | [YT] | 1
View 1 reply
Ascension Autism
π WHY DO WE KEEP UPGRADING THE WORK?
lnkd.in/p/e3iKtbhj
lnkd.in/p/eSnBVGRX
lnkd.in/p/ez9E6QXM
lnkd.in/p/ewUpS_qs
lnkd.in/p/eSnBVGRX
lnkd.in/p/etRZmjgT
Because research should never become a snapshot that we are afraid to revisit.
Today we are sharing **WP3 β *Breaking the Loops: Gut health, ethical AI and family support in Northern Ireland*.**
It takes an article first published on **28 October 2025** and brings it forward into our current Black Mountain Research Initiative working-paper framework. :chatgpt-content-reference{index="1"}
But why go back?
Why not simply publish something new?
Because **going back is part of the research.** π
New evidence appears.
New official figures are published.
Technology changes.
Services change.
And sometimes a claim that looked simple becomes much more complicated when we examine the primary sources.
That means responsible research has to be capable of saying:
**βHere is what we said then. Here is what the evidence supports now. Here is what remains uncertain.β**
### π§© Lived experience starts the question
WP3 begins with a family story.
That experience matters. It can identify problems, gaps and questions that statistics alone may never reveal.
But lived experience and scientific evidence are not interchangeable.
The working paper therefore draws an important line:
**one family can open a research door β it cannot close the science.** :chatgpt-content-reference{index="2"}
### π¦ The gut-health evidence needed upgrading too
The original article referred to an **βup to 70%β** figure for gastrointestinal problems among autistic children.
WP3 goes back to the published literature and shows why a single headline number is inadequate.
The research contains a **wide range of estimates**, with pooled estimates often around **49β55%**. It supports taking gastrointestinal problems seriously, while **not establishing that GI symptoms cause autism or that one intervention works for everyone**. :chatgpt-content-reference{index="3"}
That distinction matters.
**Association β causation.**
And upgrading a claim when the evidence requires greater precision is not a weakness.
### **It is what evidence-led work is supposed to do.** π§ͺ
### π Northern Ireland's numbers need their labels
The same principle applies to statistics.
WP3 keeps population estimates, School Census figures and assessment waiting-list figures separate because they describe **different populations and different questions**.
The paper records the **2025/26 School Census figure of 18,472 school-age children (6.2%)**, while separately referencing **17,205 children waiting for ASD assessment** as reported on 8 October 2025. :chatgpt-content-reference{index="4"}
Both numbers can matter.
But putting two different measures together as though they mean the same thing does not make the evidence stronger.
**Keeping the layers separate does.**
### π€ And then there is AI
AI is developing too quickly for yesterday's assumptions simply to be carried forward.
Our position in WP3 is deliberately straightforward:
**AI supports. It does not replace care.**
The paper preserves questions about tools offered without an adequate evidence base, automated βassessmentβ products that bypass human judgement, and marketing that promises families a βfix.β :chatgpt-content-reference{index="5"}
Technology can be extraordinary.
But extraordinary technology still needs evidence, transparency and human accountability.
---
## π± THAT IS WHY WE NEED THE UPGRADES
Not to rewrite history.
**To document it.**
Not to make every early idea look correct.
**To test it.**
Not to turn lived experience into clinical proof.
**To allow lived experience to generate better research questions.**
And not to pretend that a working paper is ever the final word.
If stronger evidence appears tomorrow, we should look at it.
If an official figure changes, we should update it.
If something we believed needs qualification, we should qualify it.
And if the evidence does **not** support a claim, we should be prepared to say so.
That is the standard we are trying to build into the **Black Mountain Research Initiative β’ 00AI Series**.
π **Find the claim.**
π **Find the source.**
π§ͺ **Test the evidence.**
π·οΈ **Label the evidence properly.**
π **Upgrade when the record changes.**
π¨βπ©βπ§βπ¦ **Keep children and families at the centre.**
Because the purpose isn't to prove that we were right yesterday.
### **The purpose is to get closer to the evidence today β so we can ask better questions tomorrow.**
π **Real people. Real evidence. Better questions. Stronger services.**
**WP3 β BREAKING THE LOOPS**
*Gut health, ethical AI and family support in Northern Ireland*
**BLACK MOUNTAIN RESEARCH INITIATIVE β’ 00AI SERIES**
**Data leads. Evidence tests. Childrenβs welfare decides.**
#BlackMountainResearchInitiative #00AI #BreakingTheLoops #WP3 #NorthernIreland #Belfast #LivedExperience #EvidenceMatters #EvidenceBased #ResearchTransparency #GutHealth #GutBrainAxis #Microbiome #AutismResearch #AutismNI #Neurodevelopment #EthicalAI #ResponsibleAI #AIInHealthcare #FamilySupport #ChildrensWelfare #BetterServices #PublicHealth #CommunityResearch #OpenEvidence #RealPeople #BetterQuestions #StrongerFamilies #StrongerCommunities #DataLeads #EvidenceTests
2 days ago | [YT] | 1
View 0 replies
Ascension Autism
π So we keep upgrading
WP1 asked what people actually find when they search.
WP2 asked what happens when evidence exists but people cannot reach it.
WP3 asks what happens when **lived experience, gutβbrain research, Northern Ireland data, ethical AI and family support are put into the same evidence trail without pretending they are all the same kind of evidence.** :chatgpt-content-reference{index="11"}
And there will be more questions.
That is not a weakness in the project.
### **That is the project.**
Ask. β
Listen. π
Find the source. π
Test the evidence. π§ͺ
Separate what we know from what we suspect. βοΈ
Publish. π
Then come back and check again. π
Because evidence changes.
Technology changes.
Services change.
Families' needs change.
And if the evidence changes, **we should be prepared to change with it.**
π **Real people.**
π¬ **Real evidence.**
π€ **Better questions.**
π± **Stronger services.**
π¨βπ©βπ§βπ¦ **Children and families at the centre.**
**BLACK MOUNTAIN RESEARCH INITIATIVE β’ 00AI SERIES**
**Data leads. Evidence tests. Childrenβs welfare decides.**
### Hashtags
#BlackMountainResearchInitiative #00AI #BreakingTheLoops #WP3 #NorthernIreland #Belfast #AutismResearch #AutismNI #GutBrainAxis #GutHealth #Microbiome #LivedExperience #EvidenceBased #EvidenceMatters #EthicalAI #ResponsibleAI #AIandHealth #FamilySupport #ChildrensWelfare #Neurodevelopment #PublicHealth #HealthResearch #ResearchTransparency #OpenEvidence #BetterServices #CommunityResearch #RealPeopleRealStories #StrongerCommunities #DataLeads #EvidenceTests
2 days ago | [YT] | 0
View 0 replies
Ascension Autism
What is going on on the island
It's very raw for many after 30 years
praying
for all the families and children
m
2 days ago | [YT] | 1
View 0 replies
Ascension Autism
π THE EVIDENCE IS NO USE IF PEOPLE CANNOT REACH IT
Why the 00AI cards sit beside the first paper
Nothing stands still.
Not technology.
Not research.
Not public policy.
Not the information people find when they search.
And certainly not AI.
That is why the 00AI cards have been updated for 2026.
The original cards helped us start a conversation.
The new cards ask a harder question:
Have we upgraded our understanding as quickly as we have upgraded our tools?
AI can now search, compare, summarise and produce material at extraordinary speed.
But speed is not the same thing as understanding.
And a confident answer is not necessarily a correct one.
That matters particularly when AI touches health information, children and families, public services, education, personal data or decisions that can affect people's lives.
βοΈ DO WE UNDERSTAND THE LAW AROUND AI?
That question needs to become part of AI literacy.
In the UK, there is not simply one rule called βAI lawβ that answers everything. Existing legal duties can apply depending on what an AI system does, what information it processes and the context in which it is used.
For example, where AI processes personal data, UK data-protection law applies. The ICO highlights lawfulness, fairness, transparency, purpose limitation, data minimisation, accuracy, security and accountability, alongside people's rights. ICO
Fairness is particularly important. AI can reproduce or amplify problems arising from training data, assumptions, design choices and the environment in which a system operates. The ICO therefore stresses that fairness cannot be reduced to a mathematical measure inside an algorithm; context and effects on people matter too. ICO
And when AI contributes to decisions about people, the principles of transparency and accountability matter. People may need meaningful information about how their information is being used and how an AI-assisted decision was reached. ICO
Even the guidance itself does not stand still: the ICO currently notes that parts of its AI guidance are being reviewed following changes made by the Data (Use and Access) Act. ICO
That is exactly the point.
π IF THE SYSTEM CHANGES, WE HAVE TO KEEP CHECKING
The 00AI method cannot be:
We researched it once, therefore we know it forever.
It has to be:
What has changed?
What does the current source say?
What does the law require now?
What guidance has changed?
What evidence has appeared?
What did we miss?
And then the uncomfortable question:
Were we wrong?
If better evidence changes the picture, we change the card.
That is not failure.
That is the method working.
π€ AI SHOULD HELP US QUESTION β NOT STOP US QUESTIONING
Roger/ChatGPT and Sean/Grok remain beside me in the updated cards for a reason.
Different systems can help search, compare, challenge and expose gaps.
But neither AI gets the final word.
Neither does a government document.
Neither does a charity.
Neither does a research paper merely because it appears authoritative.
And neither do we.
The new cards therefore return repeatedly to the same discipline:
SEARCH β COMPARE β VERIFY β READ β CONTEXTUALISE β DECIDE.
The working paper puts it even more simply:
The whitepaper carries the evidence.
The cards open the door to it.
AI assists the search.
People still decide. 00AI_Cards_Matter_Whitepaper_20β¦
βοΈ WHY KEEP THE CARDS?
Because evidence that nobody can reach has limited public value.
A visual card can make someone stop.
A question can make them search.
A source can take them to the evidence.
And the evidence may change what they think.
That is the job of the updated 00AI cards.
Not to tell people what to believe.
To make it easier to ask why they believe it.
Nothing stands still.
So neither should our research.
Neither should our methods.
And neither should our understanding of AI.
Upgrade the technology.
Upgrade the evidence.
Upgrade the questions.
And keep checking us too.
BLACK MOUNTAIN RESEARCH INITIATIVE β’ 00AI
Data leads.
Evidence tests.
People decide.
#00AI #BlackMountainResearchInitiative #EvidenceMatters #ArtificialIntelligence #AI #AILiteracy #AIGovernance #DataProtection #UKGDPR #ResponsibleAI #Transparency #Accountability #CheckTheSource #Research #NorthernIreland #Belfast #WestBelfast #AutismResearch #LivedExperience #PeopleDecide
3 days ago | [YT] | 1
View 0 replies
Ascension Autism
new systems
thats
what it takes
m
3 days ago | [YT] | 1
View 0 replies
Load more