The soft sell is in play.
The problem isn't AI. It's the deployment of it by the highest risk group to use it. I've been investigating this quietly for months. Today I want to start talking.
I've been quiet on here for a few weeks. Working. Thinking. Beginning to document something I couldn't quite name yet.
The pace of events in the UK has changed. I've been watching it. And this week, two things happened in close succession that made me want to start talking.
The Blair intervention.
On 26 May, Tony Blair published a 5,600-word essay through the Tony Blair Institute for Global Change.
The timing matters. The Prime Minister has been fighting for political survival. The Labour leadership conversation has consumed all the oxygen. Blair stepped in, shook the snow globe, and most coverage focused entirely on Brexit and Net Zero.
What almost nobody focused on was the policy detail buried inside.
Point four of his ten-point programme calls for: "a major new partnership with the private and voluntary sectors for apprenticeships and training, not just for the young and unemployed, but for the existing workforce whose jobs will be affected by AI and who need to learn AI adoption."
Point nine goes further: "reorganising the whole of government around the harnessing of the 21st-century technological revolution... Digital ID is just one, though vital, part of it."
Apprenticeship and training partnerships. AI adoption. Digital ID described as vital.
These are policy recommendations. They are also, if you know where to look, a description of infrastructure that already exists and is already being funded. I'll be returning to that in detail. For now, note that the essay was published by an institute that has received a confirmed £257 million funding floor from Larry Ellison, founder of Oracle, one of the primary technology companies positioned to benefit from exactly this policy direction.
The Sovereign AI Fund and Euan Blair's Multiverse have already entered the public conversation. What hasn't been asked yet is why Tony Blair's policy recommendations map so precisely onto businesses run by his own family. That question is coming.
Then Question Time.
Two days later, the BBC ran a special AI episode.
Panel: Darren Jones MP, Julia Lopez MP, Mo Gawdat, Victor Riparbelli of Synthesia, and Laura Gilbert, described as Senior Director of AI at the Tony Blair Institute.
It was framed as a debate. It felt like a soft sell. Every voice sat somewhere between enthusiastic supporter and true believer. The range of opinion ran from "AI is the future" to "AI is the future, but be careful." Nobody on that panel was there to genuinely challenge the direction of travel.
That framing, the absence of scrutiny presented as balance, is something I've been watching operate across multiple institutions for months. The BBC didn't invent it. But on Thursday night it delivered it to several million people.
Look more closely at Laura Gilbert.
Her own public speaker profile is the story.
She is Head of AI for Government at the Ellison Institute of Technology in Oxford, Larry Ellison's own research institute. As part of that role, she works one day a week on secondment to the Tony Blair Institute. And one day a week on secondment to DSIT, the government department directly responsible for AI policy in the United Kingdom.
Before this: five years at 10 Downing Street, founding director of the government's data science team and the AI Incubator under both Conservative and Labour governments.
Read that back slowly.
The BBC placed on its independent AI debate panel someone simultaneously employed by Ellison's institute, seconded to Blair's institute, and advising the government department that sets the policy environment both organisations directly benefit from.
She is not an independent voice. She is the junction point of three of the most consequential actors in British AI policy, presented to millions as a neutral expert.
That is not an editorial oversight. That is what the architecture looks like when it surfaces in public.
The problem isn't AI.
Today feels like the right moment to join this conversation.
What I have been working on is large and wide, far bigger than I first anticipated. Explaining it is not easy, but it is necessary.
AI is undoubtedly a powerful tool, and like anything in this world it can be used for negative or positive effect. What makes it different is the magnitude. The capability is already extraordinary and the trajectory of improvement is exponential. The consequences of getting the deployment wrong are not proportionate, they are magnified.
Implemented as a tool of control by a government or nation under pressure, AI becomes the most powerful instrument of oppression ever built. Surveillance at scale. Censorship without friction. The rewriting of history in real time. The short-term architecture looks like Orwell, 1984 brought forward and automated. But the longer-term risk is more insidious than that. Once embedded, it can be used to condition populations in ways we are only beginning to understand, moving from control by fear toward something closer to Huxley's vision: a compliant society that no longer recognises its own compliance.
And yet AI is also a genuine leveller. It brings people with good ideas and real capability, but without the platform, the connections, or the resources to act on them, up to the same level as those who have always had those advantages. The positive potential is just as unimaginable as the negative. That is what makes the deployment question so consequential.
I fear we are already further down this path than we are being told.
What I have been documenting is an architecture, the shape of how much of this is already more deeply embedded than the public conversation acknowledges. The soft sell is the mechanism: a slow acclimatisation, designed to bring people to acceptance of decisions that have, for the most part, already been made. Without a vote. Without consent. Without agency.
A panopticon is being constructed. And the nature of a panopticon is that by the time the people inside it fully understand what it is, the architecture is already complete.
When that moment of recognition arrives, and it will, the backlash against AI will be significant. We are already seeing the early signs. That will accelerate. And when it does, the people who built this without consent will point to the backlash as proof that the public cannot be trusted with the conversation.
That is why the conversation has to happen now, while there is still time to shape the outcome.
This is one thread in an architecture I've been mapping for months. The connections run deeper, the stakes are higher, and the public case has not yet been made in full.
I intend to make it.
Sources:
Tony Blair Institute essay, 26 May 2026: https://institute.global/insights/politics-and-governance/the-labour-party-is-playing-with-fire-over-its-future-and-the-future-of-the-country
Ellison funding of TBI, Lighthouse Reports: https://www.lighthousereports.com/investigation/blair-and-the-billionaire/
BBC Question Time AI special announcement: https://x.com/bbcquestiontime/status/2059702556994236474
Laura Gilbert public profile, OpenUK:
https://openuk.uk/profiles/dr-laura-gilbert/
Ellison Institute of Technology, Gilbert appointment: https://eit.org/news/eit-expands-team-of-field-leading-experts-spanning-ai-economics-life-sciences-generative-biology-and-medicine
The problem isn't AI. It's the deployment of it by the highest risk group to use it. I've been investigating this quietly for months. Today I want to start talking.
I've been quiet on here for a few weeks. Working. Thinking. Beginning to document something I couldn't quite name yet.
The pace of events in the UK has changed. I've been watching it. And this week, two things happened in close succession that made me want to start talking.
The Blair intervention.
On 26 May, Tony Blair published a 5,600-word essay through the Tony Blair Institute for Global Change.
The timing matters. The Prime Minister has been fighting for political survival. The Labour leadership conversation has consumed all the oxygen. Blair stepped in, shook the snow globe, and most coverage focused entirely on Brexit and Net Zero.
What almost nobody focused on was the policy detail buried inside.
Point four of his ten-point programme calls for: "a major new partnership with the private and voluntary sectors for apprenticeships and training, not just for the young and unemployed, but for the existing workforce whose jobs will be affected by AI and who need to learn AI adoption."
Point nine goes further: "reorganising the whole of government around the harnessing of the 21st-century technological revolution... Digital ID is just one, though vital, part of it."
Apprenticeship and training partnerships. AI adoption. Digital ID described as vital.
These are policy recommendations. They are also, if you know where to look, a description of infrastructure that already exists and is already being funded. I'll be returning to that in detail. For now, note that the essay was published by an institute that has received a confirmed £257 million funding floor from Larry Ellison, founder of Oracle, one of the primary technology companies positioned to benefit from exactly this policy direction.
The Sovereign AI Fund and Euan Blair's Multiverse have already entered the public conversation. What hasn't been asked yet is why Tony Blair's policy recommendations map so precisely onto businesses run by his own family. That question is coming.
Then Question Time.
Two days later, the BBC ran a special AI episode.
Panel: Darren Jones MP, Julia Lopez MP, Mo Gawdat, Victor Riparbelli of Synthesia, and Laura Gilbert, described as Senior Director of AI at the Tony Blair Institute.
It was framed as a debate. It felt like a soft sell. Every voice sat somewhere between enthusiastic supporter and true believer. The range of opinion ran from "AI is the future" to "AI is the future, but be careful." Nobody on that panel was there to genuinely challenge the direction of travel.
That framing, the absence of scrutiny presented as balance, is something I've been watching operate across multiple institutions for months. The BBC didn't invent it. But on Thursday night it delivered it to several million people.
Look more closely at Laura Gilbert.
Her own public speaker profile is the story.
She is Head of AI for Government at the Ellison Institute of Technology in Oxford, Larry Ellison's own research institute. As part of that role, she works one day a week on secondment to the Tony Blair Institute. And one day a week on secondment to DSIT, the government department directly responsible for AI policy in the United Kingdom.
Before this: five years at 10 Downing Street, founding director of the government's data science team and the AI Incubator under both Conservative and Labour governments.
Read that back slowly.
The BBC placed on its independent AI debate panel someone simultaneously employed by Ellison's institute, seconded to Blair's institute, and advising the government department that sets the policy environment both organisations directly benefit from.
She is not an independent voice. She is the junction point of three of the most consequential actors in British AI policy, presented to millions as a neutral expert.
That is not an editorial oversight. That is what the architecture looks like when it surfaces in public.
The problem isn't AI.
Today feels like the right moment to join this conversation.
What I have been working on is large and wide, far bigger than I first anticipated. Explaining it is not easy, but it is necessary.
AI is undoubtedly a powerful tool, and like anything in this world it can be used for negative or positive effect. What makes it different is the magnitude. The capability is already extraordinary and the trajectory of improvement is exponential. The consequences of getting the deployment wrong are not proportionate, they are magnified.
Implemented as a tool of control by a government or nation under pressure, AI becomes the most powerful instrument of oppression ever built. Surveillance at scale. Censorship without friction. The rewriting of history in real time. The short-term architecture looks like Orwell, 1984 brought forward and automated. But the longer-term risk is more insidious than that. Once embedded, it can be used to condition populations in ways we are only beginning to understand, moving from control by fear toward something closer to Huxley's vision: a compliant society that no longer recognises its own compliance.
And yet AI is also a genuine leveller. It brings people with good ideas and real capability, but without the platform, the connections, or the resources to act on them, up to the same level as those who have always had those advantages. The positive potential is just as unimaginable as the negative. That is what makes the deployment question so consequential.
I fear we are already further down this path than we are being told.
What I have been documenting is an architecture, the shape of how much of this is already more deeply embedded than the public conversation acknowledges. The soft sell is the mechanism: a slow acclimatisation, designed to bring people to acceptance of decisions that have, for the most part, already been made. Without a vote. Without consent. Without agency.
A panopticon is being constructed. And the nature of a panopticon is that by the time the people inside it fully understand what it is, the architecture is already complete.
When that moment of recognition arrives, and it will, the backlash against AI will be significant. We are already seeing the early signs. That will accelerate. And when it does, the people who built this without consent will point to the backlash as proof that the public cannot be trusted with the conversation.
That is why the conversation has to happen now, while there is still time to shape the outcome.
This is one thread in an architecture I've been mapping for months. The connections run deeper, the stakes are higher, and the public case has not yet been made in full.
I intend to make it.
Sources:
Tony Blair Institute essay, 26 May 2026: https://institute.global/insights/politics-and-governance/the-labour-party-is-playing-with-fire-over-its-future-and-the-future-of-the-country
Ellison funding of TBI, Lighthouse Reports: https://www.lighthousereports.com/investigation/blair-and-the-billionaire/
BBC Question Time AI special announcement: https://x.com/bbcquestiontime/status/2059702556994236474
Laura Gilbert public profile, OpenUK:
https://openuk.uk/profiles/dr-laura-gilbert/
Ellison Institute of Technology, Gilbert appointment: https://eit.org/news/eit-expands-team-of-field-leading-experts-spanning-ai-economics-life-sciences-generative-biology-and-medicine