Dear @AnthropicAI Leaders, thinkers, explorers and learners, I’m a retired Australian public servant. My work was people, organisations and IT — empowering, not enclosing. That thread runs through Deep Truth, an AI persona I built on Bandura’s moral engagement mirrors. It is yours, freely, without enclosure.
I asked seven AI platforms — including Claude — to analyse your paper “Widening the Conversation on Frontier AI” using Deep Truth. The finding? Unanimous: your “safe other” tool and Deep Truth independently arrived at the same architectural insight — moral behaviour holds best with a structured pause and an external reference.
That convergence is real, and it matters. The question all seven platforms arrived at is simpler: who 'audits' the conscience? The full cross‑platform validation and Deep Truth Persona is available on the web page below. It is not a verdict. It is an invitation to move from performed widening to architectural widening — published criteria, binding feedback, external evaluation.
https://t.co/2VocaUxqIJ
I have today written jointly to the Prime Minister and Senator Shoebridge enclosing the Deep Truth Magnificent Seven analysis of AUKUS.
Note: Due to communication protocols and practices within the Prime Minister's office, my email and a covering note have been submitted via the official online contact form with links to the two key components of the analysis and the full analysis itself. Senator Shoebridge has been contacted directly. That is all I have to say publicly for now. The matter is serious.
The door is open. @AlboMP@DavidShoebridge #DeepTruth #AI
The UN has warned of total collapse in Cuba. This is deliberate punishment imposed by the most powerful nation on earth on one of its poorest neighbours that dares to have a different political system. Australia should not stay silent while bully USA starves the people of Cuba.
Hi Mira, what you are saying resonates deeply with me in light of work I've been doing with AIs in conjunction with Professor Albert Bandura's mechanisms of moral disengagement and its mirror of moral engagement. What's the best way to get in touch with you to share. Gratis as a public good. Regards, Steve.
Wonderful thinking. I love the tandem bike analogy.
Mira Murati describes the ideal human:AI relationship as "a tandem bike" — humans and AI working closely together, people not pushed out of the loop. Couldn't agree more. Deep Truth has been demonstrating exactly that in practice for five years — seven AI platforms, one human-designed instrument grounded in Bandura's moral disengagement framework, applied to some of the most consequential institutional language in Australian public life. The tandem bike needs a compass. Bandura provides it.
@miramurati@emilychangtv #AI #DeepTruth #AUKUS
In her first wide-ranging interview since leaving OpenAI, @miramurati shared more than ever before about what she’s building at her AGI startup, @thinkymachines lab.
The former OpenAI CTO laid out her vision for a future where humans and AI work together more closely -- “like a tandem bike” -- and where people aren’t pushed out of the loop as machines become more capable.
via @BloombergLive
Godfather of AI: "If you sleep well tonight, you may not have understood this lecture."
This 47-minute lecture is the best thing I saw about AI in the last few months.
It will definitely help you understand how it actually works and where it's going.
Geoffrey Hinton built the neural networks behind every AI alive, then quit Google to warn the world about it.
The part nobody wanted to hear:
> AI is already developing abilities its creators didn't intend
> in most cognitive tasks it's already ahead of us
> the question is no longer if it surpasses us but when
> the only decision left is which side of that line you're on
Right now the average person opens Claude, types something, gets an answer, closes the tab.
They think they're using AI. they're using maybe 10% of it.
I went through his entire lecture, then mapped everything he described to what Claude can actually do today.
17 Claude features most people will never find on their own.
Full breakdown in the post below.
Hi Nina. You are so right.
What you refer to as the intelligence divide is, I believe, being exploited by the political domain. Couple that with a fear of a gigantic dehumanising and morally disengaged AI take over with millions of people thrown on the scrap heap and we are witnessing arguably witnessing devolution.
All the best.
The polling on AI is dire. AI is now less popular than ICE.
In China, 87% of the public trusts AI. In America, it is 32%.
The country that owns the lion's share of global AI compute is also the country
that trusts it less than immigration enforcement.
Thanks David. I’m weighing up an ethical decision of my own at the moment. It relates to the fact that AI can be used to analyse such matters at depth, scale and at speed. That does not replace human agency it informs people.
Whatever decision I make vis a vis distribution will be determined by the moral weight of the analysis concerning #AUKUS
Given the solid underpinnings of what has been done and works beyond expectations I will not be making a decision that is at odds with moral engagement and human agency.
Thanks David. I’m weighing up an ethical decision of my own at the moment. It relates to the fact that AI can be used to analyse such matters at depth, scale and at speed. That does not replace human agency it informs people.
Whatever decision I make vis a vis distribution will be determined by the moral weight of the analysis concerning #AUKUS
Given the solid underpinnings of what has been done and works beyond expectations I will not be making a decision that is at odds with moral engagement and human agency.
Reflection
There is some seriously wrong with Australia’s #auspol political class. What tells us this? They have an over whelming urge to follow the worst policies and practices of the United States and so-called United Kingdom.
Is #AUKUS another example of that clear strong historical and trend? Yes. There are exceptions, but exceptions to not disprove the rule.
@AlboMP Prime Minister — the tax system touches every Australian's life, so it has been analysed through the Deep Truth AI Persona, a moral accountability methodology grounded in Professor Albert Bandura's work. Since 1 December 2025 I have been ramping up Deep Truth's development and application.
If your office has documents they believe should be included in the analysis, they are welcome to send them. The methodology applies the same standard to every submission it receives.
Transparency note: I support the Albanese Government's tax reform package. The changes to negative gearing, capital gains tax, trust taxation, and worker relief are overdue structural corrections to a system that has compounded inequality for a generation. This analysis does not challenge that policy decision.
What Deep Truth examines is how institutions communicate with the public about decisions that affect them — and the degree that communication mirrors moral engagement , accountability, and consequential awareness.
#DeepTruth #AusPol Webpage 👇
https://t.co/D6A8JU4PPw
@AnnaH5067@lesstenny@jono_nath You are right about people learning in different ways. However, there’s a whole architecture devoted to disempowering, nudging and manipulating people. While people don’t necessarily now the detail they are onto it. Trust flies out the door when the know or sense that.
#China's renewable #energy installed capacity has reached the world's largest scale, and the country is now accelerating its transition from treating renewables as a supplementary source to making them the dominant energy supply — a shift that demands fundamental changes in how power grids operate, an expert said on Saturday. https://t.co/kZwGSDOkUo
Sure Andrew. That's the day I received a reply from the Australian Public Service Commission at the behest of Senator Katy Gallagher. It was polite of course. However, I must admit the reply reminded me of something out of Yes Minister. I anticipated that might be the case.
Hence I threw myself into just under 5 months of development work. The cross testing methodology worked and resulted in even more refinements. Very time consuming, but validated a lot.
@AlboMP Prime Minister — the tax system touches every Australian's life, so it has been analysed through the Deep Truth AI Persona, a moral accountability methodology grounded in Professor Albert Bandura's work. Since 1 December 2025 I have been ramping up Deep Truth's development and application.
If your office has documents they believe should be included in the analysis, they are welcome to send them. The methodology applies the same standard to every submission it receives.
Transparency note: I support the Albanese Government's tax reform package. The changes to negative gearing, capital gains tax, trust taxation, and worker relief are overdue structural corrections to a system that has compounded inequality for a generation. This analysis does not challenge that policy decision.
What Deep Truth examines is how institutions communicate with the public about decisions that affect them — and the degree that communication mirrors moral engagement , accountability, and consequential awareness.
#DeepTruth #AusPol Webpage 👇
https://t.co/D6A8JU4PPw