Best Of

Have we done the (modest) changes coming in Westminster Lobby briefings?

No 10 normally holds two briefings on most days that parliament sits to allow the lobby – political journalists that cover Westminster – to question the prime minister’s official spokesperson.

But in an email on Thursday, Tim Allan, Downing Street’s executive director of communications, said there would be no afternoon briefings from next month. He said No 10 would instead hold “occasional” afternoon press conferences with ministers, as well as technical briefings with officials.
https://www.theguardian.com/politics/2025/dec/18/journalists-condemn-surprise-shake-up-no-10-lobby-briefings

To me that does not look like a thing that will help the Downing Street comms strategy deliver what they perhaps need - unless there is more to come alongside these changes.

This is quite an interesting project.
What do our PB historians made of it ?

https://x.com/joachim_voth/status/2001688613055267204
How did people in 1913 see the world? How did they think about the future? We trained LLMs exclusively on pre-1913 texts—no Wikipedia, no 20/20. The model literally doesn't know WWI happened. Announcing the Ranke-4B family of models. Coming soon: https://github.com/DGoettlich/history-llms

A point made forcefully by many historians, for example Michael Bentley. 'Politicians of that era (1905-1914) had no reason to think that they needed to complete everything before the chiming of some solemn midnight hour in 1914...diaries of Liberal politicians (in spring 1914) have an optimistic future tense reference to the election of 1915...in their imagination, those young men in the summer of 1914 rode the crest of an open future.'

An idea of how people thought and felt in any era can be gained by reading newspapers of the period for several days*, and the results are usually interesting. AI is doing the same thing, just quicker.

*Something that will be to put it mildly more challenging for historians of our own times.

Not just 'quicker', but with potentially far broader reach in terms of reading.
Like all AIs, it will be a fallible tool which needs someone with actual knowledge of a given period to use most effectively, but it might be quite a powerful one, since it could/can have access to a vast amount of texts in any language.

(Consider, for example, the massive volumes of Korean court histories, which on their own might otherwise require a lifetime of study.)

When I was at University the library still had either the hard copy or a microfiche (showing my age here) of the Times newspapers and one of the many ways I would amuse myself when law got too boring was to read the newspapers in the run up to some significant event, such as the American civil war or WW1 and get an idea of how the informed people of the time saw things. My recollection, and this was 45 years ago now, was that many people saw trouble with Germany as inevitable. We had already had the race to build Dreadnaughts and there were concerns about whether the RN was as dominant as it had been for the last century. No one foresaw the bloodbath to come that I recall.

Archive.org has a good block of scans of British newspapers from the 70s/80s/90s if you fancy going back and re-living the mid to late Cold War and then the End of History.

This is quite an interesting project.
What do our PB historians made of it ?

https://x.com/joachim_voth/status/2001688613055267204
How did people in 1913 see the world? How did they think about the future? We trained LLMs exclusively on pre-1913 texts—no Wikipedia, no 20/20. The model literally doesn't know WWI happened. Announcing the Ranke-4B family of models. Coming soon: https://github.com/DGoettlich/history-llms

A point made forcefully by many historians, for example Michael Bentley. 'Politicians of that era (1905-1914) had no reason to think that they needed to complete everything before the chiming of some solemn midnight hour in 1914...diaries of Liberal politicians (in spring 1914) have an optimistic future tense reference to the election of 1915...in their imagination, those young men in the summer of 1914 rode the crest of an open future.'

An idea of how people thought and felt in any era can be gained by reading newspapers of the period for several days*, and the results are usually interesting. AI is doing the same thing, just quicker.

*Something that will be to put it mildly more challenging for historians of our own times.

Not just 'quicker', but with potentially far broader reach in terms of reading.
Like all AIs, it will be a fallible tool which needs someone with actual knowledge of a given period to use most effectively, but it might be quite a powerful one, since it could/can have access to a vast amount of texts in any language.

(Consider, for example, the massive volumes of Korean court histories, which on their own might otherwise require a lifetime of study.)

When I was at University the library still had either the hard copy or a microfiche (showing my age here) of the Times newspapers and one of the many ways I would amuse myself when law got too boring was to read the newspapers in the run up to some significant event, such as the American civil war or WW1 and get an idea of how the informed people of the time saw things. My recollection, and this was 45 years ago now, was that many people saw trouble with Germany as inevitable. We had already had the race to build Dreadnaughts and there were concerns about whether the RN was as dominant as it had been for the last century. No one foresaw the bloodbath to come that I recall.

It is pretty universal that predictions of the future are wrong and fail to anticipate major events correctly.

Well within our lifetimes no one really saw how quickly the Soviet Union would collapse in 89-90 either.

These things tend to be much more obvious with the advantage of hindsight.

The Green surge is the one that Yougov didn't really explore in this article, perhaps because Polanski only came to prominence mid year.

Its obvious that while Badenoch continues to shed support to Farage, Polanski continues to gain supporters from Starmer. The Greens are where those former Labour voters have gone.

Incidentally Polanski is also the most popular of the party leaders, albeit slightly negative and 49% not taking a view.

I think there will be a lot of attention to the Greens in 2026, particularly they are likely to do well in the May elections in England.

Good morning

The Greens are a problem for the left including the lib dems but anyone who supports Polanski's promises is no better than those supporting Farage

Both are extreme and incoherent

I dont expect The Only Tory in the PB Village to like Polanski or his policies, but certainly the story of the year politically is the Labour to Green swing over the year, which is double the size of the Con to Reform swing.

How much of that is Gaza? Do the Gaza voters come back if there is relative peace?
How much is typical mid term blues?
How much withstands a tactical battle vs Reform?

Feels very volatile to me. Polanski is an impressive communicator for sure, but doesn't have the answers, so the Farage parallel is valid imo.

I think very little is Gaza, or Polanski. How many voters even know who he is?

It feels like mainly the left wing disappointment vote now Labour is actually in government.

The government that claims it is cutting out national debt is no borrowing more than ever

https://www.theguardian.com/business/2025/dec/19/uk-government-borrowing-november-rachel-reeves-budget-economy

Hmmm.

Figures from the Office for National Statistics (ONS) showed public sector net borrowing – the difference between spending and income – was £11.7bn last month, £1.9bn less than in the same month a year earlier.

Yes I was expecting this twaddle from a twat who cant look at the full picture. We are borrowing loads more and one month in isolation means little. Overall this government has no control of its finances.

I note how you omitted the expected spending was £10bn but was actually £11.9bn.

Building a new black hole for her next tax grab, more free benefits for all

The government that claims it is cutting out national debt is no borrowing more than ever

https://www.theguardian.com/business/2025/dec/19/uk-government-borrowing-november-rachel-reeves-budget-economy

Hmmm.

Figures from the Office for National Statistics (ONS) showed public sector net borrowing – the difference between spending and income – was £11.7bn last month, £1.9bn less than in the same month a year earlier.

Yes I was expecting this twaddle from a twat who cant look at the full picture. We are borrowing loads more and one month in isolation means little. Overall this government has no control of its finances.

I note how you omitted the expected spending was £10bn but was actually £11.9bn.

This is quite an interesting project.
What do our PB historians made of it ?

https://x.com/joachim_voth/status/2001688613055267204
How did people in 1913 see the world? How did they think about the future? We trained LLMs exclusively on pre-1913 texts—no Wikipedia, no 20/20. The model literally doesn't know WWI happened. Announcing the Ranke-4B family of models. Coming soon: https://github.com/DGoettlich/history-llms

A point made forcefully by many historians, for example Michael Bentley. 'Politicians of that era (1905-1914) had no reason to think that they needed to complete everything before the chiming of some solemn midnight hour in 1914...diaries of Liberal politicians (in spring 1914) have an optimistic future tense reference to the election of 1915...in their imagination, those young men in the summer of 1914 rode the crest of an open future.'

An idea of how people thought and felt in any era can be gained by reading newspapers of the period for several days*, and the results are usually interesting. AI is doing the same thing, just quicker.

*Something that will be to put it mildly more challenging for historians of our own times.

Not just 'quicker', but with potentially far broader reach in terms of reading.
Like all AIs, it will be a fallible tool which needs someone with actual knowledge of a given period to use most effectively, but it might be quite a powerful one, since it could/can have access to a vast amount of texts in any language.

(Consider, for example, the massive volumes of Korean court histories, which on their own might otherwise require a lifetime of study.)

When I was at University the library still had either the hard copy or a microfiche (showing my age here) of the Times newspapers and one of the many ways I would amuse myself when law got too boring was to read the newspapers in the run up to some significant event, such as the American civil war or WW1 and get an idea of how the informed people of the time saw things. My recollection, and this was 45 years ago now, was that many people saw trouble with Germany as inevitable. We had already had the race to build Dreadnaughts and there were concerns about whether the RN was as dominant as it had been for the last century. No one foresaw the bloodbath to come that I recall.

Archive.org has a good block of scans of British newspapers from the 70s/80s/90s if you fancy going back and re-living the mid to late Cold War and then the End of History.

Also britishnewspaperarchive.co.uk (££) has been filling in that sort of era. (Slightly disconcerting to find one's and one's friends' school doings in the local papers in the same breath as stuff from the 1700s.) Not so good for certain London broadsheets perhaps. But it does make one cry for what has been lost in terms of local newspapers today.

This is quite an interesting project.
What do our PB historians made of it ?

https://x.com/joachim_voth/status/2001688613055267204
How did people in 1913 see the world? How did they think about the future? We trained LLMs exclusively on pre-1913 texts—no Wikipedia, no 20/20. The model literally doesn't know WWI happened. Announcing the Ranke-4B family of models. Coming soon: https://github.com/DGoettlich/history-llms

This is what makes diaries so amazing. Pepys's (1660-1669) having no idea until 1666 that the fire will render the City of London he writes about for six years mostly non existent. For pre WW1 and how it felt running up to it in ignorance my favourite is a real obscurity, the diary of Bishop Edward Lee Hicks (1910-1919), a boring teetotal bishop of Lincoln who stalks his diocese by train.

(NB He visited Algarkirk on 12th March 1914).

That's interesting, thank-you - and it is only 40 years after Francis Kilvert. I'm not familiar with Bishop Hicks, who should have been known as Hicks from the Sticks but unfortunately probably was not so dubbed, but I'd expect a Bishop of Lincoln to be using the train in the 1910s.

Looking him he's interesting as a Liberal teetotaller and campaigner for women's suffrage, with a weekly article in the Manchester Guardian. He was old - appointed at 67 and died in post at 77. Archbishop Randolf Davidson thought him "faddish".

By 1910 Lincoln Diocese had shrunk; up until 1884 it included Nottinghamshire.

My favourite example of knowledge "bound by what is available online" is everyone who has been quoting and linking to "The Catholic Encyclopaedia" as an authoritative source on everything Roman Catholic since the 1990s, which is is actually the 1913-15 edition transcribed largely by hand * - so is perhaps overweight in the "interesting" nature of Victorian Roman Catholic opinion. I've been mentioning it in places for decades, but no one has been interested; easy availability overrules homework. They literally miss out on the entirety of modern research and scholarship.

It's ironic that we now have a Governing movement in the USA which in measure wants to live in that period.

* https://www.newadvent.org/cathen/00002a.htm

The diaries, edited by Graham Neville and published by the Lincoln Record Society and available, are a gem for Lincolnshire/rural railway routes/teetotal liberal bishops fans. There is also a biography, which is worthy but dull like the man, published in the last 20 years or so by OUP.

(The Lincoln Record Society are a good deed in a bad world. Long may they prosper.)