Worries of a dystopian society apart it seems to me that in the not very distant future an AI could not be worse than the current shower in power.

Eg: read the comments from the [1]Covid inquiry "Mr Johnson was "clearly bamboozled" by some of the science." It is worrying quite how unsuited many politicians are in today's scientific age.

Yes: I would not want an AI in No 10, but please can we have some with some real understanding of the world. (I am not going to even start on characteristics such as honesty and working for the common good rather than themselves.)

[1] https://www.bbc.co.uk/news/health-67479548

They could call it TwatGPT.

Ah, you are still following the media circus that's the Covid Inquiry, and reading the BBC. I see your problem.

Surely it's the job of Valance to communicate to the PM and explain The Science if the PM is "clearly bamboozled"?

Whilst it may seem "obvious" that the Eat Out to Help Out scheme would cause an increase in transmission, if you look at the graphs for around that time it's really not very easy to pick out the time of the scheme.

https://trusttheevidence.substack.com/p/the-hallett-inquiry-update

And while Valance is now claiming, with hindsight, that we should have locked down sooner, he was on TV two weeks before saying that we should pursue a herd immunity strategy. However, somehow the inquiry hasn't pressed him on that. The whole thing is a massive waste of time and money, and it's clearly a whitewash. It's only going to state that Boris was a buffoon (we all knew that anyway) and we should have locked down sooner (to what end?)

Follow certain substack accounts, and it'll balance out some of the BS you get from the BBC.

The whole thing is a massive waste of time and money, and it's clearly a whitewash.

May I refer you to a quote from a great political documentary?

“No Minister, I beg you,” replies Sir Humphrey. “A basic rule of government is never look into anything you don’t have to, and never set up an inquiry unless you know in advance what its findings will be.

The findings will be that the civil service is 100000% the bestest system in the world, let down by somebody who has departed the scene and can't mount a political defence. And lessons have been learned etc etc etc.

"And while Valance is now claiming, with hindsight, that we should have locked down sooner, he was on TV two weeks before saying that we should pursue a herd immunity strategy. However, somehow the inquiry hasn't pressed him on that"

Actually that was discussed at the inquiry, yesterday I think. Chris Whitty said that government wasn't pursuing a herd immunity strategy at the time and that he was trying to persuade ministers to not discuss it on open, as it would give a misleading picture. But because the debate about it (herd immunity) was happening in public, it looked like it was a government policy. Whitty confessed this was a mistake saying "We didn't help the public by having a debate that I think quite rightly upset and confused a lot of people"

On the one hand, putting AI into civil servants hands probably isn't going to improve their efficiency very much, so that's probably a waste of money.

On the other hand, "No more staff budget" as in the title is a bit misleading. I've worked with a lot of civil servants, and many of them spend a lot of time just shuffling papers around or attend meetings all day just explaining why they can't deliver something. They spend weeks putting together a document, and then complain that the requirements are a bit vague. When you suggest that they call up the person who put the requirements together and actually ask them for clarification, you just get a blank stare.

I don't think having more of them on the payroll is going to help. I guess we'll find out next year as Labour are supposed to favour Big Government. I'm from the Ron Swanson school of government (keep it very small). And if you haven't seen Parks and Recreation, you are missing something great.

"Labour are supposed to favour Big Government"

They've committed to "financial prudence until the economy returns to growth", which is as good as saying they will stick to the broad spending plans set out by the current shower of piddle. On most notable policy areas, including energy, environment, transport, business there's no difference of any materiality between the two parties, and not much on justice, immigration and health. For any changed or new plans, where will they find the money if they aren't going to change spending plans? Taxing the rich has been tried before and results in lower receipts. For all the blustering, we have a uniparty government whoever gets in, but both parties specialise in pretending there's a big difference.

In any event, let's look at the cost of the civil service, that's around £11bn a year. That's to administrate total spending of £1.19 trillion, so say 1%. The nearest private sector metric would be organisational SG&A costs, and the leading analysts on that are the Hackett Group, who reckon that for European companies the median value is about 12% of turnover, and first quartile is 5-6%. Which would mean that UK government administration is (in pure cost terms) somewhere around second quartile performance. That's pretty good considering the breadth and complexity of government operations. Operationally I have no data, but I'd guess from my thirty years working in a range of large companies that UK government effectiveness is around the private sector median - I've seen private sector screwups brushed under the carpet that are every bit as bad as the very public government screwups.

"I've seen private sector screwups brushed under the carpet that are every bit as bad as the very public government screwups."

So? In the usual way, a private company that messes up too often goes to the wall.

A government mess on the other hand - we all pay for it. Always.

So where do you want higher performance?

Of course, many large private companies have some way of getting the government to pay for their messes (cost-plus contracts, "change of scope", lobbying for subsidies...).

In those cases I would say for those responsible:

"Up against that wall please. Cigarette? Blindfold?"

>” On the other hand, "No more staff budget" as in the title is a bit misleading”

Definitely misleading, as the IT will have to increase to fund the AI and it’s supporting staff.

I remember back in the mid 1990s implementing a housing benefit system based on an”AI” (ie. A rules-based system), the full project was canned because the rules handed down from central government changed faster than the programmers could convert them into a (validated) rule set, plus some of the directives didn’t readily translate into machine implementable rules, so cases were constantly being handed off to humans and thus the staff savings were not achievable. However, the system did one thing, it prevented paper being “lost” or “forgotten in a cabinet”, so when the (AI) project was cancelled, they have the data to show that the staff were both fully utilised and more productive than previously, so more needed to be recruited.

Wow, all those years I spent learning stuff, when I could have just gone along to an "upskilling event" instead!

This is unlikely to be any time soon. HMRC have refused to implement 64-bit Microsoft Office until 'at least' the end of 2024 (they're currently running a weird mix of 32 bit windows on some machines and 32bit office with 64bit unpatchd windows 10 on others), as they stated "64 bit software is too new to trust".

So AI in government will be like 2250 at the soonest.

Government on the cheap is not necessarily efficiency.

The Environment Agency, for example, has lost over half of its environmental protection budget over the last decade. One consequence is the constant discharge of sewage into rivers and seas. That's not efficiency, it's capitulation. It doesn't even have to be a cost to the public purse - it could be met by a charge on water companies.

HMRC has been saving money - for example by shutting down its enquiry lines between 12th June and 3rd September. That's not efficiency, it's an abdication of public service. Efficiency would come from simplifying the tax system so it generated fewer enquiries.

The problem is that politicians will the ends but are not prepared to will the means. There's a cost to good public administration and bad public administration simply wastes money by failing to achieve its objectives. No doubt you could save money by treating people essentially as data points and using AI to decide whether they [1]fraudulently claimed childcare benefit , or decide whether they [2]should go to jail or they can go [3]to college - but do we want to be subservient to machines simply to save a few quid?

[1] https://www.uva.nl/en/shared-content/faculteiten/en/faculteit-der-rechtsgeleerdheid/news/2023/02/childcare-benefit-scandal-transparency.html

[2] https://www.technologyreview.com/2019/01/21/137783/algorithms-criminal-justice-ai/

[3] https://www.forbes.com/sites/brennanbarnard/2023/11/21/will-ai-be-reviewing-your-college-application/

"Government on the cheap is not necessarily efficient."

Correction:

Government at *any* cost is not efficient. Full Stop.

Efficiency limits the 'opportunities' for 3rd parties to make a profit, trying to 'help' government to be 'more' 'efficient' !!! :)

('help', 'more' and 'efficient' are subject to a wide gamut of meanings, often at the same time by multiple parties. !!!

:)

I don’t really mind how Gov services the population, but the evidence is so far that humans are way better at coping with anything other than a very simple question. Begs the question of perhaps Gov should just recruit and train and pay their staff properly and forget all these AI vanity projects. Oh but some other countries might get to AI supremacy first! I don’t care, just do what works for our population and do it well.

While I’m here, kudos to the UK Gov pensions team who clearly have their ducks in a row. It’s been a pleasure talking to you.

TL:DR

We don't know what we're talking about but it won't stop us talking.

That it A in't I ntelligent to give money to Crapita, or our private medical data to Palantir?

Of course they'll never listen to any intelligent discussion!

Here's the fast and cheap version: "Everybody, repeat after me: Spreadsheets are not a database! When you have a database of data*, use a database program for it ."

* I know, I know. But we're tuning this for bureaucrats.

Down voted for being off topic, :)

please rephrase with the solution being “…use an AI for it”.

“

"The announcement came alongside a requirement to draw up plans to reduce the state's size"

I will believe it when I see it. The state is ever growing, the idea of 'austerity' was spending like slightly less of a drunken sailor, cutting quangos involved more and tax and stupid wastes of money continue to rise.

The state shrank as a % of GDP consistently between 2010 and 2019, although the consequences of that were rising hospital waiting lists, rising school class sizes, fewer police, a shrunken military, more pot holes, more congestion, and shorter prison terms, shrunken social care, plus rising student tuition charges (ie personal debt), amongst other effects. Nothing right or wrong about any of that, but if you want a smaller state then you want less public services.

There is a very British myth that "if we fixed public sector inefficiency, we could have public services like Denmark, and taxes like Chad (10% of GDP, for those not paying attention) unfortunately that is not the case. All very well railing on about waste and quangos, but there's only two main challenges - what level of public services we want, and then setting a tax level to pay for it. For many years the UK public spending has been a touch above the OECD average but well below comparator north western EU countries, and our tax take has been moderately below the OECD average, creating a deficit that's amongst the highest in the OECD, and net debt that's been inordinately high.

@Lurko

"The state shrank as a % of GDP consistently between 2010 and 2019"

I assume you mean the public sector being 46% of GDP (2010) down to 39.5% (2019)? Putting that into context the gov can only spend what it takes from the private people/business and borrowing against the private people/business. So the current utopia come about from the current 45% of the economy and 2010 was a high point of public spending after labour spent everything and left huge bills and they ramped up to about 40% of the economy!

Also remember labours splurge was 'investment' which means a return on investment. So if the increased spending must remain so to keep the level of service it is NOT an investment, and is at best a liability. And that 'investment' came at the expense of more debt, selling gold, PFI contracts, etc.

"Nothing right or wrong about any of that, but if you want a smaller state then you want less public services."

Yes. We already have more government than we are willing to pay for and so borrowing is insane. If we maintain or grow our monster it will continue borrowing and the interest payments of course eat into spending on services we want. Our larger state hasnt really improved our situation either as it steals more to redistribute to subsidies that make us poorer. In fact it has been government policy making our lives more expensive that is part of the inflation issue and energy crisis.

And with all this money we have more diversity officers but failure to follow the existing disaster plan for covid. The horrid expense of conflicting ideas that people be alone at a funeral while they party, that we must lock down and seriously harm the economy but subsidise eating out. It is general governmental incapability that spends so much and yet makes a bad situation. Some of it because they are not very good, some of it because central management is severely limited by bad information well out of date.

"For many years the UK public spending has been a touch above the OECD average but well below comparator north western EU countries, and our tax take has been moderately below the OECD average, creating a deficit that's amongst the highest in the OECD, and net debt that's been inordinately high."

That is a fair assessment of the problem. The dream that someone else will/should pay and a demand for more spending. The small pockets of activists who in a good few cases should be arrested are instead listened to. Instead of providing for the country the government spends our money appeasing strange cults. The government making expensive solutions to very simple problems.

There is always money available for shiny shit.

There is always money available for shiny shit or wars (Real or Imagined) !!!

:)

Pair programming is not only a misnomer (it should be called pair coding) and as implemented it's a disaster. Real programming involves thoughtful design before the keyboard is touched, and once the keyboard is reached there should be a consistent approach to implementation followed by testing, review and, if necessary, adjustment. "Pair programming" amounts to coming up with something that "works" by trial and error, exacerbated by quite possibly mixing two different views of what's needed. The result is typically muddle.

Really? An "upskilling event"?

Kill me now, I'm done with this planet...

"Really? An "upskilling event"?"

Typical, civil servant / Govt point of view:

My job is complicated and requires training, skill and some intelligence.

Your job is simple and I can do it by simply watching you for an hour or so and copying the process .... which is also simple. !!!

:)