Starmergeddon: They Came And Ate Us

[Biffer]

They’re just deliberately feeding the whackos now

https://www.theguardian.com/books/2025/ ... SApp_Other

[I like neeps]

They’re just deliberately feeding the whackos now

https://www.theguardian.com/books/2025/ ... SApp_Other

Nice looking coin though tbf.

[I like neeps]

It's not exactly a lack of action. They've introduced a bunch of bills and enacted some of the policies they campaigned on. They clearly do have a plan.

The problem is right now they are mostly focusing on the things that will get them better press (so they think) and coming out with hot air over immigration, culture war shit, and other Tory cover band populist nonsense doomed to failure that won't move the needle one bit. Instead of, y'know, putting people first, enacting a progressive agenda, and sticking to their guns rather than getting bogged down in a media war they can't win.

No amount of columns in the express by Starmer will change the fact that they're too chicken shit to push through genuine change that risks upsetting the right wing media and the mythological Tory swing voter.

I think the problem is IF they do have a plan (I'm not convinced as there's a different between a lot of activity and a plan) they can't verbalise it.

People just don't know what Starmer and this government stand for. Why are we making all of these sacrifices? What's the end goal?

Labour did multiple brandings and rebrandings in opposition- none cut through, they ran an election campaign hopelessly devoid of anything but the Tories are awful we can't say anything until we know how bad it is but your tax won't go up yadda yadda.

[dpedin]

It's not exactly a lack of action. They've introduced a bunch of bills and enacted some of the policies they campaigned on. They clearly do have a plan.

The problem is right now they are mostly focusing on the things that will get them better press (so they think) and coming out with hot air over immigration, culture war shit, and other Tory cover band populist nonsense doomed to failure that won't move the needle one bit. Instead of, y'know, putting people first, enacting a progressive agenda, and sticking to their guns rather than getting bogged down in a media war they can't win.

No amount of columns in the express by Starmer will change the fact that they're too chicken shit to push through genuine change that risks upsetting the right wing media and the mythological Tory swing voter.

I think the problem is IF they do have a plan (I'm not convinced as there's a different between a lot of activity and a plan) they can't verbalise it.

People just don't know what Starmer and this government stand for. Why are we making all of these sacrifices? What's the end goal?

Labour did multiple brandings and rebrandings in opposition- none cut through, they ran an election campaign hopelessly devoid of anything but the Tories are awful we can't say anything until we know how bad it is but your tax won't go up yadda yadda.

I agree they need to be clearer and more radical in their vision for the country - investing in NHS, Schools, renationalizing the railways, stopping tax breaks for private schools, etc are all things the public overwhelmingly support. However it needs more coherence. The problem is trying to get this message across, partly because they are not the best communicators but more because the political agenda is being driven by the right wing media like Musk and the Mail, Express, Telegraph, etc and amplified without any real critique by the likes of the BBC who are too scared to real journalism and present the facts rather than the 'both sides' fallacy. The one exception to this is the wonderful Ros Atkins at the BBC who can cut through all the shite in less than a. minute! Perhaps the Gov should just employ him as Head of Communications?

The big missing step Labour could take is to reverse Brexit and start the process for the UK to join the SM and/or Customs Union, for which there is significant popular support. This would have a huge economic and growth benefit for the UK very quickly.

[Paddington Bear]

It's not exactly a lack of action. They've introduced a bunch of bills and enacted some of the policies they campaigned on. They clearly do have a plan.

The problem is right now they are mostly focusing on the things that will get them better press (so they think) and coming out with hot air over immigration, culture war shit, and other Tory cover band populist nonsense doomed to failure that won't move the needle one bit. Instead of, y'know, putting people first, enacting a progressive agenda, and sticking to their guns rather than getting bogged down in a media war they can't win.

No amount of columns in the express by Starmer will change the fact that they're too chicken shit to push through genuine change that risks upsetting the right wing media and the mythological Tory swing voter.

I think the problem is IF they do have a plan (I'm not convinced as there's a different between a lot of activity and a plan) they can't verbalise it.

People just don't know what Starmer and this government stand for. Why are we making all of these sacrifices? What's the end goal?

Labour did multiple brandings and rebrandings in opposition- none cut through, they ran an election campaign hopelessly devoid of anything but the Tories are awful we can't say anything until we know how bad it is but your tax won't go up yadda yadda.

It’s just consultation after consultation. They’re right that we need planning reform but their solution is more planning officers not a big change in regulation, so at best we’ll have some building sites by the next election. As another example, they’ve put out social care funding to a consultation to report in 2028, by which time we’re into election campaigning and most Local Authorities will have gone bankrupt. It’s madness - they must have discussed in opposition how to fund it and they have a 100 seat majority, just get on with it and you have half a chance of seeing benefits before you’re turfed out

[Sandstorm]

It’s madness - they must have discussed in opposition how to fund it and they have a 100 seat majority, just get on with it and you have half a chance of seeing benefits before you’re turfed out

I hope they did have a plan or two leading up to the election, but the big problem is that the cost of borrowing is higher now that it was for the Tories and Labour are petrified they'll forever be called The Party That Bankrupted Britain or similar when borrowing goes over 5 trillion and the interest rate is 8%.

Of course borrowing now & spending it on infrastructure and building projects will build the economy, but politicians these days are too busy following social media to actually do anything useful.

[shaggy]

I thought the Labour plan was simply redistribution of wealth?

If you look at some of the policies you can see this, inheritance tax on farmers so they have to sell land (hopefully for housing), public services and pay rises, changes to private pensions forcing individuals to spend more, etc. Individually they can be unpopular and create gaps and challenges elsewhere but if you start to combine them it could be seen as stepping stones to enable that transfer of wealth?

And this is only 6 months in, each year there will be changes (large and small) that could add to it.

[Sandstorm]

I thought the Labour plan was simply redistribution of wealth?

If you look at some of the policies you can see this, inheritance tax on farmers so they have to sell land (hopefully for housing), public services and pay rises, changes to private pensions forcing individuals to spend more, etc. Individually they can be unpopular and create gaps and challenges elsewhere but if you start to combine them it could be seen as stepping stones to enable that transfer of wealth?

And this is only 6 months in, each year there will be changes (large and small) that could add to it.

Redistirbution of wealth won't happen. The Rich already have their money tied up in trusts and investments that you can't touch. The only way to move money around is by investing in infrastructure and creating jobs. Labour instead brought in the NI tax uplift which removes jobs from the market overnight. Business owners hate nothing more than a fat wage bill every month, they would 100% automate if the robots were available.

I also like dpedin's idea of revoking Brexit. Labour supported Remain anyway, so there's nothing there the media can target them with and it will help the economy in a big way.

[robmatic]

I thought the Labour plan was simply redistribution of wealth?

If you look at some of the policies you can see this, inheritance tax on farmers so they have to sell land (hopefully for housing), public services and pay rises, changes to private pensions forcing individuals to spend more, etc. Individually they can be unpopular and create gaps and challenges elsewhere but if you start to combine them it could be seen as stepping stones to enable that transfer of wealth?

And this is only 6 months in, each year there will be changes (large and small) that could add to it.

The biggest step that they took in their first budget was to load tax onto middle-income earners though, which shows a different direction of travel.

[Biffer]

I thought the Labour plan was simply redistribution of wealth?

If you look at some of the policies you can see this, inheritance tax on farmers so they have to sell land (hopefully for housing), public services and pay rises, changes to private pensions forcing individuals to spend more, etc. Individually they can be unpopular and create gaps and challenges elsewhere but if you start to combine them it could be seen as stepping stones to enable that transfer of wealth?

And this is only 6 months in, each year there will be changes (large and small) that could add to it.

Redistirbution of wealth won't happen. The Rich already have their money tied up in trusts and investments that you can't touch. The only way to move money around is by investing in infrastructure and creating jobs. Labour instead brought in the NI tax uplift which removes jobs from the market overnight. Business owners hate nothing more than a fat wage bill every month, they would 100% automate if the robots were available.

I also like dpedin's idea of revoking Brexit. Labour supported Remain anyway, so there's nothing there the media can target them with and it will help the economy in a big way.

£14bn investment announced this morning, substantially in building AI data centres and infrastructure in the UK.

[Sandstorm]

£14bn investment announced this morning, substantially in building AI data centres and infrastructure in the UK.

It's a start. But with 1 new hospital costing 1bn , it's a drop in the ocean. Borrow a trillion, get building.

[Tichtheid]

£14bn investment announced this morning, substantially in building AI data centres and infrastructure in the UK.

It's a start. But with 1 new hospital costing 1bn , it's a drop in the ocean. Borrow a trillion, get building.

If only that had happened when borrowing rates were historically low.

There's yer economically sharp party for youse.

[I like neeps]

It's not exactly a lack of action. They've introduced a bunch of bills and enacted some of the policies they campaigned on. They clearly do have a plan.

The problem is right now they are mostly focusing on the things that will get them better press (so they think) and coming out with hot air over immigration, culture war shit, and other Tory cover band populist nonsense doomed to failure that won't move the needle one bit. Instead of, y'know, putting people first, enacting a progressive agenda, and sticking to their guns rather than getting bogged down in a media war they can't win.

No amount of columns in the express by Starmer will change the fact that they're too chicken shit to push through genuine change that risks upsetting the right wing media and the mythological Tory swing voter.

I think the problem is IF they do have a plan (I'm not convinced as there's a different between a lot of activity and a plan) they can't verbalise it.

People just don't know what Starmer and this government stand for. Why are we making all of these sacrifices? What's the end goal?

Labour did multiple brandings and rebrandings in opposition- none cut through, they ran an election campaign hopelessly devoid of anything but the Tories are awful we can't say anything until we know how bad it is but your tax won't go up yadda yadda.

It’s just consultation after consultation. They’re right that we need planning reform but their solution is more planning officers not a big change in regulation, so at best we’ll have some building sites by the next election. As another example, they’ve put out social care funding to a consultation to report in 2028, by which time we’re into election campaigning and most Local Authorities will have gone bankrupt. It’s madness - they must have discussed in opposition how to fund it and they have a 100 seat majority, just get on with it and you have half a chance of seeing benefits before you’re turfed out

They don't want to reform them because the reform options will be very unpopular (with tory and maybe Lib Dem) voters. It's why we need politicians with intellectual honesty but sadly we don't have any.

[Slick]

I thought the Labour plan was simply redistribution of wealth?

If you look at some of the policies you can see this, inheritance tax on farmers so they have to sell land (hopefully for housing), public services and pay rises, changes to private pensions forcing individuals to spend more, etc. Individually they can be unpopular and create gaps and challenges elsewhere but if you start to combine them it could be seen as stepping stones to enable that transfer of wealth?

And this is only 6 months in, each year there will be changes (large and small) that could add to it.

Redistirbution of wealth won't happen. The Rich already have their money tied up in trusts and investments that you can't touch. The only way to move money around is by investing in infrastructure and creating jobs. Labour instead brought in the NI tax uplift which removes jobs from the market overnight. Business owners hate nothing more than a fat wage bill every month, they would 100% automate if the robots were available.

I also like dpedin's idea of revoking Brexit. Labour supported Remain anyway, so there's nothing there the media can target them with and it will help the economy in a big way.

They ran their election campaign on not revoking Brexit, so there is that for the media to get stuck in about.

[Sandstorm]

They ran their election campaign on not revoking Brexit, so there is that for the media to get stuck in about.

"We didn't want to kill any more ancient Tory voters from shock, so lied about it. Sorry. Not Sorry"

[JM2K6]

Somehow I missed the news about the AI investment and "having a government be on the side of the AI companies".

Fucking LOL, absolutely clueless shit. Dreadful, dreadful plan. Embarrassingly stupid.

[Paddington Bear]

The bond market issue is not going away, and if it carries on for much longer every single idea and policy they’ve ever dreamed of will be subordinated to getting it under control

[ScarfaceClaw]

Somehow I missed the news about the AI investment and "having a government be on the side of the AI companies".

Fucking LOL, absolutely clueless shit. Dreadful, dreadful plan. Embarrassingly stupid.

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

[Biffer]

Somehow I missed the news about the AI investment and "having a government be on the side of the AI companies".

Fucking LOL, absolutely clueless shit. Dreadful, dreadful plan. Embarrassingly stupid.

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

[JM2K6]

Somehow I missed the news about the AI investment and "having a government be on the side of the AI companies".

Fucking LOL, absolutely clueless shit. Dreadful, dreadful plan. Embarrassingly stupid.

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

[fishfoodie]

Somehow I missed the news about the AI investment and "having a government be on the side of the AI companies".

Fucking LOL, absolutely clueless shit. Dreadful, dreadful plan. Embarrassingly stupid.

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

It's again focusing on the wrong side of the equation so !

The different Cloud Providers are fighting each other to get an allocation of nVidia chips, & anyone else is left standing around with their dicks in the hands. It's a fools errand to go looking to build out any serious HW capacity for the next few years, because nVidia have sold their next couple of years production already.

On the other hand what is worthwhile is training people !

There are a lot of ideas on how to utilize AI, but bugger all people with the skills to work out how to implement the idea, once they get the HW.

For example, if you want to track potholes, or what roads to repair, then you'd need a couple of datasets like google streetview, & the various councils repair data, & then you'd need to train the code on what a pothole looks like, or various other things like worn tarmac, or cracks etc, & then kick it off to scan the streetview pictures to identify the worst roads that hadn't been resurfaced since the images were captured.

All of this takes people with a high level of technical aptitude & a knowledge of what the tech can, & cannot do, & they're rare as hens teeth.

If you can build a pool of these people, then they'll not just fix the UKs potholes, but they'll have shown that they can do the same for cities around the world, & that is the way to earn money !

[_Os_]

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

I bow to your knowledge on AI. The scale of the benefits certainly haven't remotely matched the costs so far, as things stand the definition of a bubble.

But there's a larger point. If you want some state led development/industrialisation/infrastructure, then there's going to be failures along the way. How people handle the failures is the real test of it's viability in the UK. If "Labour pissed away x million or billion on AI" becomes a strong attack line and vote winner, then culturally it's not possible for the state to attempt to do anything which could grow/direct the economy in a way the private sector cannot or has failed to do. Which means nothing changes. East Asia has a lot of state intervention and is hardly unsuccessful, but they've had failures too, Singapore's national shipping line no longer exists, South Korean chaebols can collapse like Daewoo.

The plans around green energy and industrialisation could work but need 5 years. The UK has a lot of wind for wind farms, Wales and Scotland have a lot of mountains and rain for pumped storage. Seems reasonable the UK could have plentiful cheap energy, even failure should still provide some wins as these are natural advantages. But they're not quick wins.

Labour's plans are ambitious, not because they are actually ambitious, but because all of it has to work perfectly to avoid punishment from voters. If culturally the UK only accepts the state as a provider of basic services, which means tolerating failure (the NHS is clearly failing but remains strongly supported), then it's probably bad for the Union too. The death of nationalised industries is linked to the growth of Celtic nationalism.

Bond yields rising, growth needed asap. Labour will get desperate if this holds. Forced by the markets the only answer will then become the opposite of Levelling Up, which means directing as much investment and immigration into London/South East as possible. Labour aren't going to say that because it's unpopular, but it's the surest bet for some quick growth. The Tories were forced into the same corner after Brexit, they achieved stagnation.

[_Os_]

[For example, if you want to track potholes, or what roads to repair, then you'd need a couple of datasets like google streetview, & the various councils repair data, & then you'd need to train the code on what a pothole looks like, or various other things like worn tarmac, or cracks etc, & then kick it off to scan the streetview pictures to identify the worst roads that hadn't been resurfaced since the images were captured.

All of this takes people with a high level of technical aptitude & a knowledge of what the tech can, & cannot do, & they're rare as hens teeth.

If you can build a pool of these people, then they'll not just fix the UKs potholes, but they'll have shown that they can do the same for cities around the world, & that is the way to earn money !

AI as a solution for potholes ... is just total madness.

In South Africa it works like this. DA run metro = residents can report a pothole by sending a photo and details, it is then put on a waiting list and repaired asap (weeks not months). ANC run metro = a lot more hit and miss, residents sometimes fill them in themselves, spray painting a circle around the pothole with an arrow pointing at it captioned "ANC" has been proven to work.

My bet is the UK's pothole problem is mostly about lack of state capacity, and private contractors gouging to the point where potholes have become too expensive. My bet is not that there's unused capacity ready to go if only they knew where the potholes were, and only HAL from 2001 A Space Odyssey can possibly locate these hidden potholes. Interested in what HAL has to say though: "Call me Dave, you implemented austerity, those doing this work were axed and contractors were used instead who overcharge, often the exact same people wearing a different hat. This in a nutshell is how you implemented austerity and grew the debt at the same time, Dave. The solution is kill the contractors, Dave.".

I was talking to a landscaper recently, he did a one day job for the NHS cutting back trees and charged his usual rate of £500. The NHS official hiring/paying was shocked he wasn't charging £2000+ like everyone else.

[TheNatalShark]

Are we bringing back the cones hotline

[Biffer]

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

It's again focusing on the wrong side of the equation so !

The different Cloud Providers are fighting each other to get an allocation of nVidia chips, & anyone else is left standing around with their dicks in the hands. It's a fools errand to go looking to build out any serious HW capacity for the next few years, because nVidia have sold their next couple of years production already.

On the other hand what is worthwhile is training people !

There are a lot of ideas on how to utilize AI, but bugger all people with the skills to work out how to implement the idea, once they get the HW.

For example, if you want to track potholes, or what roads to repair, then you'd need a couple of datasets like google streetview, & the various councils repair data, & then you'd need to train the code on what a pothole looks like, or various other things like worn tarmac, or cracks etc, & then kick it off to scan the streetview pictures to identify the worst roads that hadn't been resurfaced since the images were captured.

All of this takes people with a high level of technical aptitude & a knowledge of what the tech can, & cannot do, & they're rare as hens teeth.

If you can build a pool of these people, then they'll not just fix the UKs potholes, but they'll have shown that they can do the same for cities around the world, & that is the way to earn money !

Again, if you’d bother to read the government proposal, training people is one of the core activities. It’s a fundamental part of building the infrastructure in the UK. That’s a vital recognition, by government, that tech infrastructure isn’t just boxes and buildings. That’s not been acknowledged previously.

But you know, you do you, just react to what’s fed to you by your particular news source.

It’s only 25 pages and it’s not small type. Took me just over half an hour to read through it yesterday. There’s stuff in it that can be criticised, and I would, but there’s a lot of good stuff too.

[dpedin]

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

40 years ago I did research into the impact of new IT technologies into the office and productivity. The introduction of word processing was going to revolutionize the productivity of the office and get rid of lots and lots of admin staff. However what I found was instead of getting a document 99% right before it was typed up managers just drafted some rough old thing and then got it 'word processed' and then went through a number of redrafts, because it was easy to do now, until it was acceptable. The end result no productivity increase in terms of labour cost reductions nor in time taken to knock out a final product but an increase in IT costs! Ok, over time that position will have changed and there will have been productivity improvements but came across same issue when I looked at the into of speech recognition into the NHS and the ideal that it would reduce dramatically the cost and time associated with consultants getting reports etc typed up.

The bottom line is there are hardware, software and human ware issues with IT implementation. Hardware costs have fallen dramatically and we rarely full utilize the capability of the hardware we purchase,, software costs are still high but how we pay for them has changed ie by selling our data, however the human ware issues are the biggest barrier to implementation and seeing the cost reductions and improved productivity. As usual the investment focuses on the hardware and software and it is often the human ware that has least invested in it when it should be the opposite. For this reason I suspect any optimistic view about the returns on AI will be grossly exaggerated in terms of payback time.

[Biffer]

Included in possible proposals is using AI to identify potholes and poor roads needing maintenance. I don’t think the UK has anywhere near the computing power to handle that sort of data processing and load.

You wouldn’t even really need AI for that. An AA roadmap with a post-it note saying “all these” would suffice.

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

1. Whatever tech succeeds over the next ten years it’s likely to need massively more compute and stuff built for AI doesn’t have to be used for that
2. Training people in AI gives them core skills that are easily re used in other tech areas
3. In our fucked up world, sovereign capability is going to become more and more important


I think what we’ll see, is multiple announcements similar to this for different deep tech areas. In fact I didn’t see anyone complaining about the £2.5billion going into quantum tech a month ago, but that’s probably because people don’t understand it. I’m expecting similar plans on robotics, space, advanced pharma, advanced manufacturing, defence, clean energy and probably some others. Any of these could be called buzzwords, bubbles or fads depending on your own personal views, and some of them will be. But all of them involve training more and more people on modern technical science and engineering skills, for all levels from basic technician level through to advanced research and design. We need more skills in quantum, we need more skills in space, we need more skills in robotics, we need more skills in AI. A huge slab of those are just engineering - mechanical, electrical, systems, optical, process, production engineers. Doing that training across multiple sectors means we’re equipped to fulfil our needs in whichever of those prove the most successful.

[Biffer]

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

40 years ago I did research into the impact of new IT technologies into the office and productivity. The introduction of word processing was going to revolutionize the productivity of the office and get rid of lots and lots of admin staff. However what I found was instead of getting a document 99% right before it was typed up managers just drafted some rough old thing and then got it 'word processed' and then went through a number of redrafts, because it was easy to do now, until it was acceptable. The end result no productivity increase in terms of labour cost reductions nor in time taken to knock out a final product but an increase in IT costs! Ok, over time that position will have changed and there will have been productivity improvements but came across same issue when I looked at the into of speech recognition into the NHS and the ideal that it would reduce dramatically the cost and time associated with consultants getting reports etc typed up.

The bottom line is there are hardware, software and human ware issues with IT implementation. Hardware costs have fallen dramatically and we rarely full utilize the capability of the hardware we purchase,, software costs are still high but how we pay for them has changed ie by selling our data, however the human ware issues are the biggest barrier to implementation and seeing the cost reductions and improved productivity. As usual the investment focuses on the hardware and software and it is often the human ware that has least invested in it when it should be the opposite. For this reason I suspect any optimistic view about the returns on AI will be grossly exaggerated in terms of payback time.

Do you think large businesses, right now, would be more or less productive if you took all the word processing software out?

[Raggs]

Do you think large businesses, right now, would be more or less productive if you took all the word processing software out?

Was just thinking exactly this. Make every document need to be typed on a typewriter and lets see if there's not many/any time savings by doing it in a word processor across the whole business front (so no email attachments etc).

It takes humans time to adapt properly to new tech, and yes, at first it's likely no better, if not worse than what was in place, but it gets better, and surely having decently trained users of that tech is going to be quite handy in the future.

Sure, right now having AI spot potholes is laughable. But in the future? Building from that work? Able to spot which roads are getting more traffic, heavier vehicles, bad weather, spotting cracks before they become potholes (so before anyone would likely report them). There's clearly a good direction for this to go in.

When the AI "bubble" goes pop, there's still very much going to be AI, and it's going to continue to improve, just like the dotcom bubble, overly valued early on sure, but there's a reason why plenty of the richest people in the world owe a lot of that to current or more recent internet businesses.

[JM2K6]

If you'd read the proposal, it's to build the compute power in the UK as sovereign capability.

However Starmer did come out with some pish today. But the fundamental investments of increasing UK infrastructure, creating jobs related to that infrastructure and developing the skills for those jobs is all good. Even if AI doesn't deliver all these things, the compute and skills are going to have positive economic impact.

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

1. Whatever tech succeeds over the next ten years it’s likely to need massively more compute and stuff built for AI doesn’t have to be used for that
2. Training people in AI gives them core skills that are easily re used in other tech areas
3. In our fucked up world, sovereign capability is going to become more and more important


I think what we’ll see, is multiple announcements similar to this for different deep tech areas. In fact I didn’t see anyone complaining about the £2.5billion going into quantum tech a month ago, but that’s probably because people don’t understand it. I’m expecting similar plans on robotics, space, advanced pharma, advanced manufacturing, defence, clean energy and probably some others. Any of these could be called buzzwords, bubbles or fads depending on your own personal views, and some of them will be. But all of them involve training more and more people on modern technical science and engineering skills, for all levels from basic technician level through to advanced research and design. We need more skills in quantum, we need more skills in space, we need more skills in robotics, we need more skills in AI. A huge slab of those are just engineering - mechanical, electrical, systems, optical, process, production engineers. Doing that training across multiple sectors means we’re equipped to fulfil our needs in whichever of those prove the most successful.

I partly agree with 1) - the requirements for an AI DC are quite different to regular compute, but it's not a totally different industry. I completely disagree with 2) and would argue that the focus is on exactly the wrong kind of AI - the brain-dead LLMs etc al - and away from the useful machine learning stuff. 3) is a no brainer.

"Hey, we really need to improve our tech infrastructure especially as we used to be really good at this stuff so we're going to do something massive in Manchester, Newcastle, etc because it's about time we had our own infrastructure that isn't just an extension of AWS / GCP / whatever" would be great, no argument there.

[Biffer]

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

1. Whatever tech succeeds over the next ten years it’s likely to need massively more compute and stuff built for AI doesn’t have to be used for that
2. Training people in AI gives them core skills that are easily re used in other tech areas
3. In our fucked up world, sovereign capability is going to become more and more important


I think what we’ll see, is multiple announcements similar to this for different deep tech areas. In fact I didn’t see anyone complaining about the £2.5billion going into quantum tech a month ago, but that’s probably because people don’t understand it. I’m expecting similar plans on robotics, space, advanced pharma, advanced manufacturing, defence, clean energy and probably some others. Any of these could be called buzzwords, bubbles or fads depending on your own personal views, and some of them will be. But all of them involve training more and more people on modern technical science and engineering skills, for all levels from basic technician level through to advanced research and design. We need more skills in quantum, we need more skills in space, we need more skills in robotics, we need more skills in AI. A huge slab of those are just engineering - mechanical, electrical, systems, optical, process, production engineers. Doing that training across multiple sectors means we’re equipped to fulfil our needs in whichever of those prove the most successful.

I partly agree with 1) - the requirements for an AI DC are quite different to regular compute, but it's not a totally different industry. I completely disagree with 2) and would argue that the focus is on exactly the wrong kind of AI - the brain-dead LLMs etc al - and away from the useful machine learning stuff. 3) is a no brainer.

"Hey, we really need to improve our tech infrastructure especially as we used to be really good at this stuff so we're going to do something massive in Manchester, Newcastle, etc because it's about time we had our own infrastructure that isn't just an extension of AWS / GCP / whatever" would be great, no argument there.

So where in the plan does it say they’re concentrating on LLM?

Oh that’s right, you haven’t read it. It doesn’t say that.

And btw on the environmental cost, it specifically highlights coastal Scotland to site some of this stuff as then you have freely available low carbon electricity. Again, read the fucking thing before banging in about why it’s shit.

[JM2K6]

1. Whatever tech succeeds over the next ten years it’s likely to need massively more compute and stuff built for AI doesn’t have to be used for that
2. Training people in AI gives them core skills that are easily re used in other tech areas
3. In our fucked up world, sovereign capability is going to become more and more important


I think what we’ll see, is multiple announcements similar to this for different deep tech areas. In fact I didn’t see anyone complaining about the £2.5billion going into quantum tech a month ago, but that’s probably because people don’t understand it. I’m expecting similar plans on robotics, space, advanced pharma, advanced manufacturing, defence, clean energy and probably some others. Any of these could be called buzzwords, bubbles or fads depending on your own personal views, and some of them will be. But all of them involve training more and more people on modern technical science and engineering skills, for all levels from basic technician level through to advanced research and design. We need more skills in quantum, we need more skills in space, we need more skills in robotics, we need more skills in AI. A huge slab of those are just engineering - mechanical, electrical, systems, optical, process, production engineers. Doing that training across multiple sectors means we’re equipped to fulfil our needs in whichever of those prove the most successful.

I partly agree with 1) - the requirements for an AI DC are quite different to regular compute, but it's not a totally different industry. I completely disagree with 2) and would argue that the focus is on exactly the wrong kind of AI - the brain-dead LLMs etc al - and away from the useful machine learning stuff. 3) is a no brainer.

"Hey, we really need to improve our tech infrastructure especially as we used to be really good at this stuff so we're going to do something massive in Manchester, Newcastle, etc because it's about time we had our own infrastructure that isn't just an extension of AWS / GCP / whatever" would be great, no argument there.

So where in the plan does it say they’re concentrating on LLM?

Oh that’s right, you haven’t read it. It doesn’t say that.

And btw on the environmental cost, it specifically highlights coastal Scotland to site some of this stuff as then you have freely available low carbon electricity. Again, read the fucking thing before banging in about why it’s shit.

Correct, I haven't read it, I've only read their public communications about it. When I get a chance I will.

[Biffer]

I partly agree with 1) - the requirements for an AI DC are quite different to regular compute, but it's not a totally different industry. I completely disagree with 2) and would argue that the focus is on exactly the wrong kind of AI - the brain-dead LLMs etc al - and away from the useful machine learning stuff. 3) is a no brainer.

"Hey, we really need to improve our tech infrastructure especially as we used to be really good at this stuff so we're going to do something massive in Manchester, Newcastle, etc because it's about time we had our own infrastructure that isn't just an extension of AWS / GCP / whatever" would be great, no argument there.

So where in the plan does it say they’re concentrating on LLM?

Oh that’s right, you haven’t read it. It doesn’t say that.

And btw on the environmental cost, it specifically highlights coastal Scotland to site some of this stuff as then you have freely available low carbon electricity. Again, read the fucking thing before banging in about why it’s shit.

Correct, I haven't read it, I've only read their public communications about it. When I get a chance I will.

Their comms weren't the best tbf. The plan doesn't mention pish like potholes.

[Tichtheid]

AI can play a significant role in improving the process of identifying, managing, and fixing potholes in several ways. Here are some approaches:

1. Detection and Monitoring
AI can help identify potholes quickly and accurately using various sensors and data sources. Some potential methods include:

Computer Vision (CV): Using cameras mounted on vehicles or drones, AI-powered computer vision systems can automatically detect potholes by analyzing images or video feeds in real time. These systems can identify cracks, deformations, and other road surface irregularities that indicate potholes. This can be done using deep learning models trained to recognize potholes in various conditions (lighting, weather, etc.).
Mobile Apps and Crowdsourcing: Smartphones equipped with accelerometers and GPS can collect data when a vehicle passes over a pothole. AI can analyze this data to pinpoint the exact location of potholes and assess their severity. Apps like Google Maps can then provide real-time reports of potholes to authorities or maintenance teams.
Drones and Satellite Imaging: AI can process images or videos captured by drones and satellites to detect road damages over large areas, especially in hard-to-reach or rural locations. This can be combined with historical data to predict when and where potholes are likely to form.

2. Predictive Maintenance
AI can analyze data from multiple sources, such as road conditions, weather patterns, traffic data, and historical repair information, to predict where potholes are likely to form in the future. Machine learning algorithms can model and forecast road degradation, allowing cities to proactively schedule repairs before potholes develop into major problems.

Sensor Data from Roads: Some roads are equipped with embedded sensors that collect data on stress, temperature, moisture, and traffic load. AI can process this data to predict areas that are more likely to develop potholes, based on conditions like excessive traffic, freeze-thaw cycles, or other stress factors.
Historical Data Analysis: AI can analyze patterns from past maintenance logs, weather data, and road usage to determine which areas of a city or highway are at the greatest risk of pothole formation. This predictive approach helps prioritize repairs and allocate resources more efficiently.

3. Autonomous Pothole Repair Machines
AI-powered autonomous vehicles or machines could be used to fix potholes automatically. These machines could be equipped with robotic arms, sensors, and repair materials to detect, fill, and smooth over potholes with minimal human intervention. The steps involved might include:

Inspection: Drones or ground robots inspect the road surface, detecting potholes in real time.
Automated Repair: Once a pothole is detected, an autonomous machine could be dispatched to the location. The machine could be programmed to fill the hole with appropriate material (such as asphalt or concrete), level it, and ensure it is properly compacted.
These machines would rely on AI algorithms for path planning, damage assessment, material selection, and real-time decision-making.

4. Robotic Road Repair Systems
AI-driven robotic systems can be deployed for faster, more efficient pothole repair. For example, robots could use AI to determine the size and depth of a pothole, then choose the best repair technique (e.g., cold patch, hot mix) and apply it accurately.

Autonomous Road Repair Vehicles (ARVs): AI-powered vehicles can use computer vision and machine learning algorithms to assess pothole damage, apply the necessary patching material, and even compact it automatically. This could save time and reduce human error during repairs.
Integrated Repair Systems: AI could coordinate the work of different robots or machines (e.g., one for cleaning out debris, another for applying the patch, and a third for compaction) to ensure that repairs are done quickly and effectively.

5. Optimizing Resource Allocation
AI can optimize the deployment of repair crews and materials. By analyzing real-time data on potholes and traffic patterns, AI can help authorities prioritize which potholes need immediate attention and which can wait. This leads to better resource management and faster repairs.

Routing and Scheduling: AI can help plan the most efficient routes for repair crews, minimizing downtime and reducing the time it takes to repair potholes. It can also ensure that the right amount of materials are available at the right location at the right time.

6. Quality Control and Improvement
After potholes are repaired, AI systems can monitor the quality of the repair by using sensors and vision systems. AI can assess whether the patch is solid and properly compacted, and if it has been effectively integrated with the surrounding road surface.

Feedback Loops: AI could be used to evaluate the success of repairs by analyzing long-term data (such as how quickly new potholes appear in the same location) and adjust the repair process or materials used to improve future outcomes.

7. Public Reporting and Engagement
AI-powered platforms can facilitate public reporting of potholes. Citizens can use apps or online platforms to submit pothole locations, and AI can help prioritize these reports based on factors like severity, location, and traffic data. AI could also help cities automatically verify reports using satellite images or vehicle-based data.

Benefits of Using AI for Pothole Repair:
Faster detection and response times: AI can detect potholes almost in real time, leading to quicker repairs.
Cost-effectiveness: Predictive AI systems can reduce unnecessary repairs by focusing resources on areas with the greatest need.
Proactive maintenance: AI can predict and prevent future potholes, reducing long-term infrastructure costs.
Efficient resource allocation: AI optimizes the deployment of repair crews and materials, improving operational efficiency.
Improved safety: Timely repairs ensure that roads are safer for drivers and pedestrians.

Challenges and Considerations:
Data Quality: The success of AI systems depends on accurate, high-quality data. Poor data collection or incomplete sensor coverage could lead to missed potholes or incorrect predictions.
Cost and Infrastructure: Setting up AI-powered systems and deploying autonomous repair vehicles or drones can be expensive, requiring significant upfront investment.
Weather and Environmental Factors: AI systems must be resilient to weather conditions like rain, snow, and fog, which can impact both detection and repair processes.
In summary, AI can revolutionize pothole detection, repair, and maintenance, making it faster, more efficient, and more cost-effective. By combining sensors, predictive analytics, and autonomous machines, cities can not only address existing potholes more effectively but also prevent future road degradation.


I got the above by putting "How can AI be used to fix potholes" into ChatGBT, it took about three seconds to churn that out.

[Sandstorm]

AI can play a significant role in improving the process of identifying, managing, and fixing potholes in several ways.

Challenges and Considerations:

This is being funded and rolled out in the UK and will suffer from delays, cost over-runs and ultimately failure due to poor project planning and zero leadership. There will also be an expensive & delayed enquiry into the fiasco, at the end of which no money will be returned, no heads will roll & zero lessons will be learned.

Fixed

[Paddington Bear]

If you built a vaguely user friendly website that allowed people to report potholes with an exact location, size and photo, I can guarantee every bloke named Trevor aged 50-70 in this nation will have ensured that 100% of potholes are recorded and fully documented within a month

[shaggy]

https://www.fillthathole.org.uk/

Used this for over 20 years. Local authorities just ignore it.

[fishfoodie]

If you built a vaguely user friendly website that allowed people to report potholes with an exact location, size and photo, I can guarantee every bloke named Trevor aged 50-70 in this nation will have ensured that 100% of potholes are recorded and fully documented within a month

.... and then he'd go back to his usual haunt on a bridge, trainspotting .....

I only mentioned potholes, as someone did above, but there are more worthy uses of AI, such as 1st level medical image screening for cancers, & drug modelling etc, etc, as well as boring stuff like traffic monitoring, & logistics planning. Even Space Karen sees this as somewhere he can make some money, as he invited people to upload their medical imagery to his Nazi AI, so he can steal your data & ignore HIPAA regulations

If you have a huge amount of data, & you need excruciating detail, but some degree of fuzzy logic, then that's a niche that AI can fit into, but at a cost, because it certainly ain't cheap.

[Paddington Bear]

https://www.fillthathole.org.uk/

Used this for over 20 years. Local authorities just ignore it.

Suppose I should have added ‘if Trevor could have a high degree of confidence it will then be fixed’

[dpedin]

AI is a massive bubble and it's pissing money down the drain on something that will have a huge environmental cost.

40 years ago I did research into the impact of new IT technologies into the office and productivity. The introduction of word processing was going to revolutionize the productivity of the office and get rid of lots and lots of admin staff. However what I found was instead of getting a document 99% right before it was typed up managers just drafted some rough old thing and then got it 'word processed' and then went through a number of redrafts, because it was easy to do now, until it was acceptable. The end result no productivity increase in terms of labour cost reductions nor in time taken to knock out a final product but an increase in IT costs! Ok, over time that position will have changed and there will have been productivity improvements but came across same issue when I looked at the into of speech recognition into the NHS and the ideal that it would reduce dramatically the cost and time associated with consultants getting reports etc typed up.

The bottom line is there are hardware, software and human ware issues with IT implementation. Hardware costs have fallen dramatically and we rarely full utilize the capability of the hardware we purchase,, software costs are still high but how we pay for them has changed ie by selling our data, however the human ware issues are the biggest barrier to implementation and seeing the cost reductions and improved productivity. As usual the investment focuses on the hardware and software and it is often the human ware that has least invested in it when it should be the opposite. For this reason I suspect any optimistic view about the returns on AI will be grossly exaggerated in terms of payback time.

Do you think large businesses, right now, would be more or less productive if you took all the word processing software out?

I can guarantee right now that the average user in the average large business uses about 25% or less of the capability of word or excel etc. The question is how much more productive should the large business be if it used the current software to its max capability? Most folk will have a work PC or laptop that is specced far too high for what they need and use it for. Most large businesses find it easier to buy lots of kit and software and then never review its actual usage or their ROI for that significant investment. They will invest far, far less in training and work/process redesign than they do in hardware and software, always have and always will.