Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
Wales receives £120 for every £100 in England due to the Barnett formula
And its even more than that up here north of the wall. And yet our councils are absolutely broke because that cash isn't available to them.
We have two problems - it costs too much money to do anything, and we don't put up enough cash to cover the bill. Hence councils and public services having a dangerous lack of cash.
The solution is not to vote out the respective governments. That doesn't change the maths - we need to do things cheaper and pool budgets. Cutting the number of teachers cos we can't afford it costs us more money. And yet regardless of party they get stuck in this "cut the service and the need goes away" nonsense.
Yes - important to note that council's have had a really rough time north of the border too.
It's why the public realm looks so rubbish. And the whining about cycle lanes is misplaced, because even that comes from direct ring-fenced central government funding.
Bernard Arnault is exploring a potential acquisition of struggling French Artificial Intelligence company Mistral citing a need for France to "maintain it's edge in artisanal goods post ASI." https://x.com/techbrospod/status/1883952340023280083
Mistral just raised €600m at a €5.8bn valuation.
France is working on an authentically French rival to Deepseek, but every time they try to launch it, it goes on strike.
Won’t they just regulate it out of existence ?
France’s corporate sector has done pretty well in recent years. It manages to have twice as many companies in the Fortune 500 as Britain, and was briefly ahead of Germany on that measure, but is behind us in number of tech unicorns.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
It's all going on social care Malcolm, after years of council tax freezes and real terms cuts to general grant. You can moan about the Scottish Government and the NHS all you want - councils really deserve a pass IMO.
I need to stop checking if prices have tumbled at scan.co.uk... I think the speculation is actually going to boost demand in the short term, damn it.
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
BREAKING: Per multiple law enforcement sources, within the last hour, Border Patrol agents near Fronton, TX were fired upon from MX by suspected cartel gunmen as a group of illegal aliens were being brought across the river. I’m told BP returned fire, nobody hit on either side, and that the illegal aliens did not make it across.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
I've got a well-run local council who despite the real terms cuts in funding from central govt have managed to keep libraries and leisure centres open, services running and roads repaired, despite the increasingly heavy hgvs that destroy manholes, collapse drains and any areas washed out by mains water leaks on our road. I live in fear of increasing support for the lib dems locally, they run the neighbouring council and levelled their local leisure centre before realising they couldn't afford to replace it amongst other examples of incompetence.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
You have a point. The elected councillors, who are supposed to act like non-exec directors and protect the taxpayer interest, rarely do so. Most are simply unqualified to challenge and scrutinise. The officers run the show and all too often local authorities are, effectively, "captured" by their employees and are run on their behalf. It is almost impossible to get rid of poor performing senior dept heads.
But, tbf, they have had their funding cut too. I can't help thinking that the only hope for change is exec mayors who are much more accountable to the public.
Just fyi, @deepseek_ai collects your IP, keystroke patterns, device info, etc etc, and stores it in China, where all that data is vulnerable to arbitrary requisition from the 🇨🇳 State.
Errrr
Anyone can download the Deepseek model, and run it themselves using Ollama on their local computer. And you can even do it with no internet connection.
I actually got Deepseek (1.5bn) running successfully on my phone with acceptable performance levels. (Obviously, I have insanely powerful phone. But still.)
You lost me with Ollama. I’m well known offline for my lack of technical chops.
I get you can run it locally, but I still don’t trust it, especially since at the end of the day none of my devices are permanently offline.
The models are just like massive Excel formulas. They can't 'do' anything apart from take your input, and return some output. If anything, I trust them more than Excel these days.
I know of a couple of writers who use local models via Ollama as they need to write 'adult' scenes (whether murder/mystery or kissy-fun-times) and the commercial ones just point-blank refuse.
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
I've got a well-run local council who despite the real terms cuts in funding from central govt have managed to keep libraries and leisure centres open, services running and roads repaired, despite the increasingly heavy hgvs that destroy manholes, collapse drains and any areas washed out by mains water leaks on our road. I live in fear of increasing support for the lib dems locally, they run the neighbouring council and levelled their local leisure centre before realising they couldn't afford to replace it amongst other examples of incompetence.
LibDems are great at putting out leaflets and being "community".
Less good at running multi-million pound budget service providers.
I’d like to download DeepSeek but I’m worried about Chinese surveillance.
My attempts to get Claude to write spy thrillers haven’t been terribly successful, to be honest.
It's open source, you can download and spin it up yourself if you've got a modern GPU.
You can run it on your phone, assuming you have Android. I get remarkably snappy performance from the 1.5bn parameter model.
Nice, what's the word on Apple silicon? I just got my M4 Pro laptop delivered last week, finally replaced my old M1 Pro now that I don't have a laptop from work.
You will certainly be able to run the 8bn parameter model, and may well be able to run the 14bn one.
Just fyi, @deepseek_ai collects your IP, keystroke patterns, device info, etc etc, and stores it in China, where all that data is vulnerable to arbitrary requisition from the 🇨🇳 State.
Errrr
Anyone can download the Deepseek model, and run it themselves using Ollama on their local computer. And you can even do it with no internet connection.
I actually got Deepseek (1.5bn) running successfully on my phone with acceptable performance levels. (Obviously, I have insanely powerful phone. But still.)
You lost me with Ollama. I’m well known offline for my lack of technical chops.
I get you can run it locally, but I still don’t trust it, especially since at the end of the day none of my devices are permanently offline.
OK.
Ollama is a platform for running LLMs (i.e. neural nets) on.
Deepseek or LLama or any of these models are just a series of connected nodes with weights attached to them. They have no ability to do anything other than to spit out tokens. (Technically, they take a bunch of tokens and predict the next token. Then the next token. Then the next token.)
Now obviously if you use someone's hosted service, whether Claude or ChatGPT or DeepSeek, then they will likely store everything you tell them,
But the model itself has no ability to do that, because it's just a bunch of nodes with weights, that calculated that if the two tokens it has been passed are "I" and "am", then there's an 87% chance the next word should be "a".
I shall give it a go - I think a 16GB video card should run it OK. Might use a sandbox for paranoia's sake...
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
It's all going on social care Malcolm, after years of council tax freezes and real terms cuts to general grant. You can moan about the Scottish Government and the NHS all you want - councils really deserve a pass IMO.
In general, that's probably true. But there can be a very large difference in outcomes between well and badly run councils. Albeit complicated by particular local circumstances.
Just fyi, @deepseek_ai collects your IP, keystroke patterns, device info, etc etc, and stores it in China, where all that data is vulnerable to arbitrary requisition from the 🇨🇳 State.
Errrr
Anyone can download the Deepseek model, and run it themselves using Ollama on their local computer. And you can even do it with no internet connection.
I actually got Deepseek (1.5bn) running successfully on my phone with acceptable performance levels. (Obviously, I have insanely powerful phone. But still.)
You lost me with Ollama. I’m well known offline for my lack of technical chops.
I get you can run it locally, but I still don’t trust it, especially since at the end of the day none of my devices are permanently offline.
The models are just like massive Excel formulas. They can't 'do' anything apart from take your input, and return some output. If anything, I trust them more than Excel these days.
I know of a couple of writers who use local models via Ollama as they need to write 'adult' scenes (whether murder/mystery or kissy-fun-times) and the commercial ones just point-blank refuse.
Have they considered “writing” the scenes in question?
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
That's the problem
Yesterday, Nvidia claimed you could only run the models on their expensive hardware.
Today you can run them on a phone.
That's a problem for people who make really expensive hardware...
I’d like to download DeepSeek but I’m worried about Chinese surveillance.
My attempts to get Claude to write spy thrillers haven’t been terribly successful, to be honest.
It's open source, you can download and spin it up yourself if you've got a modern GPU.
You can run it on your phone, assuming you have Android. I get remarkably snappy performance from the 1.5bn parameter model.
Nice, what's the word on Apple silicon? I just got my M4 Pro laptop delivered last week, finally replaced my old M1 Pro now that I don't have a laptop from work.
An M4 Pro with lots of memory is an ideal platform to run current LLMs at a usable clip. LLMs really, really like lots of memory bandwidth, and short of buying extremely expensive custom hardware M4 Pro macs have the best memory bandwidth you can buy in a consumer accessible package if your LLM of choice doesn’t fit in the 8-12Gb of a typical GPU. (24 GB GPUs are available, but at vast expense & those aren’t big enough to run the 70B weight LLMs.)
I got mine with 48GB RAM so I should be fine for memory allocation, half of why I've spent the better part of £3k on this is so that I can run some cool LLMs locally.
Just fyi, @deepseek_ai collects your IP, keystroke patterns, device info, etc etc, and stores it in China, where all that data is vulnerable to arbitrary requisition from the 🇨🇳 State.
Errrr
Anyone can download the Deepseek model, and run it themselves using Ollama on their local computer. And you can even do it with no internet connection.
I actually got Deepseek (1.5bn) running successfully on my phone with acceptable performance levels. (Obviously, I have insanely powerful phone. But still.)
You lost me with Ollama. I’m well known offline for my lack of technical chops.
I get you can run it locally, but I still don’t trust it, especially since at the end of the day none of my devices are permanently offline.
The models are just like massive Excel formulas. They can't 'do' anything apart from take your input, and return some output. If anything, I trust them more than Excel these days.
I know of a couple of writers who use local models via Ollama as they need to write 'adult' scenes (whether murder/mystery or kissy-fun-times) and the commercial ones just point-blank refuse.
Have they considered “writing” the scenes in question?
Yes. Much like they've considered going back to a mechanical typewriter or quill. Or programmers consider punch-cards. There are bits of writing they just don't enjoy ... writing. So getting a first draft from an LLM is fine for them.
Imagine you were the poor sod who had to write leon's posts. You could write them out long-hand like a savage, or just churn out the copy in a second via an LLM.
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
That's the problem.
Yesterday, Nvidia claimed you could only run the models on their expensive hardware.
Today you can run them on a phone.
That's a problem for people who make really expensive hardware...
This is silly. The 1.5B distills of Deepseek-R1 that you can run on your phone are better than other people’s 1.5B sized LLMs, but they’re still mostly terrible & completely unable to cope with contexts of any reasonable size. Great for when you’re out & about and just want a question answered with a tolerable probability of the answer being wrong, not so great for anything else.
Deepseek is causing investors to panic about NVDA because they claim (quite reasonably, given their published papers) to have trained their full-size, 671B parameter model for $5million. That’s nothing compared to the amounts US AI companies have spent on training their models. This has immediately punched holes in the likely demand for NVDA hardware for training & the effectiveness of the US embargo on hardware exports to China which has been shown to be completely pointless, even counterproductive.
On the flip side, cheaper, higher quality inference (deepseek R1 is also much faster at inference than other models) is likely to drive demand for hardware for end users to use. Right now you can only run Deepseek on NVDA hardware. Tech types are buying NVDA hand over fist right now because they expect end user demand for NVDA hardware to sky-rocket as a result.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
Wales receives £120 for every £100 in England due to the Barnett formula
And its even more than that up here north of the wall. And yet our councils are absolutely broke because that cash isn't available to them.
We have two problems - it costs too much money to do anything, and we don't put up enough cash to cover the bill. Hence councils and public services having a dangerous lack of cash.
The solution is not to vote out the respective governments. That doesn't change the maths - we need to do things cheaper and pool budgets. Cutting the number of teachers cos we can't afford it costs us more money. And yet regardless of party they get stuck in this "cut the service and the need goes away" nonsense.
When Fairliered jnr. worked for the library service in your area a few years ago, there were not only small local libraries, but also mobile libraries. Now the mobile libraries are all gone, and they are closing the small local libraries.
I’d like to download DeepSeek but I’m worried about Chinese surveillance.
My attempts to get Claude to write spy thrillers haven’t been terribly successful, to be honest.
It's open source, you can download and spin it up yourself if you've got a modern GPU.
You can run it on your phone, assuming you have Android. I get remarkably snappy performance from the 1.5bn parameter model.
Nice, what's the word on Apple silicon? I just got my M4 Pro laptop delivered last week, finally replaced my old M1 Pro now that I don't have a laptop from work.
An M4 Pro with lots of memory is an ideal platform to run current LLMs at a usable clip. LLMs really, really like lots of memory bandwidth, and short of buying extremely expensive custom hardware M4 Pro macs have the best memory bandwidth you can buy in a consumer accessible package if your LLM of choice doesn’t fit in the 8-12Gb of a typical GPU. (24 GB GPUs are available, but at vast expense & those aren’t big enough to run the 70B weight LLMs.)
I got mine with 48GB RAM so I should be fine for memory allocation, half of why I've spent the better part of £3k on this is so that I can run some cool LLMs locally.
Nvidia are shipping £3k 'mac mini' like PC's that have pretty top-flight GPUs for LLM stuff.
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
That's the problem.
Yesterday, Nvidia claimed you could only run the models on their expensive hardware.
Today you can run them on a phone.
That's a problem for people who make really expensive hardware...
This is silly. The 1.5B distills of Deepseek-R1 that you can run on your phone are better than other people’s 1.5B sized LLMs, but they’re still mostly terrible & completely unable to cope with contexts of any reasonable size. Great for when you’re out & about and just want a question answered with a tolerable probability of the answer being wrong, not so great for anything else.
Deepseek is causing investors to panic about NVDA because they claim (quite reasonably, given their published papers) to have trained their full-size, 671B parameter model for $5million. That’s nothing compared to the amounts US AI companies have spent on training their models. This has immediately punched holes in the likely demand for NVDA hardware for training & the effectiveness of the US embargo on hardware exports to China which has been shown to be completely pointless, even counterproductive.
On the flip side, cheaper, higher quality inference (deepseek R1 is also much faster at inference than other models) is likely to drive demand for hardware for end users to use. Right now you can only run Deepseek on NVDA hardware. Tech types are buying NVDA hand over fist right now because they expect end user demand for NVDA hardware to sky-rocket as a result.
We’ll find out who’s right eventually.
I think it's most interesting for homebrew LLM development and training, it puts 100B+ param models into the hands of end users for people who can get a 5090 rather than big companies that buying up H100s. For Nvidia it completely kills their business model IMO, hence the market panic and drop in share price. They go back to being a GPU seller with 80% less demand for their big expensive AI products. They will still be the market leader for AI hardware but the market is just going to be smaller. It also puts the likes of Cerebras in a tough spot, in a world where running a 700B param model requires a cluster their wafer engine makes sense, in a world where it needs a rehash of the ghetto Ethereum mining servers that people used to make I don't know what kind of market there is for such a huge chip and resource.
I guess it could be that Cerebras will prosper because a model like Deepseek could potentially scale to WSE3 and suddenly we're moving to petascale models which would be the giant leap forwards for AI.
Just fyi, @deepseek_ai collects your IP, keystroke patterns, device info, etc etc, and stores it in China, where all that data is vulnerable to arbitrary requisition from the 🇨🇳 State.
Errrr
Anyone can download the Deepseek model, and run it themselves using Ollama on their local computer. And you can even do it with no internet connection.
I actually got Deepseek (1.5bn) running successfully on my phone with acceptable performance levels. (Obviously, I have insanely powerful phone. But still.)
You lost me with Ollama. I’m well known offline for my lack of technical chops.
I get you can run it locally, but I still don’t trust it, especially since at the end of the day none of my devices are permanently offline.
The models are just like massive Excel formulas. They can't 'do' anything apart from take your input, and return some output. If anything, I trust them more than Excel these days.
I know of a couple of writers who use local models via Ollama as they need to write 'adult' scenes (whether murder/mystery or kissy-fun-times) and the commercial ones just point-blank refuse.
Have they considered “writing” the scenes in question?
Yes. Much like they've considered going back to a mechanical typewriter or quill. Or programmers consider punch-cards. There are bits of writing they just don't enjoy ... writing. So getting a first draft from an LLM is fine for them.
Imagine you were the poor sod who had to write leon's posts. You could write them out long-hand like a savage, or just churn out the copy in a second via an LLM.
That’s producing copy, not writing. The distinction is important. Unless we train AI to distinguish between human generated and AI generated input you start to get feedback that causes irreversible defects in future models. Also human endeavour can’t be all enjoyable, or it’s not “endeavour”, but that’s a philosophical conundrum for another day.
CIA has concluded the Covid pandemic was “more likely” to have leaked from a Wuhan lab than emerged naturally. Chinese officials have long labelled this a “conspiracy theory” and, right from the start, were supported in that interpretation by much of the western MSM, especially the left-leaning MSM like the Guardian, NYT and BBC, some of whom smeared anybody who even gave the lab theory the time of day.
Hasn't it been obviously more likely - and generally viewed as more likely - for about four years now?
I mean, at the start there was a desperate attempt to avoid finger pointing, which was both understandable but also fundamentally unhelpful, and which damaged the reputation of science generally. But we've now swung to a point where yeah, we know that, and I don't think people in the MSM has been pushing the zoonotic line for a long time.
So it feels a little bit like punching a slow kid in the mouth again and again for something they did five years ago.
But if it's true, it is industrial malfeasance on a hitherto unimagined scale. The operators of the lab, the US and China should pay everyone affected by Covid punitive damages. Those affected by the Deepwater Horizon oil spill claimed millions in damages, but the affects of Covid were millions of times worse.
Being the most likely scenario doesn't make it a certainty.
It does if the other theory can be proven to be impossible.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
I've got a well-run local council who despite the real terms cuts in funding from central govt have managed to keep libraries and leisure centres open, services running and roads repaired, despite the increasingly heavy hgvs that destroy manholes, collapse drains and any areas washed out by mains water leaks on our road. I live in fear of increasing support for the lib dems locally, they run the neighbouring council and levelled their local leisure centre before realising they couldn't afford to replace it amongst other examples of incompetence.
LibDems are great at putting out leaflets and being "community".
Less good at running multi-million pound budget service providers.
Our main bin collection has been once a month for several years and our council tax has risen by 10.1% in 23, 9.67% in 24 and expected 10% this year
Welcome to North Wales
Possibly because the funding from central government is not keeping pace, hence the increase in the poll tax amount. It's not difficult.
They are inefficient and as they have no competition they have no reason to actually fix the issues, they can just milk the public by keeping charging higher and higher amounts. Just grifters.
I've got a well-run local council who despite the real terms cuts in funding from central govt have managed to keep libraries and leisure centres open, services running and roads repaired, despite the increasingly heavy hgvs that destroy manholes, collapse drains and any areas washed out by mains water leaks on our road. I live in fear of increasing support for the lib dems locally, they run the neighbouring council and levelled their local leisure centre before realising they couldn't afford to replace it amongst other examples of incompetence.
Which council and Leisure Centre are you referring to?
What's the earliest date for a Tory leadership challenge? Difficult to see one not happening if they're averaging 20% in the polls in third place.
Isn't the Truss/Johnson/May precedent that the rules don't really matter? If the party wants her out, they will make it clear and that's it, irrespective of whether Kemi B is holding the conch shell of immunity or whatever.
What's the earliest date for a Tory leadership challenge? Difficult to see one not happening if they're averaging 20% in the polls in third place.
Isn't the Truss/Johnson/May precedent that the rules don't really matter? If the party wants her out, they will make it clear and that's it, irrespective of whether Kemi B is holding the conch shell of immunity or whatever.
Presumably the earliest is a year from appointment is the first date for a challenge, and aren't a higher percentage of letters needed now?
I don't think Badenoch has the self doubt needed to resign, but then I didn't think Truss did either.
Includes this chart which kind of explains why UK local services have gone to fuck.
How much of that is mass academisation of schools? Before 2010, staff at most schools were employed by LEAs, now the money comes direct from the DfE. Multi academy trusts end up replicating what councils used to do, but less efficiently and with bigger salaries.
What's the earliest date for a Tory leadership challenge? Difficult to see one not happening if they're averaging 20% in the polls in third place.
Isn't the Truss/Johnson/May precedent that the rules don't really matter? If the party wants her out, they will make it clear and that's it, irrespective of whether Kemi B is holding the conch shell of immunity or whatever.
Presumably the earliest is a year from appointment is the first date for a challenge, and aren't a higher percentage of letters needed now?
I don't think Badenoch has the self doubt needed to resign, but then I didn't think Truss did either.
According to this one third of Tory MPs now need to send a letter, up from 15%. On a brief search I can't find anything official confirming this. I am sure some PBers know for certain.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
The man appears to have zero knowledge of the wealthy or of wealth. What we have at present is an economy based on the interests of wealth. Vast amounts of money extracted from electricity bill payers into the pockets of windmill-owning investment funds via green energy subsidies is an economy run in the interests of wealth. High barriers to entry so only huge corporations and established players can be in business and small businesses go under - that is an economy based on the interests of wealth. No grammar schools so there are no uppity poor people thinking they can be Government ministers - that is an economy run in the interests of wealth. Centrist 'post-capitalist' social denocratic greenery is a massive way to extract wealth from the middle and working classes. What Liz Truss was aiming for was a fast moving, fast growing economy with lots of opportunities to get wealthy, which by its nature operates as an efficient distributor of wealth, because vested interests are having to compete with new entrants in a free market that is functioning as it should.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
The jury was entitled to completely disbelieve the wife’s testimony. She could have been a terrible witness and she may well have wanted to keep her partner out of prison. As for motive, domestic violence, particularly involving alcohol and cannabis, has a multiplicity of motives, and few that could be described as “rational”.
The forensic evidence was that -
Home Office pathologist Dr Jennifer Bolton carried out a post-mortem examination and found that the kitchen knife breached the chest wall between the fifth and sixth ribs, went through her lower lung and passed into the left ventricle of the heart.
Scarlett died very quickly from blood loss, the pathologist found.
Dr Bolton said it was her opinion that the knife was being “held tightly” at the time so that when it came into contact with Scarlett, it went into her.
“That typically means a firm grip and that arm is braced in a certain way,” she said.
Asked by prosecutor Mark McKone KC if she thought the knife could have been thrown towards Scarlett, Dr Bolton said: “Kitchen knives aren’t designed to be thrown, they aren’t designed to go through the air.
“So, it is practically impossible for a kitchen knife to be thrown for it to travel in such a way that it lands on Scarlett’s clothing and then her skin at 90 degrees, so it doesn’t simply bounce off or scratch across, and then go 11cm in and apparently come out again.”
Forensic scientist Gemma Escott also studied the large knife and judged that material on the blade indicated a stabbing motion had been used.
The amount of force required to throw a knife so hard it penetrates the chest wall and and then travels four and a half inches into the body, exiting the other side, suggests javelin thrower force. To suggest he could additionally, accidentally, do it so that the tip struck at precisely 90 degrees is, from a layman’s perspective, “practically impossible”. I can completely understand why the jury found no reasonable doubt.
What's the earliest date for a Tory leadership challenge? Difficult to see one not happening if they're averaging 20% in the polls in third place.
Isn't the Truss/Johnson/May precedent that the rules don't really matter? If the party wants her out, they will make it clear and that's it, irrespective of whether Kemi B is holding the conch shell of immunity or whatever.
Presumably the earliest is a year from appointment is the first date for a challenge, and aren't a higher percentage of letters needed now?
I don't think Badenoch has the self doubt needed to resign, but then I didn't think Truss did either.
I think that the ban was on AI imagery and the output from it, and was initiated by Leon posting AI-generated imagery. However I may be wrong. There is presumably a difference between AI firms and AI concepts and AI outputs, and I don't know on which side of the ban they lie. No doubt the mods will clarify.
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
Richard Murphy is the left wing Patrick Minford (he of “Economists for Brexit”). Both are total nutters, essentially the David Ickes of the economics profession.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
The jury was entitled to completely disbelieve the wife’s testimony. She could have been a terrible witness and she may well have wanted to keep her partner out of prison. As for motive, domestic violence, particularly involving alcohol and cannabis, has a multiplicity of motives, and few that could be described as “rational”.
The forensic evidence was that -
Home Office pathologist Dr Jennifer Bolton carried out a post-mortem examination and found that the kitchen knife breached the chest wall between the fifth and sixth ribs, went through her lower lung and passed into the left ventricle of the heart.
Scarlett died very quickly from blood loss, the pathologist found.
Dr Bolton said it was her opinion that the knife was being “held tightly” at the time so that when it came into contact with Scarlett, it went into her.
“That typically means a firm grip and that arm is braced in a certain way,” she said.
Asked by prosecutor Mark McKone KC if she thought the knife could have been thrown towards Scarlett, Dr Bolton said: “Kitchen knives aren’t designed to be thrown, they aren’t designed to go through the air.
“So, it is practically impossible for a kitchen knife to be thrown for it to travel in such a way that it lands on Scarlett’s clothing and then her skin at 90 degrees, so it doesn’t simply bounce off or scratch across, and then go 11cm in and apparently come out again.”
Forensic scientist Gemma Escott also studied the large knife and judged that material on the blade indicated a stabbing motion had been used.
The amount of force required to throw a knife so hard it penetrates the chest wall and and then travels four and a half inches into the body, exiting the other side, suggests javelin thrower force. To suggest he could additionally, accidentally, do it so that the tip struck at precisely 90 degrees is, from a layman’s perspective, “practically impossible”. I can completely understand why the jury found no reasonable doubt.
Thanks. The Independent write up is a lot clearer than the BBC one (and more compelling).
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
Having read that it seems likely they are both lying. "Some tongs flicked up the knife so it penetrated four inches into her chest" seems on the face of it extremely implausible.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
I think what makes it quite technical here is that a few of us are in and around LLMs for work reasons so have a reasonably good (me) or very good (Phil) understanding of the how so that's why the conversation tends towards it rather than the "so what" which is probably a more important discussion.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
Having read that it seems likely they are both lying. "Some tongs flicked up the knife so it penetrated four inches into her chest" seems on the face of it extremely implausible.
Sad, but seems more like manslaughter than murder.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
Having read that it seems likely they are both lying. "Some tongs flicked up the knife so it penetrated four inches into her chest" seems on the face of it extremely implausible.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
I think what makes it quite technical here is that a few of us are in and around LLMs for work reasons so have a reasonably good (me) or very good (Phil) understanding of the how so that's why the conversation tends towards it rather than the "so what" which is probably a more important discussion.
I personally prefer to let people discuss whatever they want, even “the thing”.
I note the ban on that remains.
Leon is a suis generis problem. Unless he is given strict parameters he is quite capable of destroying a thread and deterring a legion of posters.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
My sister works for a major London university and has said in the 12 years she's been there the university has almost doubled the size of "support services" but she doesn't know what half of them do now, even in her department where she just continually gets extra budget for staff.
Evening all. Have we discussed the case of the father found guilty of murdering his daughter - https://www.bbc.co.uk/news/articles/cly9zx02rejo - even though the only witness, his wife, insists it was accidental?
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
Having read that it seems likely they are both lying. "Some tongs flicked up the knife so it penetrated four inches into her chest" seems on the face of it extremely implausible.
Sad, but seems more like manslaughter than murder.
Well no, he clearly held the knife and stabbed her. The flicking/throwing story is just that, a story. Hard to make that manslaughter, it is an action that is likely to kill. We haven't sat through the whole trial.
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
That's the problem
Yesterday, Nvidia claimed you could only run the models on their expensive hardware.
Today you can run them on a phone.
That's a problem for people who make really expensive hardware...
Nvidia has been trading on a P/E ratio of >50 until Friday.
Can PB tech experts tell me if it's worth fishing my 1050ti out and using it to mine...something?
No, Nvidia only started adding tensor cores to consumer GPUs with the 20 series.
This isn't even a recognisable form of English.
"Nvidia" is a manufacturer of hardware and software associated with AI "Tensor cores" are little things on a chip that make GPUs go faster "GPUs" are specialised chips originally intended for computer graphics but now repurposed for complex maths "20 series" is the term used to refer to a series of Nvidia GPUs manufactured from around 2018/9, now obsolete "GTX 1050Ti" is a Nvidia GPU prior to the 20 series
So @Eabhal was wondering if he could use his circa-2014 chip to do AI stuff. @MaxPB told him no because the stuff to make the chip go fast enough wasn't put on there until five or so years later, and even those have now been superseded. It's the geek way of telling @Eabhal he's hung like a hamster in terms of chip speed and cannot keep up with the poshos.
"NEW -- #Trump's global aid freeze has cut the salaries paid to many of the prison & camp guards responsible for securing 9,500 #ISIS militants & ~40,000 associated women/kids in northeast #Syria .
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
My sister works for a major London university and has said in the 12 years she's been there the university has almost doubled the size of "support services" but she doesn't know what half of them do now, even in her department where she just continually gets extra budget for staff.
I have an operations capability, ostensibly efficiency-saving, but who - left unchecked - create superfluous process documentation and militate for additional hires.
I keep them going because they do serve certain useful functions, but it’s v clear to me that keeping clear of the knotweed is a vital task of leadership, and bureaucracies naturally expand. Indeed, I believe there is a law that describes this native tendency.
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
If Richard Murphy thinks something you can pretty much guarantee the opposite is true. The Alex Jones of accountancy. He's a well known crank who for years (he first emerged under the coalition) has basically been saying "there are hundreds of billions of pounds we could tax now but I am the only person to have realised this".
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
The man appears to have zero knowledge of the wealthy or of wealth. What we have at present is an economy based on the interests of wealth. Vast amounts of money extracted from electricity bill payers into the pockets of windmill-owning investment funds via green energy subsidies is an economy run in the interests of wealth. High barriers to entry so only huge corporations and established players can be in business and small businesses go under - that is an economy based on the interests of wealth. No grammar schools so there are no uppity poor people thinking they can be Government ministers - that is an economy run in the interests of wealth. Centrist 'post-capitalist' social denocratic greenery is a massive way to extract wealth from the middle and working classes. What Liz Truss was aiming for was a fast moving, fast growing economy with lots of opportunities to get wealthy, which by its nature operates as an efficient distributor of wealth, because vested interests are having to compete with new entrants in a free market that is functioning as it should.
Oh I agree with you, don't worry, at least domestically. I just wanted to give Starmer a kicking and some of the points Richard Murphy made (esp the firing of the Competitions guy) combined with Reeves's adoption of Conservative policies like cutting down on disabled money[1] helped with that.
[1] I have disabled or partially disabled siblings and I remember Cameron unkindly
Can PB tech experts tell me if it's worth fishing my 1050ti out and using it to mine...something?
No, Nvidia only started adding tensor cores to consumer GPUs with the 20 series.
This isn't even a recognisable form of English.
"Nvidia" is a manufacturer of hardware and software associated with AI "Tensor cores" are little things on a chip that make GPUs go faster "GPUs" are specialised chips originally intended for computer graphics but now repurposed for complex maths "20 series" is the term used to refer to a series of Nvidia GPUs manufactured from around 2018/9, now obsolete "GTX 1050Ti" is a Nvidia GPU prior to the 20 series
So @Eabhal was wondering if he could use his circa-2014 chip to do AI stuff. @MaxPB told him no because the stuff to make the chip go fast enough wasn't put on there until five or so years later, and even those have now been superseded. It's the geek way of telling @Eabhal he's hung like a hamster in terms of chip speed and cannot keep up with the poshos.
I would hope that all PBers, of whatever technical proficiency, would have understood that my 1050ti comment (which was a perfectly respectable mid-level card back then) was tongue-in-cheek.
I have a box of random old computer components, including a 2060. That might do the trick, or else I'll bury it in a mound like Sutton Hoo and a future civilisation will think I was an AI King.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
My gut says yes. We have huge teams of staff who never see a student nor do they do any research. Partly this is in response to demands for accountability (externally, such as funding bodies). But there is no doubt that academics have a long logistical tail behind us nowadays.
I have a box of random old computer components, including a 2060. That might do the trick, or else I'll bury it in a mound like Sutton Hoo and a future civilisation will think I was an AI King.
There is apparently a thriving trade in "mining" landfill for old computer parts to reclaim the gold used on the contacts
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
My sister works for a major London university and has said in the 12 years she's been there the university has almost doubled the size of "support services" but she doesn't know what half of them do now, even in her department where she just continually gets extra budget for staff.
There's an awful lot of paper shuffling and an awful lot of pointless meetings, particularly on larger projects. Processes seem to become more arcane.
On costs, an interesting one recently was flights - unusually I looked myself at the options as we're between administrators in the group and the next person we're supposed to go to is crazy overworked, so I knew it would be quicker to go with the exact flights to be booked. But bookings have to be done via the University's travel provider - third party, not University owned, which did the exact same flights at more than 50% extra cost compared to what I could have booked at my desk with a credit card. Still, it was out of a research grant budget, which covered it, getting someone to buy direct on the retail site would require specia authorisation and me buying it and claiming it back would take about two months before I saw the money. So we sucked it up and the University's travel provider got rich, I guess.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
It would be interesting to see if UK academe has suffered the same runaway administration costs that U.S. higher education has.
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
My sister works for a major London university and has said in the 12 years she's been there the university has almost doubled the size of "support services" but she doesn't know what half of them do now, even in her department where she just continually gets extra budget for staff.
At Bath lectures start on the quarter hour (e.g. 9.15). You can always tell when a member of staff has no teaching or interaction with students because they arrange meetings on the hour. an hour meeting thus blocks two 1 hour teaching slots. All too common.
Lots of staff have been brought in to satisfy outside bodies (EDI, climate change, widening participation). Very often they waste academics time getting us to give them information.
Unis DO need admin and support staff, but like HR, some of these processes become an end in themselves.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
One wonders whether the NHS and HEIs have suffered the same problem in the last forty years: the true professionals (hospital consultants and senior academics) who ran their little worlds have had to be by-passed and "controlled" by business unit types with MBAs who want to run things "efficiently" because the spreadsheet says X and Y?
I left the uni world (as a staff person not student) in early 2000s. Already the air was wracked with talk of students as "customers" who had to be "serviced".
No longer a community of learning but a business administration centre.
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
The man appears to have zero knowledge of the wealthy or of wealth. What we have at present is an economy based on the interests of wealth. Vast amounts of money extracted from electricity bill payers into the pockets of windmill-owning investment funds via green energy subsidies is an economy run in the interests of wealth. High barriers to entry so only huge corporations and established players can be in business and small businesses go under - that is an economy based on the interests of wealth. No grammar schools so there are no uppity poor people thinking they can be Government ministers - that is an economy run in the interests of wealth. Centrist 'post-capitalist' social denocratic greenery is a massive way to extract wealth from the middle and working classes. What Liz Truss was aiming for was a fast moving, fast growing economy with lots of opportunities to get wealthy, which by its nature operates as an efficient distributor of wealth, because vested interests are having to compete with new entrants in a free market that is functioning as it should.
Murphy is wrong about both the problems and the solutions. Always.
Yes, he is even more wrong about everything than Dom Cummings.
Can PB tech experts tell me if it's worth fishing my 1050ti out and using it to mine...something?
No, Nvidia only started adding tensor cores to consumer GPUs with the 20 series.
This isn't even a recognisable form of English.
"Nvidia" is a manufacturer of hardware and software associated with AI "Tensor cores" are little things on a chip that make GPUs go faster "GPUs" are specialised chips originally intended for computer graphics but now repurposed for complex maths "20 series" is the term used to refer to a series of Nvidia GPUs manufactured from around 2018/9, now obsolete "GTX 1050Ti" is a Nvidia GPU prior to the 20 series
So @Eabhal was wondering if he could use his circa-2014 chip to do AI stuff. @MaxPB told him no because the stuff to make the chip go fast enough wasn't put on there until five or so years later, and even those have now been superseded. It's the geek way of telling @Eabhal he's hung like a hamster in terms of chip speed and cannot keep up with the poshos.
I would hope that all PBers, of whatever technical proficiency, would have understood that my 1050ti comment (which was a perfectly respectable mid-level card back then) was tongue-in-cheek.
I have a box of random old computer components, including a 2060. That might do the trick, or else I'll bury it in a mound like Sutton Hoo and a future civilisation will think I was an AI King.
No need to protest, we all, sort of, understand your predicament, n-eabhal-s 🐹, if I may doxx your true identity for a moment.
(...I feel a wink emoji ought to be extraneous here).
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
One wonders whether the NHS and HEIs have suffered the same problem in the last forty years: the true professionals (hospital consultants and senior academics) who ran their little worlds have had to be by-passed and "controlled" by business unit types with MBAs who want to run things "efficiently" because the spreadsheet says X and Y?
I left the uni world (as a staff person not student) in early 2000s. Already the air was wracked with talk of students as "customers" who had to be "serviced".
No longer a community of learning but a business administration centre.
Starbucks University, as an old friend in Academia calls it.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
Only on my computer. Though when that becomes sentient I might be in trouble.
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
CIA has concluded the Covid pandemic was “more likely” to have leaked from a Wuhan lab than emerged naturally. Chinese officials have long labelled this a “conspiracy theory” and, right from the start, were supported in that interpretation by much of the western MSM, especially the left-leaning MSM like the Guardian, NYT and BBC, some of whom smeared anybody who even gave the lab theory the time of day.
Hasn't it been obviously more likely - and generally viewed as more likely - for about four years now?
I mean, at the start there was a desperate attempt to avoid finger pointing, which was both understandable but also fundamentally unhelpful, and which damaged the reputation of science generally. But we've now swung to a point where yeah, we know that, and I don't think people in the MSM has been pushing the zoonotic line for a long time.
So it feels a little bit like punching a slow kid in the mouth again and again for something they did five years ago.
But if it's true, it is industrial malfeasance on a hitherto unimagined scale. The operators of the lab, the US and China should pay everyone affected by Covid punitive damages. Those affected by the Deepwater Horizon oil spill claimed millions in damages, but the affects of Covid were millions of times worse.
Being the most likely scenario doesn't make it a certainty.
It does if the other theory can be proven to be impossible.
Come on: while lab leak is most likely, we have no idea how it took place, and therefore no reasonable way of assessing relative culpability.
Let's assume it happened.
At the one end of the scale it could be because someone on a transport flight bringing a bat to Wuhan got bit, and then his wife worked at the wet market.
At the other end of scale, there was an accident that allowed a virus to escape that was created as part of gain of function research.
And there are a dozen possibilities along that continuum. If it was something akin to the former, then why should the US have any responsibility at all?
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
Only on my computer. Though when that becomes sentient I might be in trouble.
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
My dabbling with various AI tools has left me deeply unimpressed tbh.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
Only on my computer. Though when that becomes sentient I might be in trouble.
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
My dabbling with various AI tools has left me deeply unimpressed tbh.
I oscillate between very impressed and rather disappointed. I think that’s because every time I’m impressed it raises my expectations of the technology for next time.
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
I can help - fire all the bullshit admin staff that waste our time (e.g. EDI)
One wonders whether the NHS and HEIs have suffered the same problem in the last forty years: the true professionals (hospital consultants and senior academics) who ran their little worlds have had to be by-passed and "controlled" by business unit types with MBAs who want to run things "efficiently" because the spreadsheet says X and Y?
I left the uni world (as a staff person not student) in early 2000s. Already the air was wracked with talk of students as "customers" who had to be "serviced".
No longer a community of learning but a business administration centre.
Trust me, it's got way, way worse. Especially since lockdown/covid times for reasons I'm sure academic papers will be written on. But there was a very steep rise in bringing in 'professional management' post-2020 with - as you can imagine - an equivalent very steep decline in productivity and morale.
If I was a cynical person, I might think a lot of proven-failure managers from the private sector had been sacked during covid-times and shifted over the to public sector. But... thankfully I'm not that cynical. I welcome my new very well compensated overlords.
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
Only on my computer. Though when that becomes sentient I might be in trouble.
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
My dabbling with various AI tools has left me deeply unimpressed tbh.
I oscillate between very impressed and rather disappointed. I think that’s because every time I’m impressed it raises my expectations of the technology for next time.
I find the mathematicians, coders, gamers and financial experts tend to be very impressed by the latest AI.
Literary, philosophical, political and historical people, less so. It doesn't perform their tasks and there's a tidal wave of hugely-funded bullshit in that area.
I use Copilot everyday at work ( science technician) it's brilliant. AI has its uses other than making dumb memes.
I find the problem is it's sometimes brilliant, yet wrong, yet convincing. As stands - even the SOTA 'reasoning' models need a lot of oversight (whether by humans or other LLMs). Which is annoying.
I use Copilot everyday at work ( science technician) it's brilliant. AI has its uses other than making dumb memes.
What relationship did Bill Gates have with Jeffrey Epstein?
"I'd really like to help, but it seems this topic is off-limits for me. Sorry about that!"
Funny - it started to go into quite some detail and then wiped itself.
I have a lot of problems with the LLM's - but I can give them a break for 'don't say something that will cause lawyer-grief'. Certainly in the regular chat UI's.
To continue the point rcs1000 just made: It is entirely possible that no one knows what happened with the COVID origins -- and that we never will know, since the ChiComs won't let us investigate.
"Germany’s Habeck warns of an Austria-style right-wing wave
The economy minister and Green Party chancellor candidate says Germany could follow Austria’s far-right path, as the CDU signals willingness to erode its firewall against the AfD ahead of a pivotal election."
The problem with banning AI talk is that it’s like banning atomic bomb chat in 1945, or dreadnought discussions in 1910.
However perhaps we need a way to discourage the more technical stuff.
Indeed, today it is more geopolitics than anything.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
It's has, however, probably managed to add your name to a 'list' for future reference.
Only on my computer. Though when that becomes sentient I might be in trouble.
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
My dabbling with various AI tools has left me deeply unimpressed tbh.
I oscillate between very impressed and rather disappointed. I think that’s because every time I’m impressed it raises my expectations of the technology for next time.
I find the mathematicians, coders, gamers and financial experts tend to be very impressed by the latest AI.
Literary, philosophical, political and historical people, less so. It doesn't perform their tasks and there's a tidal wave of hugely-funded bullshit in that area.
I went through our 'hardest' exam questions with a lecturer recently. Hardest in a 3rd-year Engineering sense in this case. The 'o1' models took - at most - 2.5 minutes to solve each question with full working and correct results.
They are really becoming very capable at maths and coding. But still make - to my mind - baffling mistakes. I gave the o1 model a fairly simple python script the other week and asked it to add a pre-flight check of the options it had been passed before the main/expensive part of the code kicked off.
It did a great job - verification code all perfect. But it added the 'pre-flight' check after the expensive part of the code. I tried a few times with different prompting - each time the same result.
Can PB tech experts tell me if it's worth fishing my 1050ti out and using it to mine...something?
No, Nvidia only started adding tensor cores to consumer GPUs with the 20 series.
And the 20 series, despite its age, is still useful for AI. I've been playing about with Deepseek and the 14b model on my elderly RTX 2080 and it is entirely useable, if a bit slow.
I use Copilot everyday at work ( science technician) it's brilliant. AI has its uses other than making dumb memes.
What relationship did Bill Gates have with Jeffrey Epstein?
"I'd really like to help, but it seems this topic is off-limits for me. Sorry about that!"
Funny - it started to go into quite some detail and then wiped itself.
I have a lot of problems with the LLM's - but I can give them a break for 'don't say something that will cause lawyer-grief'. Certainly in the regular chat UI's.
It didn't have any trouble with other people and Epstein. So it fails a test.
For the second time in the past few days, Trump is “joking” about violating the Constitution and running for a third term:
“I’ve raised a lot of money for the next race that I assume I can’t use for myself… I think I’m not allowed to run again. I’m not sure, am I allowed to run again?”
Comments
It's why the public realm looks so rubbish. And the whining about cycle lanes is misplaced, because even that comes from direct ring-fenced central government funding.
I don't actually get why this is so fatal for Nvidia. Surely a massive advance in AI is going to be brilliant for them, long term. One thing we know from tech is an insatiable demand for bigger, better, faster.
BREAKING: Per multiple law enforcement sources, within the last hour, Border Patrol agents near Fronton, TX were fired upon from MX by suspected cartel gunmen as a group of illegal aliens were being brought across the river. I’m told BP returned fire, nobody hit on either side, and that the illegal aliens did not make it across.
I live in fear of increasing support for the lib dems locally, they run the neighbouring council and levelled their local leisure centre before realising they couldn't afford to replace it amongst other examples of incompetence.
But, tbf, they have had their funding cut too. I can't help thinking that the only hope for change is exec mayors who are much more accountable to the public.
I know of a couple of writers who use local models via Ollama as they need to write 'adult' scenes (whether murder/mystery or kissy-fun-times) and the commercial ones just point-blank refuse.
"Is Keir Starmer a worse Prime Minister than Liz Truss?", Richard J Murphy, 7mins
https://www.youtube.com/watch?v=rJBCfPWkIbA&list=UUGlMOIZ1A3zluwLXTGpzZfw&index=3
"...Reeves is actually working against the best interests of the people of this country and in favour of the hierarchy of wealth. That isn't what Labour should be doing!...and if Labour does deliver austerity in March as (I expect) then I have little doubt at all that she will become worse than Liz Truss, because she actually is going to deliver what Liz Truss fail to do: an economy totally biased to the interests of wealth.
How did we get to a point where Labour is so corrupted, has so lost touch with its core values, is so out of touch with the people of this country who put it into office?...positively working against the best interests of most people in this country? I wish I knew the answer to that question. I can only put it down to corruption: political corruption, intellectual corruption, and plain straightforward alignment of the personal interests of those who are in office with those who have wealth..."
Less good at running multi-million pound budget service providers.
Lets see what it knows about Winnie the Pooh.
But there can be a very large difference in outcomes between well and badly run councils. Albeit complicated by particular local circumstances.
Yesterday, Nvidia claimed you could only run the models on their expensive hardware.
Today you can run them on a phone.
That's a problem for people who make really expensive hardware...
Imagine you were the poor sod who had to write leon's posts. You could write them out long-hand like a savage, or just churn out the copy in a second via an LLM.
Deepseek is causing investors to panic about NVDA because they claim (quite reasonably, given their published papers) to have trained their full-size, 671B parameter model for $5million. That’s nothing compared to the amounts US AI companies have spent on training their models. This has immediately punched holes in the likely demand for NVDA hardware for training & the effectiveness of the US embargo on hardware exports to China which has been shown to be completely pointless, even counterproductive.
On the flip side, cheaper, higher quality inference (deepseek R1 is also much faster at inference than other models) is likely to drive demand for hardware for end users to use. Right now you can only run Deepseek on NVDA hardware. Tech types are buying NVDA hand over fist right now because they expect end user demand for NVDA hardware to sky-rocket as a result.
We’ll find out who’s right eventually.
https://www.nvidia.com/en-eu/project-digits/
I guess it could be that Cerebras will prosper because a model like Deepseek could potentially scale to WSE3 and suddenly we're moving to petascale models which would be the giant leap forwards for AI.
All you need it a muslin (or any other finely straining) bag, a funnel, a bowl, full fat yoghurt, and salt
Add whatever you want to mix into the yoghurt, then the salt, transfer to bag, into funnel, into bowl
Leave it in the fridge overnight, and in the morning you have have the freshest cream cheese
Can PB tech experts tell me if it's worth fishing my 1050ti out and using it to mine...something?
I don't think Badenoch has the self doubt needed to resign, but then I didn't think Truss did either.
"A lie spreads around the world before the truth gets its boots on".
It's even more true in the world of Social Media.
https://www.standard.co.uk/news/politics/tories-mps-iain-duncan-smith-kemi-badenoch-gb-news-b1191649.html
@hendopolis
·
35m
STAR: War of the psycho scumbag chatbots #TomorrowsPapersToday
Seems, from the reporting, to rest in part on expert testimony that the knife could not have accidentally caused the fatal wound - https://www.bbc.co.uk/news/articles/c04n7yyqz9po - i.e. that it's just not plausible for an accidentally thrown knife to hit so that it pierces ribs cage and reaches vital organs in that way. However, the alternative, that an apparently untrained man managed a single stab wound with that precision, without apparent motive and with his wife in the room and that his wife would then be complicit in passing it off as an accident - well, that seems vanishngly unlikely too.
Maybe there's more to it (reporting often isn't great) but it seems to rest on that testinmony and a bit of confusion in stories of what actually happened. Even if the man was holding the knife, you surely have to completely disbelieve the wife's testimony to come to premeditated murder rather than manslaughter. Just seems a very odd case, but maybe there's compelling evidence that wasn't considered interesting enough to make the story?
@gsoh31.bsky.social
I'm very much afraid that the first of the really big explosions is about to drop in Higher Education: at Cardiff University, where management speak of an 'immediate existential crisis'. And this is just the start. Solidarity with and to all colleagues at Cardiff.
https://bsky.app/profile/gsoh31.bsky.social/post/3lgqwcduitc2i
The forensic evidence was that -
Home Office pathologist Dr Jennifer Bolton carried out a post-mortem examination and found that the kitchen knife breached the chest wall between the fifth and sixth ribs, went through her lower lung and passed into the left ventricle of the heart.
Scarlett died very quickly from blood loss, the pathologist found.
Dr Bolton said it was her opinion that the knife was being “held tightly” at the time so that when it came into contact with Scarlett, it went into her.
“That typically means a firm grip and that arm is braced in a certain way,” she said.
Asked by prosecutor Mark McKone KC if she thought the knife could have been thrown towards Scarlett, Dr Bolton said: “Kitchen knives aren’t designed to be thrown, they aren’t designed to go through the air.
“So, it is practically impossible for a kitchen knife to be thrown for it to travel in such a way that it lands on Scarlett’s clothing and then her skin at 90 degrees, so it doesn’t simply bounce off or scratch across, and then go 11cm in and apparently come out again.”
Forensic scientist Gemma Escott also studied the large knife and judged that material on the blade indicated a stabbing motion had been used.
https://www.independent.co.uk/news/uk/crime/simon-vickers-guilty-murder-darlington-forensics-play-fight-b2686977.html
The amount of force required to throw a knife so hard it penetrates the chest wall and and then travels four and a half inches into the body, exiting the other side, suggests javelin thrower force. To suggest he could additionally, accidentally, do it so that the tip struck at precisely 90 degrees is, from a layman’s perspective, “practically impossible”. I can completely understand why the jury found no reasonable doubt.
However perhaps we need a way to discourage the more technical stuff.
I've just installed Deepseek locally, and it has gone into a tizz because I asked it about Winnie the Pooh and whether the character was ever compared to any national leaders.
So there are some things it just won't do...
Essentially, the cost of actual professors has stayed pretty much the same or even lower, but the “professionalisation” of universities has resulted in bloated and unaccountable administration overheads.
I note the ban on that remains.
Leon is a suis generis problem. Unless he is given strict parameters he is quite capable of destroying a thread and deterring a legion of posters.
"Tensor cores" are little things on a chip that make GPUs go faster
"GPUs" are specialised chips originally intended for computer graphics but now repurposed for complex maths
"20 series" is the term used to refer to a series of Nvidia GPUs manufactured from around 2018/9, now obsolete
"GTX 1050Ti" is a Nvidia GPU prior to the 20 series
So @Eabhal was wondering if he could use his circa-2014 chip to do AI stuff. @MaxPB told him no because the stuff to make the chip go fast enough wasn't put on there until five or so years later, and even those have now been superseded. It's the geek way of telling @Eabhal he's hung like a hamster in terms of chip speed and cannot keep up with the poshos.
"NEW --
#Trump's
global aid freeze has cut the salaries paid to many of the prison & camp guards responsible for securing 9,500
#ISIS
militants & ~40,000 associated women/kids in northeast
#Syria
.
Many are no longer turning up for work."
Well done Donald you very stable genius.
I'm out.
I keep them going because they do serve certain useful functions, but it’s v clear to me that keeping clear of the knotweed is a vital task of leadership, and bureaucracies naturally expand. Indeed, I believe there is a law that describes this native tendency.
[1] I have disabled or partially disabled siblings and I remember Cameron unkindly
I have a box of random old computer components, including a 2060. That might do the trick, or else I'll bury it in a mound like Sutton Hoo and a future civilisation will think I was an AI King.
Literally buried treasure...
i cant believe ChatGPT lost its job to AI
On costs, an interesting one recently was flights - unusually I looked myself at the options as we're between administrators in the group and the next person we're supposed to go to is crazy overworked, so I knew it would be quicker to go with the exact flights to be booked. But bookings have to be done via the University's travel provider - third party, not University owned, which did the exact same flights at more than 50% extra cost compared to what I could have booked at my desk with a credit card. Still, it was out of a research grant budget, which covered it, getting someone to buy direct on the retail site would require specia authorisation and me buying it and claiming it back would take about two months before I saw the money. So we sucked it up and the University's travel provider got rich, I guess.
Lots of staff have been brought in to satisfy outside bodies (EDI, climate change, widening participation). Very often they waste academics time getting us to give them information.
Unis DO need admin and support staff, but like HR, some of these processes become an end in themselves.
I left the uni world (as a staff person not student) in early 2000s. Already the air was wracked with talk of students as "customers" who had to be "serviced".
No longer a community of learning but a business administration centre.
Yes, he is even more wrong about everything than Dom Cummings.
The perfection of his ignorance is magnificent.
(...I feel a wink emoji ought to be extraneous here).
This might be a hit to NVIDIA and some AI companies, but is it a hit to users? Perhaps making it cheaper or better for everyone won't be a net negative.
Let's assume it happened.
At the one end of the scale it could be because someone on a transport flight bringing a bat to Wuhan got bit, and then his wife worked at the wet market.
At the other end of scale, there was an accident that allowed a virus to escape that was created as part of gain of function research.
And there are a dozen possibilities along that continuum. If it was something akin to the former, then why should the US have any responsibility at all?
The first [known] Jan6er post-pardon to get in trouble with police was shot for resisting arrest.
https://www.fox32chicago.com/news/indiana-man-fatally-shot-deputy-identified-jan-6
If I was a cynical person, I might think a lot of proven-failure managers from the private sector had been sacked during covid-times and shifted over the to public sector. But... thankfully I'm not that cynical. I welcome my new very well compensated overlords.
Literary, philosophical, political and historical people, less so. It doesn't perform their tasks and there's a tidal wave of hugely-funded bullshit in that area.
"I'd really like to help, but it seems this topic is off-limits for me. Sorry about that!"
Funny - it started to go into quite some detail and then wiped itself.
(And, as usual, one should not reject a possible Hanlon Razor explanation: https://en.wikipedia.org/wiki/Hanlon's_razor )
What we can say that the official reactions to the disease in the UK and, even more, in the US were not optimal. Example: The Loser was directly responsible for spreading the disease at the nomination of Amy Coney Barrett: https://www.cnn.com/2020/10/03/politics/trump-covid-amy-coney-barrett-event/index.html
The economy minister and Green Party chancellor candidate says Germany could follow Austria’s far-right path, as the CDU signals willingness to erode its firewall against the AfD ahead of a pivotal election."
https://www.politico.eu/article/germany-robert-habeck-austria-style-right-wing-wave-germany-warning/
They are really becoming very capable at maths and coding. But still make - to my mind - baffling mistakes. I gave the o1 model a fairly simple python script the other week and asked it to add a pre-flight check of the options it had been passed before the main/expensive part of the code kicked off.
It did a great job - verification code all perfect. But it added the 'pre-flight' check after the expensive part of the code. I tried a few times with different prompting - each time the same result.
https://www.bloomberg.com/energy
Republicans against Trump
@RpsAgainstTrump
For the second time in the past few days, Trump is “joking” about violating the Constitution and running for a third term:
“I’ve raised a lot of money for the next race that I assume I can’t use for myself… I think I’m not allowed to run again. I’m not sure, am I allowed to run again?”
https://www.electoralcalculus.co.uk/homepage.html
Kemi won both the Tory MPs and Tory members vote and is secure until the next general election