Hacker News new | past | comments | ask | show | jobs | submit login
'Moonshot' hype illustrates the UK government's obsession with tech hyperbole (theguardian.com)
28 points by gemmamilne 11 days ago | hide | past | favorite | 36 comments

This is a good piece.

They also take a similar approach to regulation - the age verification strategy was going to be "world leading" and fell apart before it launched. The new Online Harms bill will also be "world leading" but is now supposed to solve so many complex cross societal problems it will also inevitably implode.

It's not really tech specific I think, just that the current government in particular is obsessed with thinking they are super naturally gifted because they own a lot of flags, and like to repeat slogans over any nuanced thinking.

It isn't the government's obsession, it is from Dominic Cummings, an unelected beaurocrat, the likes of which many people voted in favour of brexit to get rid of (so they thought).

Current Secretary of State for the Department of Health and Social Care has pushed "better tech" as a solution for all of the NHS's problems since before he got the job.

I reckon the one thing "Project Moonshot" actually demonstrates is how much media coverage of Covid-19 is partisan horseshit. It's basically identical to what the New York Times and other American media publications have been claiming the US desperately needed for months, and they portrayed it as a neat, easy solution that the US would have had months ago if not for Trump's incompentence. Yet when the UK government and Boris Johnson - who is seen as our equivalent of Trump - attempts something similar, suddenly there's a wave of news articles in the UK context about how it's impossible, the tech doesn't even exist, it would have far too many false positives, and so on. Pretty much the exact same plan is being spun in the UK as a delusional distraction by right-wing populists that pushes a quick fix where none exists, and in the US as a easy fix that would've happened if not for the incompetent right-wing populists refusing to do it as part of an evil cover-up. (Needless to say, pretty much all the problems with false positives, inadequate technology, the diffculty of scaling up, etc would have been much bigger problems in the US a few months ago given the much larger population and the improvements in Covid-19 testing technology since then.)

My view on this is that there’d be a lot less grief being handed out over Project Moonshot if the existing solutions were being given the same level of focus on making them work. The current plan from the UK government is equivalent to a software developer responding to a bug report by rewriting the whole application - it may well work, but it’ll be massively more expensive, and come with a whole host of new unknown problems.

Right now in the UK the basic testing infrastructure is a mess, people are being told to make 500 mile round trips to get tested, and then thousands of tests are being disposed of undone because of logistical issues.

Moonshot has a budget roughly equivalent to running the entirety of the NHS for a year, I can think of much better uses for that money than a project with no clear path to success, such as massively scaling up test and trace infrastructure. But of course we’re Great Britain, and so everything must be the very best. Such as the failed attempt to implement tracing apps with a custom solution rather than building on top of the Android and iOS SDKs for the purpose.

Right now in the UK, the basic testing infrastructure is more-or-less world class as far as I can tell, with a level of testing that's basically unprecedented outside of the US (though some other European countries are catching up over the last few weeks) - it's still having trouble coping right now, but that seems to be the case throughout Europe and elsewhere too. That is very much not the impression you'd get from the media. They keep pushing stupid narratives like the UK needing South Korea-style mass testing when in reality the UK is testing about an order of magnitude more people for Covid, and South Korea charges most people with symptoms for testing in order to discourage them from getting tested and using up the limited capacity whilst the UK tests for free.

Also, the government also seems to be working on the testing infrastructure - from what I can tell, the problem is that as it turns out that things like expanding testing capacity has a lot of lead time, and the capacity expansion is a couple of weeks later than it ideally should have been.

> Right now in the UK, the basic testing infrastructure is more-or-less world class as far as I can tell

If you believe the government's figures it is, but they have history in misrepresenting the data.

The leak a day or two ago suggests they are inflating the figures again (62k tests per day actually carried out v. claimed capacity for 375k a day).

See e.g. https://news.sky.com/story/coronavirus-covid-testing-program...

I appreciate the challenges are significant, but I'd have a lot more confidence in the system if they were just honest about the problems they were facing.

That's not what the article you linked says. It actually says that only 62,000 people are getting tested for the first time every day. This isn't officially published as a daily stat anymore because it's confusing and not all that useful - there are good reasons why people get tested more than once, for example because they develop potential symptoms again or perhaps more commonly in order to protect hospitals and care homes by testing people before they're admitted to hospital, during their stay, and before they're discharged to a care home. The number of repeat tests is also going to keep going up as time goes on for really obvious reasons that have nothing to do with testing failing.

Now, admittedly it is worded in such a way as to trick people into making the exact mistake you just did. These two sentences are a work of art: "Labelled "official: sensitive", the Department of Health and Social Care report seen by The Sunday Times says most British laboratories are processing fewer tests than their stated capacity due to problems in supply chains. The newspaper said that despite government claims that its system has capacity for 375,000 tests a day, the equivalent of just 62,000 people a day were newly tested in the first week of September." It's very carefully structured to make it sound like only 62,000 people are being tested a day and the rest of the capacity is going idle due to supply chain issues, even though in reality that's not what's happening and there is pretty much no casual link between the two facts it's using to make people think that.

(The testing delay figures also look, off-hand, very similar to what I've seen from other countries around the globe lately.)

OK, I am sure you are correct. But the government is doing a lousy job of presenting the data and their history of massaging the figures does not inspire confidence.

Their assignment of tests is also awful. I know several healthcare workers who have been out of work for a week while trying to get tested, and surely they should be prioritised.

I don't think total number of tests processed is the only criterion for labelling the system "world class".

"the basic testing infrastructure is more-or-less world class as far as I can tell"

No, it isn't. There's a significant backlog on providing tests and significant backlog on analysing tests. The BBC's More Or Less radio program (available on BBC Sounds) has been explaining for weeks how the test numbers aren't reliable and aren't as impressive as people think.

I don't know who "They" who "keep pushing stupid narratives" about SK-like testing are (perhaps a link would help?). I haven't seen suggestions to that effect.

> Right now in the UK, the basic testing infrastructure is more-or-less world class as far as I can tell

Today for the second day in less than a week the country has no testing availability whatsoever according to the official website.

World class?

> Right now in the UK, the basic testing infrastructure is more-or-less world class as far as I can tell, with a level of testing that's basically unprecedented outside of the US

I get tested every week (for months now) as part of the Oxford vaccine trial, so I have a lot of exposure to the UK government testing system. I'm also an immigrant from the US and I have very little UK political bias since I don't really even understand the political parties here yet and I can't vote here.

Very early in the pandemic, testing was terrible and basically non-existent. By July, the system was great - efficient, easy-to-use and quick. Through July and most of August, it was getting better every week with very short turn-around times and it was very impressive overall. But in the last ~3 weeks, it's has kind of fallen apart. Instead of getting results the next day, I'm lucky to get them the same week. The vaccine trial used to hand out stacks of new tests like candy to take home. Now they are rationing them very closely and clearly feeling the pinch.

I think there are a lot of hard challenges here, from expanding capacity in a country that didn't really have an existing lab infrastructure, to trying to figure out the right social incentives so that the right people are getting tested but we aren't wasting limited resources on lots of people who don't need tests until testing is either faster/easier or more available. And until about 3 weeks ago, I think the government had overcome the challenges. But now they are struggling again due to the uptick in cases and increased need for tests.

In any case, it's unfortunate that the whole situation is so politicized, especially by the media. The media is mostly failing to contextualize the actual situation and just picking sides in a US-style 'my team vs. your team' battle where everyone loses.

I feel like its partly an unwillingness to make difficult decisions and maybe even to look bad. A discord perhaps in the reality versus a desired fantasy where it looks like we can defeat this virus without sweating. It disappoints me that we'd let the system _obviously_ fall apart instead of actually expecting this outcome and rationing use (especially giving the NHS priority).

I think one disappointing scenario that contributes is the desire to push people back into work (as is happening right now), it appears that desire is ultimately in conflict with our desire to ration testing so its bizarre that we'd allow these competing notions to organically compete as opposed to specifically picking a winner based on our inherent capacity.

If we stop sending people back to work or school then the amount of testing required drops significantly. It not people's fault for "wasting tests" the increased risk we've specifically chosen to take on as a country as we ease restrictions creates the conditions for this outcome. Lets not forget we permitted shielded individuals back out into society and to accommodate that we should be "wasting" tests for the people that directly interact with them.

I think part of the larger argument is that modern FPTP politics is so partisan its very hard to judge anymore what criticism is valid and which is just mud slinging. _Everything_ the opposition does is the _worst_ thing ever and nobody ever congratulates the other side on anything they do that could be good. This makes it hard to separate out hyperbolic criticism of borderline cases and legitimate criticism of clear outliers.

At its core I personally believe the issue is the "side taking" but its rare to find positions that are not explicitly called in advance of the news hitting.

This is an impression that TV and Twitter give. But politicians in all countries often say positive things about policies and ideas they like. But that’s not newsworthy so it appears to not be common.

I think you're just conflating two quite different approaches. The US, for example has ordered 150m tests in it's latest deal - or enough to test half the US population. The UK is announcing a project to do 10m tests per day - or about the whole UK population in a week. The NYT currently has this infographic [1] about whether states are testing enough and whilst it varies depending on how the states are doing, but no one is close to suggesting that they need 15,000 tests per capita per day - it's in the hundreds per capita per day.

Also, there's just an obvious messaging side to this - the plan is being deliberately amplified in the UK with "leaked" documents claiming the plan is incredibly difficult and using technology that needs to be invented, but let's use their analogy - the moonshot. The moonshot required genuine new technology to be invented, it took the best part of a decade. Suggesting that you're going to invent new technology for a problem you're also claiming to solve within 3 months is just not credible.


(Needless to say, pretty much all the problems with false positives, inadequate technology, the diffculty of scaling up, etc would have been much bigger problems in the US a few months ago given the much larger population and the improvements in Covid-19 testing technology since then.)

But had we initiated aggressive testing earlier, the population needing to be tested would've been much, much smaller. A single test now, with the virus widespread in the population, is worth much, much less than a test back in February. Testing effectiveness should not be measured in tests/population, but in tests/active infections. We had a great deal more leverage back in February and March, and a little testing back them would've been equivalent in effectiveness to massive amounts of testing now. It's a moving goal post, not for political reasons, but for clear epidemiological ones.

Not quite accurate. It's not "basically identical". The "moonshot" proposal is that you'd be able to walk up to, say, a cinema entrance and get a PCR test right there which would instantly tell you whether you're +ve for Covid. (Not antibody-positive, but infected/infective-positive.)

That's very different from what has been suggested for both the US and UK, which is to have widespread and easy access to PCR testing that could be turned round in, say, 24 hours so that you'd go to a testing station and then be contacted the next day to be told your status.

Given the biotech capacity of the UK and US, that sort of testing capacity should have been well within reach in about April. (However because the US decided not to go with the WHO recommendation on what RNA elements to test for, they ran into problems. I can't find out whether this is the same for the UK.) Germany was able to ramp up its testing capacity that quickly and thus didn't get the sort of criticism that the UK and US governments did.

The gap between the moonshot proposal and the suggested solution is gigantic. Speed, accuracy, availability - all of them come into it. Plus you also have to ask: why do it, when you have alternatives which could come on stream well before the moonshot tech is developed, in the form of multiple vaccine programs? And why suggest you're going to spend that amount? (£100bn sounds made up, if I'm honest. Why not 10? 50? Because 100 sounds big and round.)

So this isn't about media coverage somehow being "partisan horseshit". It really is about a proposal which, given the multiple missed targets and continual lying (about testing capacity, testing times, test facility availability and contact tracing success - all are worse than ministers admit in public statements, but are revealed in data from government departments) from the UK government, simply can't be taken seriously.

Coronavisrus seem like complete hyperbole at this stage. 1% of deaths in the UK were from Covid yesterday.

Right, because of the drastic measures taken. If these measures are halted, that figure will skyrocket. The pandemic isn't over in the UK, it's currently under control, despite incredible expense.

Also, people who get the virus and survive may still have permanent lung damage. There's more to it than the death rate.

It not orders of magnitude worse than a bad flu season, yet the response has been unprecedented. Sweden has managed fine without any lockdown (and it's not a competition over death rates before someone starts comparing). Hospitals didn't get overwhelmed anywhere which was what the reason for lock down was all about (originally). They appear to have reached some level of herd immunity.

If hardly anyone is dying can it still be classed as a pandemic? No one bothers about a "cold pandemic", which is effectively what coronavirus is at this stage. Deaths per day are in single digits.

And you can't claim "permanent lung damage" for something that has only been around for 7 months.

It’s not “orders of magnitude worse than a bad flu season” precisely because the entire country shut down for months

> precisely because the entire country shut down for months

The way to test your hypothesis would be to look at countries that didn't lock down.

Sweden didn't shut down and had a similar outcome to most European countries. Hospitals were not overwhelmed. Looking at Euromomo they had it relatively mild in terms of excess deaths for the time of year. https://www.euromomo.eu/graphs-and-maps/

Last I saw the doubling time for infections was 1 week.

Does it matter if deaths are not going up?

> Does it matter if deaths are not going up?

Deaths lag infections. People take a long time to die. This naive way of thinking that infections are going way up but the death rate isn't and so it was all overblown keeps playing out again and again, and it needs to stop.

>Deaths lag infections.

But the numbers don't show that at all if you look at the "second wave" that we are having here in Spain. There is a second wave of infections, but deaths are staying very low despite all the media hype.


The hype is what needs to stop.

With deaths lagging by ~21 days, though, that data only shows deaths for cases from around the 19th of August - so the data is showing a lower rate, but it still seems sensible to be cautious about the death rate.

I don't think it's at all reasonable to classify caution as hype... in addition to the increased risk of fatalities there are plenty of reports about damage to heart tissue even in young people who didn't have serious symptoms[0] - and if people start to ignore the risk and reduce measures like distancing and mask wearing there IS much greater scope for increased fatalities and long-term health impacts.

[0] https://www.mayoclinic.org/diseases-conditions/coronavirus/i...

If you look at the two graphs in the link I posted we have passed the second peak of infections more than 21 days ago, so we should have expected the spike in deaths already. The graphs don't show that.

Interesting. Could be either test capacity going up, or medical practices improving, or more infection of young people?

Good advice I heard was that hospitalisation with Covid symptoms is a better measure than deaths (happens sooner) or positive tests (not altered by increasing testing).

This report has that for Uk but alas only updated every week and then nearly a week late. https://www.gov.uk/government/publications/national-covid-19...

You can see up to date data on hospitalisations at https://coronavirus.data.gov.uk

These are the measures to watch as they were unavailable at the start of the pandemic. Hospital admissions, Covi-19 occupied beds and ventilator beds have all roughly doubled from 1-14 Sep.

Good article about the PHE Covid-19 Dashboard here - it's quite recent: https://publichealthmatters.blog.gov.uk/2020/09/04/the-covid...

Yes, but doubling from practically none is still a low number, compared to the peak a few months back.

The covid survey seems to show current rates are roughly a quarter what they were towards the end of the peak in the UK.


Exponentials are very unintuitive.

Applications are open for YC Winter 2021

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact