NHS to investigate Palantir influencer campaign as possible contract breach
goodlawproject.org
goodlawproject.org
Ideally the NHS would just build it in house, but sadly it's a slow and bloated organisation unable to innovate (as most government managed things usually become).
As an example for the second case in DoD space, there’s Advana.
They are using palantir foundry, which is palantir's big data platform, or how they call it: "The Ontology-Powered Operating System for the Modern Enterprise"
That is sweetly naive, unless you are talking about their marketing department
To see why, you can simply ask yourself: do you think that the unelected officials overseeing government agencies that embark on enterprise software development projects have sufficient expertise and enterprise software project management experience to be able to do this well?
Furthermore, do you think that the quality of engineers that the NHS or DoD can attract with less than half of the compensation of an actual software company stands a chance at developing something good in house?
It’s unfortunately almost impossible for these projects to go right.
That experience speaks more to the perils of in-housing, not to why Palantir is the best COTS for specific needs here. Are there specific leading COTS here you view it so far ahead of for such a contract?
Closer to our own practice.. Modern LLMs have basically reset the field for SOTA in this space, with Palantir, by definition, being behind OpenAI in the most basic tasks, and thus being in the same race as everyone else to retool. Speaking from our own USG experience, we are deep tech leads in some other intelligence areas (graph, ...), and before OpenAI, often chose to adopt prev-gen leading LLM models (BERT, ...) for tasks closer to the NLP side as we recognized that wasn't where our deep tech had an inhouse advantage. We basically had to start over on some of those projects there as soon as GPT4 came out because it just changed so much that the incumbent advantages of already being delivering on a contract were a dead end for core functionality, and almost a year later, it's now obvious that it was the right choice when we get compared to companies that haven't been. Palantir has been publicly resetting as well for using GenAI era tech, which suggests the same situation.
Palantir is multiple main things, and a whole ton of custom software projects on top, and a good chunk of them rely on the quality of their NLP & vision systems for being competitive with others. My question relates to the notion that they are inherently the best when, by all public AI benchmarks, they don't make the best components and, in the context of air-gapped / self-hosted government work, don't even have access to them. Separately, I'm curious how they relate to their COTS competitors (vs gov inhouse) given the claims here. For example, their ability to essentially privatize and resell the government's data to itself and make that into a network effects near-monopoly is incredible, but doesn't mean the technology is the best.
I've seen cool things with them, and on the flip side, highly frustrated users who have thrown them out (or are being forbidden to.) It's been a fascinating company to track over the years. I'm asking for any concrete details or comparisons as, so far, there is zero in the claims above, which is more consistent with their successful gov marketing & lobbying efforts than technical competitiveness.
It just doesn’t make sense to be having this conversation through the lens of AI.
* Databricks data management innovations, now that basics are in, are half on the AI side, like adding vector & LLM indexing for any data stored in it, moving their data catalog to be LLM-driven, adding genAI interfaces to accessing data stored in it, ...
* Data pipelines spanning ingestion, correction, wrangling, indexing, integration, and feature & model management, and especially of the tricky unstructured text, photo, and video nature, and wide nature of event/log/transaction recordings important to a lot of the government, are all moving or have already moved to AI. Whether it is monitoring video, investigating satellite photos, mining social media & news, entity resolution & linking on documents & logs, linking datasets, or OCR+translation of foreign documents, these are all about the intelligence tier. Tools like ontology management and knowledge graphs are especially being reset due to the ability of modern LLMs to drastically improve their quality and improve their scalability & usability through automation.
* Data protection has long been layering on AI methods for alerting (UEBA, ...), classification, policy synthesis, configuration management, ...
Databricks is a pretty good example of a company here. They don't preconfigure government datasets on the governments behalf and sell that back to them, but we do see architects using it as a way to build their own data platforms, and especially for AI-era workloads. Likewise, they have grown an ecosystem of data management providers on top vs single-sourcing, eg, it's been cool to see Altana bring supply chain data as basically a Databricks implementation. For core parts, Databricks keeps adding more of the data management stack to their system, such as examining how a high-grade entity resolution pipeline would break down between their stack and ecosystem providers.
I suspect underneath it's mostly Hadoop but it's impossible to separate the roadmap from the implementation without getting my hands on it.
It is kind of amazing what you can create if you have hundreds (if not thousands) of really well paid SWEs with industry experience of working on real products used by people who pay money for it. As opposed to a bunch of really smart (and really underpaid, looking at the stipend amounts) grad students who know a lot about their specific domain, but lack the industry experience and proper product management.
Add on top of that that Palantir will dedicate engineers just for their big customers, just so that they can assist with integrating and customizing the product the way the customer wants. Palantir literally has a role for that called Forward-Deployed Engineer, and they will fly out to wherever the customer is to address whatever issues they have. Even if all else was equal, there is no way a university would be able to do that.
P.S. Don’t get me wrong, universities can produce groundbreaking and amazing software that will live on outside of the university walls. But it would be usually rough cutting edge prototypes with zero customer support of the type that companies and governments typically look for.
Several government departments have robust in-house data platforms already. The Ministry of Justice does, the Department for International Trade as well. I've personally worked on the former and have pulled code from the latter.
(the code I've looked at from DfIT was https://github.com/uktrade/fargatespawner )
While there might be some good intentions, buying a vendor product is often better for large organisations rather than running their warehouses, ETL and Databricks. Just too much overhead and specialist skills required.
I wouldn't be surprised if this Palantir stuff involves an extremely onerous contract, and probably terms that the NHS cannot even afford long term, and the Tories knowingly entered into it any way.
By that I mean it is impressive how this could happen in one of the foremost liberal society with a large well-funded media representing the whole political spectrum, fiercely adversarial two-party politics, respect for institutional whistleblowers (so long as they don't involve foreign policy) and finally an independent civil service. All of those layers failed to put a stop to this or even to shed an intense light on it for the public. Far more than the conservatives had failed to let this get through.
In fact what is amazing is the public barely aware of its existence as a news story. Try asking a few of your non-tech friends about it. Even the twitter mob failed on this.
I'm reasonably sure the UK media tilts mostly rightwards with only a tiny strand on the left and a small chunk of centrism.
Also performed by the conservative government?
I mean that sets the precedent for kick-backs as well, so I wouldn't be surprised if, after losing the next election, a bunch of sitting Conservative MPs go and mysteriously work for Palantir (Greensill all over again and - oh and hey, David Cameron's reputation is being rehabilitated right now by having him in government again, funny that)
Can you explain how someone can "hate the NHS" while drowning them in cash far beyond what they need to operate?
Drowning them in cash that they have siphoned off to their own cronies and companies that give their MPs kickbacks.
Not unlike this Palantir contract, which is probably going to be Greensill II
while somehow simultaneously providing it with the most amount of funding ever:
https://www.statista.com/statistics/317708/healthcare-expend...
https://www.statista.com/statistics/632289/nhs-england-healt...
but even with these colossal amounts, it’s still underfunded according to the BMA:
https://www.bma.org.uk/advice-and-support/nhs-delivery-and-w...
W Bush did the same thing with the Medicare expansion -- somehow the anti-Socialist party expanded social welfare benefits... to the one group that would vote for them.
This is patently false, a narrative often parroted by people who don't understand or have never worked in the NHS.
There are many intelligent people in technical/digital roles in the NHS that innovate on a daily basis - heck, many build systems internally at a hugely reduced cost compared to outsourcing yet decisions come from parliament/gov agencies that overrule internal decision making and waste insane amounts of cash on vanity projects, or scrapping internal work to be redone by the likes of Accenture.
After all, the UK's health spending is increasing at a rate 2,000% higher than close neighbor France's is -- indicating massive fraud and waste.
For example, suppose in 24 hours France's spending goes up +0% and the UK's spending goes up +0.001%: "Oh my god! The UK's spending is increasing at a rate infinity times more!!11"
My favorite thing about dysfunctional UK political rhetoric is when someone will say "the Conservatives are slashing the NHS!" when what they really mean is that they've slightly decreased the rate of increase, so by every metric it's still skyrocketing (but not enough to those who want to dump infinite cash in the middle management growth machine).
Another fun factoid: the pro-Brexit crew were chided for claiming they'd increase NHS funding by 350m Euros a week. Since Brexit, NHS funding has actually increased by more than double that.
Nothing can stop the NHS cash burning. Expand middle management at all costs. Nevermind that we're outspending comparable countries with comparable health pressures.
To be specific, I mean these charts of those countries' total spending [0] and govt/compulsory spending [1] during that time period. (Using the OECD stats visualization tool [2].)
For total spending [0] UK's spend-growth was less than France's (not greater) and for govt/compulsory spending [1] the UK's larger growth over that period was still in the same ballpark of ~1.31x versus France's ~1.28.
In other words, the nearest number I can derive for your "increasing at a rate X% higher than France's" is around 1.31/1.28 = ~2.4%. Yet you said 2000%! That's a gap of three entire orders of magnitude which desperately need explanation.
[0] https://data.oecd.org/chart/7jcQ
I don't think your data source correctly accounts for inflation, making it essentially useless to see true costs.
As you know, France is a similar country with similar health concerns and similar hospital pressures.
Yet the UK lights taxpayer money on fire on NHS middle management salaries (care certainly isn't improving!) and France has no such problem.
Source:
- https://data.worldbank.org/indicator/SH.XPD.CHEX.PC.CD?locat...
- https://data.worldbank.org/indicator/SH.XPD.CHEX.PC.CD?locat...
I'm still not seeing it, what are the actual digits you are you math-ing in order to get 2000%? Are you sure you aren't comparing two different time ranges?
Because putting both country-lines on on the same 2010-2019 graph doesn't show anything too shocking:
https://data.worldbank.org/indicator/SH.XPD.CHEX.PC.CD?start...
That the NHS has a few old systems doesn't mean jack shit.
Like, there are still mainframes in use in many places, and more than you'd think.
Corruption. Bro leading Palantir has friends in high places. It’s easy.
By subcontracting it out, gov has an easy out.
- “oh there was a bad subcontractor misusing the system. It WaSnT uS!1!!”
- “oh it was just one bad actor within government. We sent person to black site. ItS NoT a SyStEmIc iSSuE”
The NHS could be as slow and bloated as a whale corpse, but doing everything in-house would probably still be more successful than contracting fraudster companies.
Palantir will do a number of things to gain influence. They'll find out who are the decision makers on the space they want a contract, and they'll target them specifically. For example: if they know you take the train to work every morning, and read the news paper, they'll take a full page advertisement on the paper they know you like and that you'll see. They gain this information by, you guessed it, spying on their targets.
What do you think they do that’s so above-and-beyond what a non-government organization should be able to do? I wonder if you’re falling for a lot of the marketing hype…
Publicly traded companies are required to report certain financial facts about the company, but they’re still extremely opaque.
In this case espionage and surveillance, traditionally domains of the state.
>What do you think they do that’s so above-and-beyond what a non-government organization should be able to do?
They're not subject to privacy legislation and limitations that the police is. There's a concrete exapmle of this, commercial storage of license plate data in the US is largely unregulated, so police units that have legal limitations on how long they can hold citizen data simply go to the private market and buy that data from those firms. That kind of circumvention of the law and outsourcing of responsibility is probably a large part of the Palantir business model.
Seems kind of insane to take your tax payer money and use it to buy surveillance data on you, a good chunk of which probably isn't even legal to otherwise collect and completely undermine the rule of law and democratic control of policing.
And they’re not police, so of course they’re not held to that standard, just like you or I aren’t held to that standard.
That's a strong claim given that Palantir has sold and deployed so called "predictive policing" technology to the police forces of New Orleans[1] (without city council knowledge, so much for the accountability) and even directly advertises itself as a defense company[2]. Their revenue is overwhelmingly contract with national governments, again mostly for police or defense work. That's not how you, I or most Fortune 500 companies do business.
[1]https://www.theverge.com/2018/2/27/17054740/palantir-predict...
And it absolutely is how you, I, and every other employed organization works. You seem to not understand what a contractor is.
I think it's practically guaranteed that Palantir is, or will be, blackmailing officials. Even if they never do, any lucid official will be very conscious of stepping on the toes of a private intelligence company.
The incentives are aligned that way so it would require surveillance and conspiracy to prevent it from happening.
Furthermore, collecting large amounts of information on your own population creates a lowest effort point of entry for an adversary to gather that info, and it's more useful to them.
If you write a letter to the newspaper, you might hope to influence one person, or the general audience of the paper. If you feel so strongly about something that you compose an op-ed, you'll also have to do the work of persuading a feature editor to publish it (although this isn't too hard, if it seems like it fills the news hole). That's a bit less accessible than a letter to the editor. If you're buying full page ads, you're leveraging capital that most people don't have to go outside of regular interpersonal communication channels.
I've priced doing billboards just for funsies --- to make a (snarky) point, not to make money. It's not that expensive! Also: how could it be? Look at some of the terrible billboards out there. There are a series of I-55 billboards advertising a single self-published book of poetry.
Your original statement was "normal humans don't try to communicate directly with each other through the media". I think you're using a pretty idiosyncratic definition of "normal" there. The phenomenon you're labeling abnormal is I think actually pretty common.
* "ok, extrovert-feely, DJ this party and be quick about it!"
* "but the CEO said the quarterly results are due by EOD tomorrow!"
* "are you going to oppose the MBTI? you sure aren't a culture fit are you?"
edit1: formatting
edit2: the point of the above flippant story is that the 'test' marked a person for failure and made their departure a self-fulfilling prophecy.
Pseudoscience yields anecdotes. Science yields empirical results.
It was deeply creepy, and left a very unpleasant taste in my mouth.
(Something about building a company of power-seeking people who don't notice, or aren't bothered by, ethical questionability? Some companies have herded those well enough, and thrived for decades, like that.)
This isn’t a cryptocurrency company. It’s at the core of a public service handling some of the nations most sensitive data.
That we fund this company with our taxes is repugnant.
From Wikipedia:
"A major theme of palantír usage is that while the stones show real objects or events, those using the stones had to "possess great strength of will and of mind" to direct the stone's gaze to its full capability.[T 2] The stones were an unreliable guide to action, since what was not shown could be more important than what was selectively presented. A risk lay in the fact that users with sufficient power could choose what to show and what to conceal to other stones: in The Lord of the Rings, a palantír has fallen into the Enemy's hands, making the usefulness of all other existing stones questionable. Commentators such as the Tolkien scholar Paul Kocher note the hand of providence in their usage, while Joseph Pearce compares Sauron's use of the stones to broadcast wartime propaganda"
If you're wondering, yes he is related to Oswald and Max
Really, Palantir's management couldn't have made a finest choice! Absolutely no undertones here. There were absolutely no better candidate.
A combination of terms you'd never, ever want to see together in a sentence.
For example, while it wasn't GLP but Gina Miller challenging article 50, it established that laws can't be revoked except by parliament, and that the SC can police this. Our supreme court is young (est 2006 or so) and is still coming into its powers because those are defined through lawmaking in the courts as well as parliament.
This is not the case at all. The "Supreme Court" in the UK is the latest version of the UK's "court of last resort". Previous to this it was the House of Lords one would appeal to (it had functions to sit as UK's top court as well as the 2nd chamber). The House Of Lords was the Court of last resort for the UK since at least the 1800s.
From the Wikipedia page on the "new" UK supreme court:
"It assumed the judicial functions of the House of Lords, which had been exercised by the Lords of Appeal in Ordinary (commonly called "Law Lords"), the 12 judges appointed as members of the House of Lords to carry out its judicial business as the Appellate Committee of the House of Lords."
Indeed, the first 12 justices on the "new" supreme court where just the exact same 12 Law Lords from the previous House of Lords system.
The 2006 "Supreme Court" rebrand was a continuing reform of the existing system, not some new level of appellate court the UK had never had before. In reality, they both perform the same top appellate court/court of last resort role. You could bring the same types of cases to the House of Lords as you can to the Supreme Court, and the Law Lords presided over many of the UK's most significant constitutional cases.
There are various political reasons why this "rebrand" to Supreme Court was done, but I've typed enough and can't be bothered delving further into this in a reply. It's fundamentally a continuation of the previous long standing Court, and precedent from the many, many pre-existing HoL Court decisions is equal in power to that of the Supreme Court's decisions.
Had Gina Miller's case been say 10-15 years earlier, the HoL would have heard and decided the case too - she did not need the existence of a "new" "Supreme Court".
> https://en.wikipedia.org/wiki/Judicial_functions_of_the_Hous...
> https://www.parliament.uk/about/mps-and-lords/about-lords/lo...
The system of appointment to the SC is rather different to the HoL as it draws from the judiciary and merely being a Lord isn't enough. There was initial continuity in membership, but there has now been nearly two decades of divergence. The HoL was hereditary for much of its history as I'm sure you well know. It is of course now a house of political appointees, cronies, and people who have probably paid _someone_ to be in charge.
https://www.supremecourt.uk/about/appointments-of-justices.h... explains more about the system of appointments. It has an independent commission which makes the selections.
This is where the discontinuity occurs: the fact that as time goes by the judges of the SC are increasingly apolitical. That gives them more standing to make decisions independent of government, and also independent of the historic bias of the HoL (which has been majority Tory for rather a lot of its history).
The thinking goes that the fusion of the SC and HoL as was the case in the days of the Law Lords made the highest court a political one. The splitting of them in terms of form and personnel has fundamentally changed the character of the judgements we can expect. I'd have expected the old HoL to have gone against Miller, which is why I used her as an example.
It is worth noting that had the SC not been split from the HoL after life peerages became the majority of that house, we'd be in danger of those political appointees and the like being the basis of the court. It was therefore necessary and not a "rebrand".
The rebrand to Supreme Court and change to appointment processes do indeed remove some of the risk of perception of political interference, but if you look at the historical operations of the HoL as an appelate Court in the last 80 years, political interference was much more a hypothetical problem than a real one - the HoL routinely passed decisions that greatly upset the ruling parties of the day, and the appointments of Law Lords was greatly removed from the appointments of rank and file life peers etc.
> The system of appointment to the SC is rather different to the HoL as it draws from the judiciary and merely being a Lord isn't enough.
The law lords where almost always drawn from the Judiciary too in modern history, sitting here right now I can't think of one who didn't have a very significant judicial career prior to becoming a law lord. The government was able to simply grant them a peerage permitting them to become a law lord, which is not 1000 miles removed from the Supreme Court's life-long appointments. The HoL required retirement at 75 for a Law Lord exactly the same as is expected for a new SC justice too.
> It is worth noting that had the SC not been split from the HoL after life peerages became the majority of that house, we'd be in danger of those political appointees and the like being the basis of the court. It was therefore necessary and not a "rebrand".
This is nonsense. Only the law lords sat on the court. It was not just "any" peer who could sit on the Court, of which there were only 12 members.
"To be appointed a Lord of Appeal in Ordinary under the 1876 Act, a person was required to have been a practising barrister for a period of fifteen years or to have held a high judicial office—as Lord Chancellor (before 2005) or judge of the Court of Appeal, High Court or Court of Session—for a period of two years. Lords of Appeal in Ordinary were required to retire from judicial office at 70 or 75 years of age, though as barons they continued to serve as members of the House of Lords in its legislative capacity for life."
> https://en.wikipedia.org/wiki/Lords_of_Appeal_in_Ordinary
There are many things that are arguably wrong/undesirable about an unelected second legislative chamber such as the HoL, but the operations of its Court were arguably one of its least egregious short-comings, and very separate from the general workings of the HoL at large.