Presumably because the article linked above says that the rumoured deadline for gaining sufficient market share is this year? Also, yes revenues have been growing, but is GCP profitable after all this time? As you point out, it's been 4 years since that article was published.
So, they suggest that you spin up the dependent services as part of the test runner and run tests for your service against those? I've worked on a codebase that did this extensively. Over time it lead to a test suite that takes a long time to run, slowing down the release process considerably, and inevitably requiring further engineering effort to make the tests run in a reasonable amount of time. I can't say I would recommend it as an approach based on my experience with it, but YMMV.
> FYI, The Independent has a reputation for writing ridiculous clickbait articles. When you look into the details things almost always turn out to be much less dramatic than the article made it sound.
While more reputable news sources have not yet picked up this story, if it is indeed true, there is a very marked difference between what is claimed in the article and what has been said by the government to date. The article claims that the Home Secretary will attempt to enact secondary legislation to end freedom of movement, while the gov.uk link in your tweet claims that it will be done ASAP through primary legislation.
If true this is an important difference as we likely won't know until close to the October 31st deadline whether no deal is actually going to happen (though it is clearly more of a risk now than ever), and it implies that the government plan not to wait until they achieve a majority in parliament for the bill to abolish freedom of movement, but instead plan to abolish it any way they can.
At Genomics plc we are working to transform drug discovery by understanding the human wiring diagram. Our algorithms overcome the limitations of other methods, and are capable of working at an unprecedented scale of data: millions of potential links between genetic variation and thousands of human traits and diseases. We recently raised a £25M series B round (https://www.genomicsplc.com/genomics-plc-raises-series-b-rou...) and announced a collaboration agreement with Vertex Pharmaceuticals (https://www.genomicsplc.com/vertex-and-genomics-collaboratio...).
We are building a team of collaborative software engineers to transform drug discovery and improve human health. We value well-crafted software and clean architecture, developed through a steady process of incremental change. Our software developers continually strive to master new technologies and techniques, and always aim to choose the right tool for the task at hand. We reserve 10% of our working week for the exploration of new ideas, and share what we have learned with each other through mentorship, pair programming, and tech talks. We work in productive partnerships with our data scientists, statisticians, and domain experts to advance our understanding of the human wiring diagram.
I cannot believe that the paragraph you cite actually appears in a journal article. It's really rather unprofessional and silly.
I don't want to end up defending the methods ProPublica have used as I am certainly not qualified to do that and have no skin in this game anyway. I posted here initially in response to a very general question about bias in models, and I'd rather not be drawn into a lengthy discussion about this specific piece.
However, I do have one or two issues with the conclusions you seem to be drawing in your comment:
> What I find remarkable is that in the ongoing coverage ProPublica has published on this subject in December 2016 they interviewed a bunch of more people, but none of the folks that have criticized their analysis (published in September). Make of that what you will
I'm not sure it's possible to conclude anything from that actually. There could be plenty of non-nefarious reasons for the omission. For instance, one individual they cite in the follow up review [1] has written a paper citing the paper you linked to showing that "the differences in false positive and false negative rates cited as evidence of racial bias in the ProPublica article are a direct consequence of applying an instrument that is free from predictive bias to a population in which recidivism prevalence differs across groups". [2] The Flores et al paper seems to claim that showing that predictive bias does not exist is enough, which it would seem is not the case. If racial bias might appear anyway in the situations in which the model is often applied in reality, then perhaps the ProPublica authors felt that the paper cited below [2] adequately addressed the criticism of the paper you cited and decided not to reference the FPJ article for reasons of clarity in their follow up? I think discounting their work because of a single omission would be throwing the baby out with the bathwater.
The ProPublica authors cite plenty of other research in the area in their follow ups. Sure, this is all largely in agreement with their conclusions or go further, but does this matter unless those publications are incorrect? The ProPublica authors are writing for a news publication not an academic journal and are therefore not obligated to cite every relevant publication when they're publishing. So long as they can do this without forcing a conclusion then I don't see the problem. Perhaps they deliberately ignored the paper. Who knows?
> That's not a good facsimile for deciding sentencing
Sure. I'm not saying my example is the smartest (although there are people trying to use ML to do this to be fair - http://www.fakenewschallenge.org/) and I was unaware of your experience with ML, so I was working under the assumption you had no experience with it and trying to pick a simple example (even if it's dumb) to explain how bias can sneak into models in general, rather than in the specific case of sentencing criminals.
Let's loop back to what you originally said:
> If you're writing a machine learning application to take a dataset and match future inputs to past results I don't see how these biases can sneak into the program
You then go on to describe a number of factors that you think should go into sentencing models that leave plenty of scope for bias:
* "If they meant to do it" - this is a judgement made by a person and clearly reflects the view of the person making the decision.
* If they feel bad about doing it - again, someone has to judge whether someone is legitimately remorseful or is trying to pretend they are to get themselves a lighter sentence.
* "If they have done it before" - This will reflect things like policing tactics. For instance poorer areas might be subject to higher rates of policing especially in areas adopting the broken windows theory of policing (https://en.wikipedia.org/wiki/Broken_windows_theory#New_York...) and often in these areas petty crimes are cracked down on more frequently. This means that people are more likely to have run ins with the law, meaning they're less likely to get jobs due to convictions showing up in background checks, which in turn increases their likelihood to reoffend.
* What severity this crime is - I'm not sure what you mean by this. Do you mean e.g. murder being more severe than petty theft, or things like how severe an assault was? I'm assuming the latter since the former is often just covered by things like sentencing guidelines anyway. If someone commits an assault, then how do you rate this in a way that a model can understand? How do you ensure consistency across different cases and judges?
At any rate, the point of these models is usually to remove the biases that judges might have about people of certain backgrounds from sentencing guidelines and produce a score that informs the likelihood of the convict reoffending (I believe), so your proposal isn't how this works in practice. In practice they're trying to avoid exactly these kinds of subjective assessments you proposed and replace them with supposedly objective predictors for the likelihood of the person in question to reoffend. From the article linked from the post: https://www.propublica.org/article/machine-bias-risk-assessm...
> Northpointe’s software is among the most widely used assessment tools in the country. The company does not publicly disclose the calculations used to arrive at defendants’ risk scores, so it is not possible for either defendants or the public to see what might be driving the disparity. (On Sunday, Northpointe gave ProPublica the basics of its future-crime formula — which includes factors such as education levels, and whether a defendant has a job. It did not share the specific calculations, which it said are proprietary.)
> Northpointe’s core product is a set of scores derived from 137 questions that are either answered by defendants or pulled from criminal records. Race is not one of the questions. The survey asks defendants such things as: “Was one of your parents ever sent to jail or prison?” “How many of your friends/acquaintances are taking drugs illegally?” and “How often did you get in fights while at school?” The questionnaire also asks people to agree or disagree with statements such as “A hungry person has a right to steal” and “If people make me angry or lose my temper, I can be dangerous.”
Given that independent research seems to confirm that the company's model seems to favour higher sentences for people of colour, it's pretty clear from that description where biases could sneak in to the model, I hope?
As with anything, it's hard to generalise without oversimplifying, but here goes. You don't generally just have a data set and a machine learning algorithm that somehow magics outputs from a data set. Usually decisions have to be made by people either in training the model, selecting variables that are included in a model, etc.
Here's a simple example. Say you're trying to come up with an algorithm that decides whether articles in a data set are "fake news" (topical, I know). We have to tell the algorithm whether a given article in the training set is fake or legitimate, otherwise how would it know? Clearly this will reflect the views of whoever is tagging the articles. When we run the model on a training set we need to score how well it did, again this will reflect the opinion of the person doing the scoring.
There's precedent of sorts for this though. Currently a non-EU citizen with a visa to Ireland still needs a visa to travel to the UK. This could be policed in the same way as that case.
Passwords should usually not be encrypted. Saying a password is encrypted implies that there is a decryption function that can produce the original password in plaintext. This should not be the case. Instead, typically, a key derivation function like bcrypt, scrypt, etc. is applied. The output of these are sometimes referred to as "password hashes" because like hashes they are not reversible.
This really needs a "disturbing content ahead" warning on the parts of the site that have graphic images. I really don't expect to see decapitated bodies when I randomly click a link.
It's possible for a library to return an error to indicate that a file is not found when you try and open a non-existent file. Then it's possible for a function that calls the open file function to decide whether that error should be translated into an exception i.e. should that file being missing break out of the current execution? There's no race condition there.
I don't think anyone reasonably expects a library to know the context of the file being opened and its importance in the logic of the rest of the program.
I should have qualified that by saying that there were a lack of services that suited our particular needs. We ended up concluding pretty quickly that for our purposes services like Mailchimp would have been prohibitively expensive. Also, we weren't particularly interested in running a campaign oriented service, which seem to be what Mailchimp and most of the services on your list are for. I don't remember seeing Sendgrid though. That might be what we were looking for, a simple API for sending mails that aren't tied to campaigns, but which provide the template niceness of services like Mailchimp.
Seems to me like this has the potential to be something I'd pay for, but confusingly the article focuses on MIME, when the real problem you seem to be solving is providing a more usable abstraction over the complexity of sending email in general.
Having wanted a service to send emails to customers recently, I hit upon a distinct lack (or apparent lack at least) of services that allowed me to do this. While this should sound ridiculous, sending email reliably is a hard thing to do these days.
I understand that. Just from reading your comment I didn't know whether you were referring to the dead language, a nonexistent one, or the dialect...I'll go sit with the rest of the pedants ;)