Something that I've worked on a bunch professionally, but never written about here: online identity verification is really important for access to government services, access to which is increasingly intermediated via the internet. The current approach to verifying identities generally relies on verification of information in credit records, which is bad and foolish for lots of reasons. The proper solution is the remote presentation of mobile driver's licenses. That's ~3 years off, I regret.
In August 2021 I wrote an article on this topic for the Federal Reserve Bank of San Francisco’s publication, Community Development Innovation Review, in an issue dedicated to fintech, racial equity, and an inclusive financial system, for anybody who wants to learn more about the inequities baked into the paradigm of online identity verification that exploded into public use in 2020. https://www.frbsf.org/community-development/wp-content/uploads/sites/3/jaquith-americans-need-a-digital-identity-system-stat.pdf
The American Association of Motor Vehicles (a coalition of every state DMV) created the mobile driver’s license standard. Right now it's goofy—a way to store your license on your phone to show in person instead of a card. Which nobody asked for. But they're working toward something vastly more useful: a way to store your license on your phone and present it to a remote party. Kind of like signing into a website with your Google credentials, but instead your DMV credentials.
I forecast a huge freakout about mobile drivers licenses (mDLs), ginned up by industry.
Here's why: At the moment, a dozen private vendors are successfully inserting themselves into the online equivalent of a mundane, decades-old interaction: a member of the public showing their state-issued ID to a government employee. This interaction used to cost literally nothing. Now big contracts go to vendors to do this, because there *is* no online equivalent of a state-issued ID.
You know how we've reached the point where half of congress says that it's totally inappropriate for people to be able to file their taxes online directly with the IRS? That the Good and Right thing is to have to pay H&R Block or Intuit?
That's the play that private identity vendors are making. They’re racing for the perception of incumbency before states can catch up with mDLs, before Login.gov can work with those states to ensure that anybody with a DMV record can access government services.
That will restore the normal order of things in the U.S., where you show a state ID to a local, state, or federal agency to prove your identity. Private vendors—especially credit agencies—are going to fight this tooth and nail, insisting that mDLs are some kind of a horrible encroachment on your privacy, and that the best thing would be for them to get billions of dollars in government contracts to mine your financial records and then quiz you on them to prove your identity.
The actual mechanics of complying with this supposed norm are dizzyingly impractical. Which aspect of a post might strike people as uninteresting? Through which lenses might they not want to learn about what I'm writing? Crafting hashtags for discoverability is one thing, but for *anti*-discoverability? Am I to red-team my own posts?
I receive these cheerful suggestions any time I write anything that’s widely-boosted. I've ignored them so far, and intend to continue to ignore them.
One weird and off-putting thing about Mastodon culture is the idea that one should use hashtags not for discoverability, but for the opposite—so that people need not be exposed to concepts that they aren't interested in…but for some reason they follow people who boost such uninteresting posts.
Here somebody suggests that I use a hashtag about US politics to keep him from seeing something I wrote that has nothing at all to do with US politics.
@jessamyn Anecdotally, I think these requests have always come from people on different instances, and I think a large percentage have been from other countries? I’ll have to observe more closely.
Blog entry: The process by which government estimates the cost of a custom software project is opaque, non-interrogatable, and harmful for budgeting and oversight. But there's a better way—measuring level of effort in scrum team years. https://waldo.jaquith.org/blog/2023/08/scrum-team-years/
Eight days ago I lamented that Pennsylvania's timeline to fix the destroyed I-95 overpass was measured in months. This afternoon, Gov. Josh Shapiro announced it will be open again this weekend. The state has maintained a popular livestream of the repairs, which the governor uses to monitor progress. This is a great example of government doing the work and doing it right. Very impressive! https://www.inquirer.com/news/95-live-cam-shapiro-repair-quinta-20230621.html
My work focuses on budgeting, oversight, and procurement of major government software projects, but I’ll be damned if the problem space and solutions don’t map neatly onto transportation. As a nation, we need to be able to make such emergency repairs in *days*. https://wapo.st/3CpQONX
Google says that Bard is “is intended to…not replicate existing content at length” and, that if it does quote at length, it'll cite the page. But I wrote most of this Wikipedia entry, and I immediately recognized that Bard is copying it, word-for-word, at length. Sure, there's a footnote, but this is a straight-up duplicate of a webpage. https://en.wikipedia.org/wiki/Southwest_Mountains
I couldn't convince ChatGPT that the capital of West Virginia is Huntington (it's not, it's Charleston), until I told it to check wv.gov (which of course it cannot do), which caused it to finally concede the point.
Consider this your periodic reminder that ChatGPT doesn't "think" anything but is just completing sentences with statistically-likely conclusions.
Thought follower. Male software developer. Alumnus of 18F, the Obama White House, Georgetown's Beeck Center, the Biden-Harris Transition Team, and the Biden administration. Speaks only for self. he/him