Jan Strehmel:
We find that open source code containing swearwords exhibit significantly better code quality than those not containing swearwords under several statistical tests. We hypothesise that the use of swearwords constitutes an indicator of a profound emotional involvement of the programmer with the code and its inherent complexities, thus yielding better code based on a thorough, critical, and dialectic code analysis process.
A year after outcry, IRS still doesn't offer taxpayers alternative to ID.me:
When the IRS announced last year that it was working to roll out a government-administered tool to securely grant taxpayers access to online services, it seemed like a triumph for critics calling on the agency to end its controversial partnership with ID.me, a service that largely uses automated facial recognition to verify a taxpayer's identity.
But a year and another tax season later, the IRS still offers ID.me exclusively with no alternative vendor or in-person options. The lack of progress raises questions about the federal government's slow rollout of Login.gov and its reliance on a growing industry of private vendors to verify Americans' identities in order for them to access public services online.
Launched in 2017 to offer the public an easy way to securely log in to federal websites, Login.gov was supposed to streamline the way Americans interact with the federal government. Five years later and more than $187 million in government investment later, however, it still hasn't been widely rolled out.
Troll (2022): Norwegian Gojira! It's awesome. And it hits all the compulsories: plucky scientist; 150' tall metaphor; aircraft swatted from sky; terrible secret weapon; the power of love.
Fun (not fun) fact: Frazetta doesn't even get a thank you in the credits for either movie, even though he designed this whole thing. Wikipedia says "After securing Oliver Stone's services, Pressman approached Frank Frazetta to be a 'visual consultant', but they failed to come to terms."
I just woke up and Twitter is telling me "You are over the daily limit for sending Tweets". So things are still going great over there.
Predicting Financial Crime: Augmenting the Predictive Policing Arsenal
In this paper we have presented our state-of-the-art model for predicting financial crime. By incorporating public data sources with a random forest classifier, we are able to achieve 90.12% predictive accuracy. We are confident that our model matches or exceeds industry standards for predictive policing tools.
Crucially, our model only provides an estimate of white collar crimes for a particular region. It does not go so far as to identify which individuals within a particular region are likely to commit the financial crime. That is, all entities within high risk zones are treated as uniformly suspicious.
Recently researchers have demonstrated the effectiveness of applying machine learning techniques to facial features to quantify the "criminality" of an individual.
We therefore plan to augment our model with facial analysis and psychometrics to identify potential financial crime at the individual level. As a proof of concept, we have downloaded the pictures of 7000 corporate executives whose LinkedIn profiles suggest they work for financial organizations, and then averaged their faces to produce generalized white collar criminal subjects unique to each high risk zone. Future efforts will allow us to predict white collar criminality through real-time facial analysis.
Previously, previously, previously, previously, previously, previously, previously, previously.
Please enjoy jwz mixtape 238.
It's been a little while since the last one. Not as long as the previous gap, but still, my rate of discovery has noticeably slowed in the last year....
ChatGPT is, in technical terms, a 'bullshit generator'. If a generated sentence makes sense to you, the reader, it means the mathematical model has made sufficiently good guess to pass your sense-making filter. The language model has no idea what it's talking about because it has no idea about anything at all. It's more of a bullshitter than the most egregious egoist you'll ever meet, producing baseless assertions with unfailing confidence because that's what it's designed to do. [...]
Of course, the makers of GPT learned by experience that an untended LLM will tend to spew Islamophobia or other hatespeech in addition to talking nonsense. The technical addition in ChatGPT is known as Reinforcement Learning from Human Feedback (RHLF). While the whole point of an LLM is that the training data set is too huge for human labelling, a small subset of curated data is used to build a monitoring system which attempts to constrain output against criteria for relevance and non-toxicity. It can't change the fact that the underlying language patterns were learned from the raw internet, including all the ravings and conspiracy theories. While RLHF makes for a better brand of bullshit, it doesn't take too much ingenuity in user prompting to reveal the bile that can lie beneath. The more plausible ChatGPT becomes, the more it recapitulates the pseudo-authoritative rationalisations of race science. [...]
ChatGPT is a part of a reality distortion field that obscures the underlying extractivism and diverts us into asking the wrong questions and worrying about the wrong things. Instead of expressing wonder, we should be asking whether it's justifiable to burn energy at "eye watering" rates to power the world's largest bullshit machine. [...]
Commentary that claims 'ChatGPT is here to stay and we just need to learn to live with it' are embracing the hopelessness of what I call 'AI Realism'. The compulsion to show 'balance' by always referring to AI's alleged potential for good should be dropped by acknowledging that the social benefits are still speculative while the harms have been empirically demonstrated. Saying, as the OpenAI CEO does, that we are all 'stochastic parrots' like large language models, statistical generators of learned patterns that express nothing deeper, is a form of nihilism. Of course, the elites don't apply that to themselves, just to the rest of us. The structural injustices and supremacist perspectives layered into AI put it firmly on the path of eugenicist solutions to social problems.
If the CEO of OpenAI thinks that you are a stochastic parrot, then that means that he doesn't really recognize you as a person. We have a word for that kind of systemic lack of empathy and that word is "psychopath".
I confess myself a bit baffled by people who act like "how to interact with ChatGPT" is a useful classroom skill. It's not a word processor or a spreadsheet; it doesn't have documented, well-defined, reproducible behaviors. No, it's not remotely analogous to a calculator. Calculators are built to be *right*, not to sound convincing. It's a bullshit fountain. Stop acting like you're a waterbender making emotive shapes by expressing your will in the medium of liquid bullshit. The lesson one needs about a bullshit fountain is *not to swim in it*.
Previously, previously, previously, previously, previously, previously, previously, previously, previously.
The House committee investigating the riot avoided detailed discussion in its report for fear of offending Republicans and tech companies:
Congressional investigators found evidence that tech platforms -- especially Twitter -- failed to heed their own employees' warnings about violent rhetoric on their platforms and bent their rules to avoid penalizing conservatives, particularly then-president Trump, out of fear of reprisals. [...]
"The sum of this is that alt-tech, fringe, and mainstream platforms were exploited in tandem by right-wing activists to bring American democracy to the brink of ruin," the staffers wrote in their memo. "These platforms enabled the mobilization of extremists on smaller sites and whipped up conservative grievance on larger, more mainstream ones." [...]
That focus on Trump meant the report missed an opportunity to hold social media companies accountable for their actions, or lack thereof, even though the platforms had been the subject of intense scrutiny since Trump's first presidential campaign in 2016, the people familiar with the matter said.
Confronting that evidence would have forced the committee to examine how conservative commentators helped amplify the Trump messaging that ultimately contributed to the Capitol attack, the people said -- a course that some committee members considered both politically risky and inviting opposition from some of the world's most powerful tech companies, two of the people said. [...]
The Washington Post has previously reported that Rep. Liz Cheney (R-Wyo.), the committee's co-chair, drove efforts to keep the report focused on Trump. But interviews since the report's release indicate that Rep. Zoe Lofgren, a Democrat whose Northern California district includes Silicon Valley, also resisted efforts to bring more focus in the report onto social media companies.
Facebook secretly killed users batteries, worker claims in lawsuit:
The practice, known as "negative testing," allows tech companies to "surreptitiously" run down someone's mobile juice in the name of testing features [...]
"I said to the manager, 'This can harm somebody,' and she said by harming a few we can help the greater masses," said Hayward, 33, who claims in a Manhattan Federal Court lawsuit that he was fired in November for refusing to participate in negative testing. [...]
"Any data scientist worth his or her salt will know, 'Don't hurt people,'" he told The Post.
Killing someone's cellphone battery puts people at risk, especially "in circumstances where they need to communicate with others, including but not limited to police or other rescue workers," according to the litigation filed against Facebook.
"I refused to do this test," he said, adding, "It turns out if you tell your boss, 'No, that's illegal,' it doesn't go over very well." [...]
He said he doesn't know how many people have been impacted by Facebook's negative testing but believes the company has engaged in the practice because he was given an internal training document titled, "How to run thoughtful negative tests," which included examples of such experiments being carried out.
"I have never seen a more horrible document in my career," he said.
We don't know much else, because Hayward's employment contract included a non-negotiable binding arbitration waiver, which means that he surrendered his right to seek legal redress from his former employer. Instead, his claim will be heard by an arbitrator -- that is, a fake corporate judge who is paid by Facebook to decide if Facebook was wrong. Even if he finds in Hayward's favor -- something that arbitrators do far less frequently than real judges do -- the judgment, and all the information that led up to it, will be confidential, meaning we won't get to find out more.
SF's Market Street Subway Is Running on Reagan-Era Technology:
"Our train control system in the Market Street subway is loaded off of five-and-a-quarter inch floppy drives," SFMTA director Jeffrey Tumlin told KQED's Priya David Clemens this week. "We have to employ programmers who are experts in the programming languages of the nineties in order to keep running our current system. So we have a technical debt that stretches back many decades."
Ah yes, the withered programmers from the Nineteen-Hundreds. I'm glad they included these historical fun facts about floppy disks!
For those who've never had to use one -- and keep it far away from any magnets -- floppy disks are still used as the "save" icon on computers worldwide, but they've been obsolete for 30 years or more. Originally a full eight inches across when introduced in the 1970s, floppies later shrunk to five-and-one-quarter inches, and, eventually, a mere three. Early generations were indeed bendable, although the final generation was made of a harder, thicker plastic, with no flop at all. Eventually, CD-ROM technology superseded them before becoming nearly extinct, too.
Here's an article from 2015 about some upgrades, but this sounds like an even older, pre-floppy system:
Muni train control system gets biggest upgrade since the '90s:
In a dank walkway tucked away inside the Van Ness Avenue Muni station is the mainframe that controls Muni trains underground.
Relay racks extend down a cramped hallway nearly 15 feet deep, where dozens of 4-inch copper filaments clatter up and down like teeth. The speedy "click, clack" sound signifies a smooth running Muni light rail automatic control system [...] The clapping beat signifies the proper alignment of the railway's 83 axle (controls) and numerous other track switches. [...]
When this happens, or a track switch is damaged, a green relay cable running the length of the underground track is tripped. This signals the maintenance crew, which then checks each ticking copper relay for those with an off beat, or ones that have stopped altogether. [...]
"Last year, we had water up to here" Haley said, gesturing with his hand to his waist. The relay system is under a city drain, he said. A yellow tarp is strung at the roof of the relay room, while sandbags, a bucket and more tarps are tucked away in the corners. [...]
"These things are hard to replace," Kelly said. "You can't just go to Relays R Us." [...]
The crew that maintains and inspects the relays every night said it is hopeful for the new system -- but it also respects the old one. "I trust it to run another 20 years," said Hoa Huynh, a 20-year maintenance crew member. But, he said, "It's slow. Very slow."
I love the idea that part of the diagnostic process is to stand in this room and listen to it -- "that's not how the tunnel song is supposed to go." An interface that draws big red boxes using a supercomputer is not necessarily easier or more expressive than one that communicates ambiently through the system's inherent musicality.
Previously, previously, previously, previously, previously, previously, previously.
