Forecasting what advanced bullshit generators will do to society under the control of antisocial media platforms financed by advertising.

It confuses the issue by calling them "AI".

The new UK government said it would stop using a barge to house asylum seekers.

This raises two big questions which I do not recall much writing about.

1. Why the UK's handling of asylum requests takes years.
2. Why the barge was such a bad place for asylum seekers to live in.

For (1), I suspect the causes include the Tories' desire to discourage people from seeking asylum in the UK, budget cuts that affect bureaucracy, computerization that replaces human workers, and an attitude of rigidity among functionaries that leads to delay whenever they encounter an unusual situation that the rules don't explicitly address.

For (2) I have seen little concrete coverage of what makes it so bad to live there. There may be aspects of its arrangements and policies that make life on board painful, but I have no idea what they are.

The one thing I do understand is the barge's remote location and consequent inconvenience of access. I think that makes it unfeasible for anyone living there to visit most parts of the UK without staying away overnight — and they can't afford to rent a room for that. I think that would in most cases cut them off from meeting friends, relatives and support organizations.

If that is the main problem, could it be solved by mooring the barge in London, convenient to public transit?

If there are other major problems of living there, I would like to understand them.

An Illinois thug deputy has been charged with first degree murder for shooting dead a black woman he had been sent to help.

At least it is good to see the state authorities filed the appropriate charges without trying to avoid it.

The article linked to just above displays symbolic bigotry by capitalizing "black" but not "white". (To avoid endorsing bigotry, capitalize both words or neither one.) I denounce bigotry, and normally I will not link to articles that practice it. But I make exceptions for some articles because I consider them important -- and I label them like this.

*Human Rights Groups Call on Biden to Investigate Israeli Influence Campaign Against American Lawmakers.*

*Developing countries face worst debt crisis in history, study shows. Spending on health and education being cut as nearly half of budgets are used to pay creditors, campaigners say.*

Side point: I dislike the euphemism of "developing countries." But that doesn't diminish the importance of the injustice.

Identifying the biggest falsehoods in Netanyahu's speech to the US Congress.

*[Rep.] Rashida Tlaib holds up "war criminal" sign during Netanyahu address.*

Bravo, Rep. Tlaib.

Kamala Harris strongly called for reforms to stop the unjust violence of uniformed thugs.

*"Oil Kills" protesters disrupt flights at airports across Europe in wave of action.*

Causing many flights to be canceled is quite annoying, but since the alternative is the collapse of civilization into suffering and death…

Scientific surprise: metallic nodules (lump of mixed meals) that rest on some flat areas of ocean bottoms generate oxygen in a slow but steady way.

They may be crucial for some life on the bottom.

Scientists already warn that mining them could destroy sea-bottom life. This shows another way it could have that effect.

*JD Vance writes foreword for Project 2025 leader's upcoming book.*

Barely six weeks after the XScreenSaver for Android Privacy Policy fiasco, Google, is up to brand new fuckery. As far as I can tell, they're saying that if I don't give them a copy of my driver's license, they're going to remove XScreenSaver and XDaliClock from their "Play" [sic] store.

So that's obviously not going to happen.

And gosh, I'm all broken up about the consequences.

Director	We can't just have someone looking through binoculars, they need to be fancy binoculars with tech on them.
Art dept.	Got it boss. "In porta sed elemen tum eu mattis morbi lectus congue portit or leo", just like you asked!
Director	Wait what
Art dept.	Sorry, I meant: "Harestra mec eget amet arcu eu curabitur mulla dec auctor integer pharetra congue sapien tellus morbi vitae ergis socales ligula malesurdia curobutyr qyus sei vuvrutys duan gravida."
Director	This is an action movie, why are these binoculars summoning the devil?
Art dept.	165.9547 71.9077 17.9717 27.7862 64.9883! Loading! Initializing_datas_acad_tyu191!

Previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously, previously.

I keep posting about the importance of functions inside of deep neural networks being sublinear but haven’t given an exact definition of that before. It’s sublinearity in the computer science asymptotic sense. The Taylor expansion should not only have a linear bound but either going to zero or at least have the positive and negative directions go to different asymptotics. If the function is defined by different formulas in different sections that criterion should apply to all of them.1

With that out the the way here’s how common activation functions look with a new suggestion at the bottom.

RELU

This is as nonlinear as you can get a monotonic function sublinear function to be and is trivial to compute. The one big problem is that it has that kink at 0. Is it too much to ask for a function to be continuously differentiable?

Leaky RELU

It’s okay to have insecurities about some areas being completely occluded and hence stop responding to training but if it’s that much of a problem for you you don’t have what it takes to work with DNNs and should go back to playing with linear functions.

GELU

We get it, you got rid of that kink, but the requirements specified a function which is monotonic, can you not read? Also maybe don’t use functions which are so obscure that I can’t figure out how to enter them into Wolfram Alpha.

Softplus

Thank you for you following the requirements and there’s some argument to using Softmax here since you’re probably using it elsewhere anyway. But it does seem to take a large area to smooth that kink out and never quite gets to exactly RELU in either direction.

RELU with Sinusoidal Smoothing (RELUSS)

This is my new idea. Not only is the kink completely smoothed out, it’s done with a simple quick to calculate function which meets the requirements and reverts completely to RELU outside of that area.

Pondering out loud:

Sometimes I post things to Ye Olde Blogge, and sometimes I post them on Mastodon only. How do I decide which? It has been somewhat random, but if I examine my decisions, I think what I have been doing is:

• Linking to a long funny article, or to Actual Artwork: Blog.

• Someone's good shitpost: Blog if it is a banger for the ages; Mastodon boost if it is an ephemeral Sensible Chuckle.

• Actual News, that will still be interesting next year: Blog.

• Breaking Bullshit News, but funny, but that won't matter tomorrow: I often just boost someone else's Mastodon post, which does not show up on the blog. (E.g. "Republican says something dumb"). Unless it is, like, such a self-own that it amounts to a Quality Shitpost, then I'll re-blog it.

  One reason for boosting others' posts is that I do not see the replies and sometimes that is for the best.

• My own "Sensible Chuckle" shitpost: sometimes I do these on the blog, and sometimes Mastodon only. The problem with the Mastodon ones is that sometimes I misjudge the popularity of these, and my throwaway one-liner gets a dozen angry replies from the Mastodon HOA or the anarcho-syndicalists, and my blog readers might enjoy skimming that pile-on.

  Plus there's the whole "the blog functions as an archive" aspect of things.

So, I dunno. Maybe I should fall into a pattern of: never make top-level Mastodon posts, only post to the blog and mirror that to Mastodon.

Here's a good "where should this have gone" example.

"Always blog" is in the spirit of POSSE, but in my case there's not a lot of "E = everywhere" in that these days -- pretty much just Mastodon -- as I no longer use Facebook or Twitter. (I do still auto-post a screenshot-breadcrumb to Instagram, because I have a few friends who seem to use that the way I use a feed reader.)

Previously, previously, previously, previously, previously.

• Suggesting that the vast majority of the systems that failed due to reliance on not just Windows but on some third-party Windows rootkit were part of the Bullshit Economy and are societally "surplus to requirements".

  (Yes, I know you have read about one counterexample. Please sit down.)

• Suggesting that taxing billionaires is good actually.

  (And several of the people mad about this take were dorm-room-bong anarcho-syndicalists eager to tell me things that I "just don't understand", about like, the futility of voting, that both sides are the same, and presumably the CIA's malign influence on their skunk weed.)

---

Anyway, send BSODs. Reply here with the best photos of Clownstrike BSODs in the wild. Please do a modicum dilligence and don't send years-old BSODs, only the freshest.

Previously, previously, previously.

If you're wondering why the billionaire owners of the New York Times, Washington Post, et. al. are trying to sabotage the Democrats in general and Biden specifically, it's because of tangible accomplishments like this.

For years, the tax agency simply didn't try collecting sizable debts owed by 1,600 people with annual incomes of at least $1 million:

"The tax bill wasn't even in dispute -- the taxes were clearly owed by these people," IRS Commissioner Danny Werfel said in a call with reporters. "But we didn't have the people or the resources. ... It takes time and staffing to work through these cases."

The tax agency, boosted by the 2022 Inflation Reduction Act, has hired hundreds of skilled accountants over the past year and a half after years of shrinking staffing. Some of them have focused their efforts on specific groups of tax delinquents, including a group of 1,600 households with annual incomes above $1 million who were all known to owe at least $250,000 in back taxes, based on their previous tax returns. Facing understaffing in the past, IRS leaders admit, their agents simply didn't try to collect these taxes for years. [...]

In other cases, it took much more effort, and in some instances levies on the taxpayers' assets, to collect the money. Werfel described months of letters back and forth between the IRS's new agents and the millionaires' accountants and lawyers.

The push that IRS officials described this week targeted people who filed tax returns but hadn't paid. The agency is also pursuing a far larger group: wealthy people who haven't filed returns at all.

Previously, previously.

Spotted one in the wild!

This is definitely an improvement. Now the extremely loud fan is running 24/7 to display a PSA against using MICROS~1 products, instead of whatever advertising bullshit was there before.

Advertising shits in your head, so thanks for enabling this upgrade, MUNI!

Sam Altman is the owner of a massive, invasive, parasitical toxic sludge that respects no boundaries and ruins everything it touches, and that he thinks someone else should clean up.

Also his new house has mold.

OpenAI CEO's $27 million San Francisco mansion is ridden with 'hazardous' mold, lawsuit says:

The 950 Lombard St. property was once San Francisco's most expensive single-family listing, boasting six bedrooms and eight bathrooms across its 9,500 square feet along with expansive views of the San Francisco skyline. Friday's lawsuit said it "seemed obvious why" the property was so expensive, and called the home's infinity pool, cantilevered over a hillside, its "crowning achievement." But according to the lawsuit, which was filed in San Francisco's Superior Court against the property's developer and contractors, living there has been one headache after another.

Womp womp.

Previously, previously, previously, previously, previously, previously, previously.