Barry Zuckerkorn @ BarryZuckerkorn @beehaw.org

He's very good.

---

Read more

Posts

2

Comments

190

Joined

2 yr. ago

Oh. That kind of tinder.

As long as no one is getting hurt I don’t really see the problem.

It'd be hard to actually meet that premise, though. People are getting hurt.

Child abuse imagery is used as both a currency within those circles to incentivize additional distribution, which means there is a demand for ongoing and new actual abuse of victims. Extending that financial/economic analogy, seeding that economy with liquidity, in a financial sense, might or might not incentivize the creation of new authentic child abuse imagery (that requires a child victim to create). That's not as clear, but what is clear is that it would reduce the transaction costs of distributing existing child abuse imagery, which is a form of re-victimizing those who have already been abused.

Child abuse imagery is also used as a grooming technique. Normalization of child sexual activity is how a lot of abusers persuade children to engage in sexual acts. Providing victimless "seed" material might still result in actual abuse happening down the line.

If the creation of AI-generated child abuse imagery begins to give actual abusers and users of real child abuse imagery cover, to where it becomes more difficult to investigate the crime or secure convictions against child rapists, then the proliferation of this technology would make it easier to victimize additional children without consequences.

I'm not sure what the latest research is on the extent to which viewing and consuming child porn would lead to harmful behavior down the line (on the one hand, maybe it's a less harmless outlet for unhealthy urges, but on the other hand, it may feed an addictive cycle that results in net additional harm to society).

I'm sure there are a lot of other considerations and social forces at play, too.

I mean Rust is a godsend as a decision for the language to use.

I have no dog in the fight, but Penguincoder has been pretty vocal about Rust being the wrong choice for a web service: slow to develop and modify, easy to make mistakes that take much more work to fix later (and blames this fact for the state of the lemmy codebase). Its greatest strength is the speed of execution, but that doesn't really matter for web servers, that are basically never CPU limited.

I can’t imagine anything even major needing changing, let alone a full rewrite.

I think the moderation tool examples given sound pretty broken, and it isn't just Beehaw admins complaining about them. Lemmy.world and a few others have instance admins complaining about how hard it is to remove images from the server (deleting posts/users/comments just orphans the image file without deleting the associated file), how all the moderation functions seem not to contemplate the federation issue (removing an abusive comment or banning a user on one instance does nothing to address that same problematic content already federated to another instance).

The best conversations happen among small groups of people selected out of a huge, huge pool of people.

Niche interests and discussions need to be able to advertise their existence to millions in order to persuade that dozen people to actually participate.

@penguincoder@beehaw.org is a pretty active dev for Beehaw, has been very open about his views that the lemmy software is built on very shaky foundations, including the programming language and architecture choices underpinning the whole thing, making moderation unnecessarily difficult and making it hard to comply with legal requirements of hosting such a service, and providing severe limits to scale. It might make more sense to build up a new forum from the ground up, compatible with ActivityPub, than to try to fork Lemmy, or persuade Lemmy's existing maintainers to start accepting big patches.

It's not even swipes. It's an overlay showing which potential swipes have are recommended by your chosen recommenders (who can't message or interact with any users). The first step of actually choosing to swipe left or right remains with the user.

It means his segment was more nuanced than I expected.

They always are. I always cringe when I learn that a generalist journalist is covering something in my areas of expertise, because I worry that they'll feed into misconceptions, but the two segments I remember John Oliver doing in areas that I have professional experience, he explained things better than an expert, in a way that laypersons could understand.

So I've come to trust him a lot more than even more serious journalists, despite his show being a comedy show.

Training AI models takes a lot of development on the software side, and is computationally intense on the hardware side. Loading a shitload of data into the process, and letting the training algorithms dig down on how to value each of billions or even trillions of parameters is going to take a lot of storage space, memory, and actual computation through ASICs dedicated to that task.

Using pre-trained models, though, is a less computationally intensive task. Once the parameters are defined on that huge training set, that model can be applied by software that just takes the parameters already defined in training and applies it to smaller data sets.

So I would expect the AI/ML chips in actual phones would continue to benefit from AI development, including models developed many chip generations later.

Google has a great track record at fulfilling its promises of support, but a terrible track record of giving un-promised support. So when they promise support, that should go into the "good track record" column.

Most of the normal apps on the phone are using AI on the edges.

Image processing has come a long way using algorithms trained through those AI techniques. Not just the postprocessing of pictures already taken, like unblurring faces, removing unwanted background people, choosing a better frame of a moving picture, white balance/color profile or noise reduction, but also in the initial capture of the image: setting the physical focus/exposure on recognizable subjects, using software-based image stabilization in longer exposed shots or in video, etc. Most of these functions are on-device AI using the AI-optimized hardware on the phones themselves.

On-device speech recognition, speech generation, image recognition, and music recognition has come a long way in the last 5 years, too. A lot of that came from training on models using big, robust servers, but once trained, executing the model on device only requires the AI/ML chip on the phone itself.

In other words, a lot of these apps were already doing these things before on-device AI chips started showing up in 2013 or so. But the on-device chips have made all these things much, much better, especially in the last 5 years when almost all phones started coming with dedicated hardware for these tasks.

Because this instance doesn't use downvotes. The default interface doesn't show downvoting, and attempts to downvote through another interface are literally discarded by the server.

Rand Paul is in the Senate and isn't a swing vote. There are more than enough votes in the Senate for a CR, with or without him. In fact, the Senate passed a CR 77-19 on Wednesday, that included funding for Ukraine.

The shutdown is happening because the House won't bring a bill to the floor, with or without Ukraine funding.

I'm gonna push back against that defeatist attitude that things aren't worth doing if success can't be guaranteed. First off, as a general matter it's still doing because we don't want that one-way ratchet where only one side occasionally tries while the other will always bring their A game and pull off a few upsets, for an overall winning record. I think that most progressives/liberals are unnecessarily handicapping themselves by not showing up for every fight.

Second, specific to this type of regulation, the "cost of doing business" issue doesn't even really apply. If the punishment for violating a regulation is a fine, then maybe you pay a few fines and it works out. But that's not generally how the FCC works, because although they do have the power to issue fines, the big thing they have is the power to actually order compliance with their rules.

If the punishment for not building a house to code is a thousand dollars in fines, that's not going to stop home builders when they're making hundreds of thousands in profit per building. But if the punishment for not building a house to code is you're not allowed to sell it until you tear it down and do it right, well, then we're talking about a punishment that cuts hundreds of thousands of dollars into lost profits/revenue.

The FCC's regulations are more like that. If the FCC orders the ISPs that "oh that contract where you're accepting money in exchange for fast lane access is illegal, so you can't do that anymore," that now-illegal contract between two big businesses turns into a more complicated effort of lawyers figuring out what they're supposed to do. Does the other side still pay, if they're not getting anything in return? Or if the FCC says that a particular QoS rule on their routers needs to be removed, do the network engineers go back to the drawing board to implement their own traffic shaping stuff that does comply with the regs?

That might be, but the tug back and forth at least gives the ISPs pause before going full bore into engineering (or contracting for) non-neutral arrangements. Why invest the time, money, and effort into something that is only sometimes legal?

Commissioner Anna Gomez was sworn in yesterday. Up until then, the FCC has been deadlocked 2-2 between Democrats and Republicans, so the FCC has been unable to push net neutrality.

They just announced that with their 3-2 majority, one of their top priorities is to get Net Neutrality regs passed. This is an important step, announced like literally the first day they've had control of the commission.

That's true of everyone who worked for Trump and came into the government as a Trump appointee or staffer, but I would argue Milley is in a different boat. Milley was appointed Army Chief by Obama, and never engaged with the political side of things. He did have some eyebrow-raising proximity to crazy Trump things, but even then was viewed as the least crazy person in the room.

I'm sympathetic to the idea that an individual user should be able to override their instance admins' preferences on access for content-related reasons, but I don't think it would be workable from an administrative viewpoint to allow users to allowlist instances that were blocklisted for administrative reasons.

Lemmy.world dealt with (and is probably still dealing with) a series of malicious actions designed to actually bring down the service or otherwise tie up its resources (including moderator/admin attention and effort, and exposure to literal criminal charges), using maliciously crafted requests to bring down servers, literally illegal content posted to their servers, etc. Defederation in response to these types of attacks would be defeated if a user could let the content come through anyway.

I imagine most instances are dealing with similar issues.

So ideally we'd need to be able to create 4 categories of relationships with other instances:

1. Blocked no matter what
2. Blocked by default for users, can be user overridden
3. Allowed by default for users, can be user overridden
4. Allowed no matter what (not sure what the use case for this status would be, but seems to be trivial to implement since it already exists as default).

But I think you'd find that the typical scenario that justifies blocking would actually put the typical block into category 1, not category 2.

"Well technically this isn't a 100% plant based meal because the mushrooms are not part of the plant kingdom"

Where I am, electricity is pretty cheap, but natural gas is tremendously cheaper per jule… so we can actually pay less by using the “inefficient” fuel for our home.

Most of the push towards rapid adoption of heat pumps is happening in Europe, where geopolitical developments (to put it mildly) caused gas prices to spike last winter. The nature of the natural gas logistics means that different continents can have wildly different prices (unlike petroleum, where you can always throw it on a ship and send it from where it's cheap to where it's expensive), so a lot of European countries are seeing these debates play out against the backdrop of their own energy markets. Germany passed a law this year that would phase out new gas furnace installations, so that's why a lot of the debate is happening with a focus on German markets.

Whether (or how quickly) a transition to heat pumps pays for itself in euros will depend a lot on what happens in the future to gas and electricity prices.

I’m a little surprised she’s toeing the line for the SAG strike, but not WGA.

What she's doing doesn't actually violate any of the contracts. Any covered work remains struck: no WGA writing for her show, no SAG-AFTRA members promoting struck work. So in a sense, she's not violating any legally enforceable line on either contract.

Choosing to produce and air the show without WGA work is shitty. It's crossing a picket line, and it's breaking solidarity. But it isn't a breach of contract, so she's following the strict letter of the law.