Kissaki

@Kissaki@beehaw.org

This profile is from a federated server and may be incomplete. Browse more on the original instance.

Kissaki,

How do you want us to push for peace there too? Because we have been since the beginning of the war in my eyes.

What do you mean by “won’t recover from”? Because they have lost things that can’t be recovered since the beginning of the war. Russia is losing things they can’t recover too; thousands of its people for example, it’s money reserves, its military inventory, its non-military-sector economy. Where do you draw the line for Russia and Ukraine of what is “won’t recover from”? Western nations have already committed to helping rebuild the country and especially its destroyed infrastructure.

How is the war in Ukraine “quickly turning into a much bigger global conflict”? Fighting is still only within Ukraine and the border to Russia. Western material support has been the case since the beginning.

I have to assume by pushing for peace you mean Ukraine should accept losing large parts of its territory and human atrocities in order for the fighting to end. Is letting Russia win going to reduce conflict long term though? They’ll have more resources to invade other countries next. And proof that it’s a worth investment. That works and they win from. There was precedent before the current war in Ukraine, which is why they started this invasion in the first place. Only this time it didn’t go as smoothly.

Kissaki,

and noted that the planes won’t be delivered to Israel for years

…but will it get better or worse in those years?

Kissaki,

I think you can learn it when you’re younger than 20 years old

Kissaki,

Seems fairly obvious and to be expected that they would do that, given their intelligence and environment and utility use.

Still, important and significant to observe and prove in the wild. Especially as not one-off random anecdotal observations.

Kissaki,

What are you referring to as obviously fake? The article about that bot account?

Kissaki,

The only way to meaningfully advocate for it after your company already announced their conditions and offerings is to present value gain.

What do you suggest concretely? What should be offered under what conditions? What would that mean as cost? What would the benefit be? How substantial is it?

Reaching out privately to them is certainly going beyond what you are employed for. I don’t know about ill-advised - if you never disclose it or are at least mindful of that. But it’s a personal assessment. You seem to be willing to invest a lot into a single customer, who tries to do something not offered or considered by the company. Whether it’s personal interest, or first a broader better understanding of the use case, I can see how it could be worth or worthwhile. But I wouldn’t get my hopes up about changing the opinions of your company [from their information alone].

Your company offered API access. So there is an interface available. They won’t make it free unless they see and deem it worth it to do so.

Kissaki,

What are those options?

Domain and IP block lists?

Kissaki,

I don’t think it seems like too few samples for it to work.

What they train for is rather specific. To identify anger and hostility characteristics, and adjust pitch and inflection.

Dunno if you meant it like that when you said “training people’s voices”, but they’re not replicating voices or interpreting meaning.

learned to recognize and modify the vocal characteristics associated with anger and hostility. When a customer speaks to a call center operator, the model processes the incoming audio and adjusts the pitch and inflection of the customer’s voice to make it sound calmer and less threatening.

Kissaki,

So what you’re saying is that we need AI do interface in-store as well? /s

Kissaki,

Does it apply if you don’t say that you are posting under the license? It may be implied, the intent is reasonably clear, but an argument of ambiguity can be made. You’re merely linking to a license.

Does it apply if the link label mismatches the license? CC by-nc-sa does more than deny commercial AI training. It requires attribution, requires general non-commercial use, and requires share-alike.

Personally, I prefer when it’s at least differently formatted to indicate it as a footer and not comment content. I’ve seen them smaller and IIRC italic on other commenters, which seems more appropriate and less distracting and noisy [for human consumption]. When the comment is no longer than the license footer… well…

Kissaki,

There was no reason they had to implement it with scan though, was there?

Kissaki,

PC. Most often mouse or keyboard and mouse. Sometimes gamepad, then maybe streaming to my TV and sofa. I have a SteamDeck, but it’s not getting much use. Like my Vive.

Kissaki, (edited )

In Advanced Users they link to computefreely.org for

approachable and friendly for people curious about free and open source operating systems and Linux distributions

which is definitely not the content there. Looks like the original website went defunct in 2022 between January and May. Their website repo is archived.

/edit: I saw the website is open source and created a change request.

Anyway, I’d be careful about how up-to-date this website is, and what it links to.

'LLM-free' is the new '100% organic' - Creators Are Fighting AI Anxiety With an ‘LLM-Free’ Movement (www.theatlantic.com)

As soon as Apple announced its plans to inject generative AI into the iPhone, it was as good as official: The technology is now all but unavoidable. Large language models will soon lurk on most of the world’s smartphones, generating images and text in messaging and email apps. AI has already colonized web search, appearing in...

Kissaki,

the $80 billion start-up

lol, can you still call that start-up?

Kissaki,

What is dissenter

Kissaki,

Mozilla Foundation is a non-profit. Mozilla Corporation is a subsidiary of Mozilla Foundation.

You’re claiming those forms are the same as any stock market company?

Kissaki,

A title, a teaser text, two paragraphs, and only in the one after inside a quote can I read that it’s limited blocking within Russia, not a general removal or blockage.

Kissaki,

Right now it’s entirely timestamp-based. That means it can interface and work on simple playback terms. On time, jump, jump to time, etc. Having to get frame data and hash it, and make playback depend on it adds a lot of technical complexity.

If ad length varies you don’t even know how far to jump ahead. And if you haven’t prebuffered the data until after the ad, you can’t find out from a hashed after-frame-hash-value either.

Xbox Boss Phil Spencer Addresses Studio Closures (www.ign.com)

“The closure of any team is hard obviously on the individuals there, hard on the team,” Spencer said. "I haven’t been talking publicly about this, because right now is the time for us to focus on the team and the individuals. It’s obviously a decision that’s very hard on them, and I want to make sure through severance...

Kissaki,

ensures we can continue to do shows like the one we just did.

Doing shows, not creating games /s(?)

Kissaki,

Are they gonna pay them out? /s

Every employee could create their own companies and products. Opportunities and diversification!

Alternatively, each and every one of them can make 101 millionaires and still be a millionaire themselves!

Kissaki,

Their description reads

I made a moving digital card of Blue Archive Mika using two flexible OLED displays.

As a bonus, I made it possible to switch between the ones I made in the past. Prototype

One commenter says you can buy these kinds of displays on Amazon. Another responds AliExpress. Another reply says

650 USD for just one display. Did I read that right?

So they’re not exactly inexpensive. Which is, of course, not unexpected, given it’s a new kind of product, relatively recent tech.

Kissaki,

Only 250 €? That’s not expensive/overpriced for the product. And far from the 700$ mentioned by a comment.

When I looked on DE Amazon I didn’t find any.

Kissaki,

How did you come to the interpretation that protection of anonymity were violating copyright?

How do you mean anonymity could give better control over copyrightable information?

Kissaki,

From the article-linked ruling press release - what it means in practice, what this was about:

In order to protect works covered by copyright or related rights against offences committed on the internet, a French decree introduced two personal data processing operations. The first operation consists of the collection, by rightholder organisations, of IP addresses which appear to have been used on peer-to-peer websites to commit such offences and the referral of those IP addresses to the Haute Autorité pour la diffusion des œuvres et la protection des droits sur internet (High Authority for the dissemination of works and the protection of rights on the Internet) (Hadopi) 1. The second operation, carried out by the internet access providers at Hadopi’s request, consists, inter alia, of matching the IP address with the civil identity data of its holder. Those data processing operations enable Hadopi to initiate a procedure against the persons identified, combining educational and punitive measures, which may lead to a referral to the public prosecution service in the most serious cases.

I find the ruling press release is much more understandable (and much more informative) than the OP-linked article.

Kissaki, (edited )

If you read more of the ruling, the ruling allows EU nations to impose requirements on ISPs. So the storage duration would be up to national law. (Which of course one may call into question bring before court on whether they are too long.)


The question of whether you are liable as a provider of open access is an independent question. Yes, it becomes more relevant if you as a provider can’t bet on anonymity anymore. But it’s independent.

Looking at DE Wikipedia, looks like previous EU court rulings were dismissing being held accountable, but there’s still one open. German law freed it in 2018. No mention of EU specifically in this article, so maybe it’s national concern - at least until the EU court makes a ruling.

Kissaki,

I only purchased this toothbrush from Amazon because that was the only way to get the water-resistant Alexa speaker that I wanted for the bathroom.

Kissaki,

They’re starting this as an experiment in their PDF editor, yes. They then want to extend to PDF reading, and then hope to extend to the general web browsing.

will be available as part of Firefox’s built-in PDF editor

Firefox is able to add an image in a PDF using our popular open source pdf.js library[…] Starting in Firefox 130, we will automatically generate an alt text and let the user validate it. So every time an image is added, […]

In the future, we want to be able to provide an alt text for any existing image in PDFs, except images which just contain text (it’s usually the case for PDFs containing scanned books).

Once the alt text feature in PDF.js has matured and proven to work well, we hope to make the feature available in general browsing for users with screen readers.

Kissaki,

Where did you read this? The article says the opposite.

will be available as part of Firefox’s built-in PDF editor

Firefox is able to add an image in a PDF using our popular open source pdf.js library[…] Starting in Firefox 130, we will automatically generate an alt text and let the user validate it.

See also my other quotes in this comment.

will be available as part of Firefox’s built-in PDF editor

Kissaki, (edited )

From your OP description:

EDIT: the AI creates an initial description, which then receives crowdsourced additional context per-image to improve generated output. look for the “Example Output” heading in the article.

That’s wrong. There is nothing crowd sourced. What you read in the article is that when you add an image in the PDF editor it can generate an alt text for the image, and you as a user validate and confirm it. That’s still local PDF editing though.

The caching part is about the model dataset, which is static.

Kissaki, (edited )

So, planned experimentation and availabiltiy

  1. PDF editor when adding an image in Firefox 130
  2. PDF reading
  3. [hopefully] general web browsing

Sounds like a good plan.


Once quantized, these models can be under 200MB on disk, and run in a couple of seconds on a laptop – a big reduction compared to the gigabytes and resources an LLM requires.

While a reasonable size for Laptop and desktop, the couple of seconds time could still be a bit of a hindrance. Nevertheless, a significant unblock for blind/text users.

I wonder what it would mean for mobile. If it’s an optional accessibility feature, and with today’s smartphones storage space I think it can work well though.


Running inference locally with small models offers many advantages:

They list 5 positives about using local models. On a blog targeting developers, I would wish if not expect them to list the downsides and weighing of the two sides too. As it is, it’s promotional material, not honest, open, fully informing descriptions.

While they go into technical details about the architecture and technical implementation, I think the negatives are noteworthy, and the weighing could be insightful for readers.


So every time an image is added, we get an array of pixels we pass to the ML engine

An array of pixels doesn’t make sense to me. Images can have different widths, so linear data with varying sectioning content would be awful for training.

I have to assume this was a technical simplification or unintended wording mistake for the article.

Kissaki, (edited )

MDN Which makes sense to me.

figure does not invalidate or change how img is to be used. The caption may often not but can differ from the image description. If alt describes the image, figcaption captions it.

What the fuck is Lemmy doing, breaking with HTML in code formatting?? Man it’s completely broken. I committed sth so it doesn’t remove the img lol.


<span style="color:#323232;"><</span><span style="color:#63a35c;">figure</span><span style="color:#323232;">>
</span><span style="color:#323232;">  img src="party.jpg" alt="people partying" />
</span><span style="color:#323232;">  <</span><span style="color:#63a35c;">figcaption</span><span style="color:#323232;">>Me and my mates</</span><span style="color:#63a35c;">figcaption</span><span style="color:#323232;">>
</span><span style="color:#323232;"></</span><span style="color:#63a35c;">figure</span><span style="color:#323232;">>
</span>
Kissaki,

Looking at Wikipedia on arrays, I think I’m just not used to array as terminology for multi-dimensional data structures. TIL

Kissaki,

Where is that quote from?

Kissaki, (edited )

I don’t see a link. Post content source is empty too.

screenshot 1

screenshot 2

Kissaki,

Given that it’s not in the comment source I doubt it’s a browser issue. But if you can see it… wtf

When I open the comment in your original instance context it’s there. Your comment was edited. Did you edit it in? I guess it got lost between instance communication lol.

Kissaki,

Interesting. It also made me look at the MDN docs again. img alt is consistent to that. I wasn’t aware of the empty for omittable images.

I also looked at figure again, and in my interpretation it does declare that figcaption is to be used.

figure represents self-contained content. figcaption provides the accessible name for the parent. The accessible name is is the text associated with an HTML element that provides users of assistive technology with a label for the element.

The resolution order being aria-labelledby, aria-label, input[type=button][value], input[type=image]|img|area[alt], …

So figcaption takes priority over img alt.

Kissaki,

oh god, would suck if it’s another broken Lemmy release

I had other formatting problems with HTML inside code blocks being removed and bleeding out of them generating other closing tags. Maybe that was also related.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • fightinggames
  • All magazines