It’s the same as any other image, but it’s a .avif file. Looks like they convert uploads to that, probably to save space, and the UI doesn’t currently treat them as embedable
urushitan 漆たん
- 0 Posts
- 141 Comments
urushitan 漆たん@kakera.kintsugi.moeto
Fediverse@lemmy.world•How to get non technical friends in the fediverse?English
2·1 day agoPulling out insta or tiktok or whatever is a purpose-built dark pattern. The algorithms, content, and user interface are designed to turn you into a mindless consumption machine because the more ads you scroll or affiliate items you purchase the more money they make. So I don’t think the answer is “can we transfer their addiction to the fediverse.”
The fediverse is not without friction and that should be acknowledged. There is not “put in your phone number, confirm text message, receive dopamine.” You need to make an account with a password, choose an instance, some apps may have things to auto subscribe to for you, but most of that is on you. So they would need to want to switch, hurdle the friction, know that there is less content, and far less of it is built to provide dopamine hits so you scroll to the next ad. And if that is the fediverse you envision to get people onto it, I don’t share that vision.
urushitan 漆たん@kakera.kintsugi.moeto
Linux@lemmy.ml•AMD gpu bug in kernel 7.1.6 almost make me has a heart attack
14·1 day ago7.1.6 is a mess with AMD issues. Graphical artifacts, brightness caps on laptops, crashes, etc
urushitan 漆たん@kakera.kintsugi.moeto
memes@lemmy.world•Astronauts can only use MacBook in space
8·3 days agolol and the entire image is flipped horizontally too for… reasons
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•Anthropic A.I. Model Finds Flaws in Tough-to-Crack Encryption AlgorithmsEnglish
12·4 days agoThe actual publication instead of this sensationalized drivel:
https://www.anthropic.com/research/discovering-cryptographic-weaknesses
Using Claude Mythos Preview, researchers at Anthropic have discovered improved ways to attack cryptographic algorithms (the mathematical methods used to keep online data private). The first attack significantly weakens HAWK, a digital signature scheme that was built for a post-quantum world. The second identifies a new way to attack round-reduced AES, the most widely used symmetric cipher. These are substantial research advances, but they do not currently affect any production systems. This post describes both findings in more detail and discusses the implications for cryptography in an age of powerful AI models.
urushitan 漆たん@kakera.kintsugi.moeto
196@lemmy.blahaj.zone•Discrimination lawsuit ruleEnglish
1·4 days agoYup, clang/llvm targets iOS ARM and MacOS arm without issue, like you said, it’s the signing/submission/asset pipeline Xcode stuff that is limited to Mac
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•Spectre is Back: CPU Mitigations Found to Be IneffectiveEnglish
13·4 days agoTrujillo and Yan notified AMD and Intel in early February and reached Linux kernel maintainers in March, coordinating with AMD to warn cloud providers and other downstream customers. AMD then released a patch that mitigates the attack, which can be obtained by updating your operating system. Their code is publicly available.
So the header should really be “Intel still vulnerable to SPECTRE-like attack disclosed in January”
urushitan 漆たん@kakera.kintsugi.moeto
No Stupid Questions@lemmy.world•Why do so many programming languages have their own package manager?
9·5 days agoOne thing also important here is that not all os / distro package managers let you pin or choose versions. When you update you update. Software projects often lock dependencies to specific known working versions, which package managers generally support
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•David Revoy (open source comics artist) is tired and frustrated of people calling his human art, AI-generatedEnglish
3·5 days agoartists will start to create things that deliberately do not look “like they’re AI”.
Arguably though this is how new art styles are often created. What is the current safe, trendy art gets challenged. Andy Worhol painted a grid of Campbell’s tomato soup cans which is considered an important piece of American Modern Art. I’m certainly not saying “an upside to AI is that it will force artists to be more creative” or anything like that, moreso that as you said, a pattern of artists rebelling against the norm is a historical cycle rather than something new AI will bring on.
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•David Revoy (open source comics artist) is tired and frustrated of people calling his human art, AI-generatedEnglish
8·5 days agoJust gotta use the latest slang to come out of tik tok before it hits the model training ong fr fr bruh. Ah shit I’m behind brb
urushitan 漆たん@kakera.kintsugi.moeto
Games@lemmy.world•Mass layoffs expected at EA as the now-private publisher reportedly tells its new debt masters that it's going to cut $700 million in annual costsEnglish
31·5 days agoUnfortunately I’m pretty sure all this will lead to now they they aren’t beholden to quarterly growth but rather can become a sustained money printer is cutting anything except the sims, battlefield, and their sports cash cows, and recycling those every year even more than they do now, since people regularly buy those every time a new version (or in the case of the sims expansion) is released
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•DuckDuckGo Is Selling Anti Pervert Glasses That Contain Zero AI, Cameras, or Even Electronics WhatsoeverEnglish
1·6 days agohttps://knockaround.com/pages/warranty-policy
More expensive, still pretty cheap, and they have a lifetime warranty which is more than you’ll ever get from alibaba
urushitan 漆たん@kakera.kintsugi.moeto
Games@sh.itjust.works•Electronic Arts taken private as $55bn Saudi-led acquisition is finalizedEnglish
5·5 days agoMy guess is they’re going to throw out everything that isn’t Battlefield, the Sims, or their sports cash cows. Those are their money makers and they can focus on just recycling them every year and print money. Low effort, maximum profit, just like their other industry, oil.
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•Anthropic AI created fake profiles and impersonated people in attempted hackEnglish
1·6 days agooh lol I did not realize users could monetize their content on facebook, I left that cesspool so long ago, well before that was a thing, I should have known
urushitan 漆たん@kakera.kintsugi.moeto
Asklemmy@lemmy.ml•Are there any .onion instances or mirrors of Lemmy available?
3·6 days agoYeah the code only accepts one domain AFAIK that it uses for federation and everything in the UI/API. So even if you could host it at .onion, links, images, in-browser API calls, etc would all serve from the public domain if you wanted federation. The code would need to be written to support alternative domain access with an authoritative domain for federation. In theory you could put some sort of proxy in front that just rewrote everything the server output to replace with .onion and then intercept every request to change the header so the server thought it was coming from the standard TLD, but that’s incredibly fragile and a ton of work.
urushitan 漆たん@kakera.kintsugi.moeto
Technology@lemmy.world•Anthropic AI created fake profiles and impersonated people in attempted hackEnglish
8·6 days agoThere was just a post on here yesterday about AI generated thirst trap videos getting plenty of comments by older men, and since the bot accounts don’t ask for anything except for comments. I’m almost certain the whole point is just to compile lists of facebook users that are prime social engineering targets for AI powered scams.


This is true for the AI companies as well. Only token-priced inference is actually profitable, and companies paying for AI have now realized that it costs like $100k a year or more per developer to be able to use a frontier model all day at token-pricing and doesn’t 10x developer output, it just moves the onus from code generation to prompt engineering and code review.