That looks good asf. 🍕
我呸！呸！呸！呸！呸！ 胡说八道 …… 忘了我吧。 🔸 可道！非常道！我自求我道。☯️ 🔸 Just re-post good stuff 🔁 + ramble to the æther 🤷♂️. Like technology 💻 + photography 📷 + aviation ✈️. Extremely online 🤡. 🔸 Don't discuss politik with me. 🤫 我不知道也不关心。（但我却是江泽民的铁杆粉丝。）
28 followers 40 following 517 posts
view profile on Bluesky
That looks good asf. 🍕
It sucks. But, also, everyone is doing it.
truly incomprehensible
Let them smoke.
i saw someone point out that the selector for time on iphone alarms is actually just a really long list and not truly circular. everything i’ve known is a lie
It's wild all the people that are constantly and performatively being mean about AI. Constant dismissive memes. Constant "Fuck you" replies. Constant antagonism. What a boring and corrosive way to live. And it's completely ineffective slop. It's self-congratulation masquerading as praxis.
Evangelion: 1.11 You Are (Not) Alone
Nostr is like the XMPP of decentralized networking.
But, anyway, my argument was that: just like how Mastodon kinda took over their own swath of Fedi & have a very diff. culture & user-base than rest of Fedi + all use Mastodon as a server/client……we could do the same on Nostr. Mastodon is software + distinct culture/user-base on shared protocol (AP).
There are a few clients that have their own local database which cache the entire known network you've loaded for offline-first engagement w/ engagement upload whenever back online, so very much SSB-like w/ offline capabilities.
&, my problem w/ Nostr clients aren't the lack of features (almost all are feature-rich), it's just that they're all kinda janky & clunky &, if anything, *too* feature-rich & messy. I'd love a cleaner, simpler, more reliable client like Bsky or some of the Fedi clients I use. It's for sure doable.
Yes. That's Relay List Metadata event (kind 10002). The standardized way in Nostr to broadcast your current relay set. Followers' clients are supposed to fetch your latest 10002 event & update their subscriptions to you accordingly. Think of it as the DNS records for your Nostr identity.
Nostr is great; It's kinda "ideal" in many respects—closest thing I've seen protocol-wise to everything everyone talks abt wanting/needing. Just need to carve out a good culture + community on it & develop some nice clients for it. Everything else is there ready to go. Heavy lifting is already done.
I've had discussions w/ ppl on Fedi abt building an intention community on Nostr distinct from the extant culture/user-base w/ our own relays. Additionally, a not-shit Nostr client we could use that utilizes specific non-typically-Nostr relays ootb for better onboarding experience.
Essentially cached by the relays. Identities are otherwise entirely ephemeral. If you want something analogous to a PDS or Fediverse instance you can just host our your own relay for yourself + your friends & set your own usage & data retention policies.
Mitra dev is working on implementing full portable objects and nomadic identities into the AP standard. But, ultimately, Nostr still stands as the dominant protocol architecturally speaking.
A picture is worth a thousand words: https://arewedecentralizedyet.online/ Here's another nice website: https://burningtree.github.io/bsky-dominance/ I am happy that I don't need to waste my time anymore debunking the "decentralization" of Bluesky. @ricci
📷️Nikon Zf 📸NIKKOR Z 50mm F1.8 S #photography #fotografía #Nikon #Zf
Will be interesting to see if Putin-controlled Trump destabilizes or invades Russian-ally Venezuela as the US did Russian-ally Syria. What is Putin's end-game in manipulating the US administration into upsetting Russian interests abroad? Some kind of 5d chess? How devious!
FUCK
Patlabor 2: The Movie
Summer scenery with the Eizan train in Kyoto🚃 The Eizan Railway runs through the northern part of Kyoto City. #photography #train #countryside #kyoto #japan #summer #landscape
I'm convinced, by many vectors, that Nostr is the closest thing to a silver bullet we have.
These age verification laws are in effect a lock-in that means we only can use Meta products, or from companies that will follow their economic model, and the idea of an open source open protocol solution is impossible. It’s all about combining state and corporate power.
Imho Nostr = the best protocol design-wise. It's simple + coherent + extensible. There're currently locally-hosted relay apps available for Android & clients that retain their own db for offline support. It's basically some of the better aspects of SSB hybridized w/ relayed signed event propagation.
Nostr is def. the right direction; It's the best designed protocol of the lot (vs. AT Proto, ActivityPub) w/ the greatest amt of potential. It's prob. the protocol that I + others ought be focusing on. Steeper learning curve than alts., but that's bc of Bsky's centralization + AP's client support.
Definitely already happening.
As the internet becomes less free, I expect forms of "dark webs" to become far larger.
After dusk in Kyoto today✨ #night #dusk #kyoto #japan #photography #sunset #photographersofbluesky
autumn vibes.
Every anti-AI hot take is like "Wow, this is incredibly sad. Truly pathetic. This is where civilization ends. Dark times ahead." And then it is just a screenshot of someone posting a silly, casual 10-second experiment they did with ChatGPT before presumably moving on with their life.
man what are you talking about
Yeah, I was going to say, I have seen equal but opposite sentiments expressed often on places like the Fediverse. It's kinda funny & new to me to see support for IPv6 bc I'm usually only seeing nothing but hate towards it in favor of IPv4.
I've finished reading RFC 9413: Maintaining Robust Protocols. It is a very important document, and I strongly recommend it to anyone developing ActivityPub applications. The robustness principle, also known as Postel's law, is simple and seemingly non-controversial: _be conservative in what you […]
Patlabor: The New Files Episode 13 The Dungeon Again
the idea of a 'parental control' makes me want to vomit uncontrollably
I don't think that really changes anything.
Rly interested in fep-ef61 & its integration into the Fediverse. It would solve so many things re the ephemeral state of the Fediverse + provide a sort of "best of all worlds" approach to protocols like ActivityPub, Nostr, & AT Protocol.
Should have included this link too: codeberg.org/fediverse/fe...
AP also has building nomadic/migratory identities: codeberg.org/silverpill/f...
Think ATP has a lot of superior protocol choices from tech standpoint, but AP has better de facto application + deployment: more coherent, easier to integrate, & is a lot more community-centric. Functionally more aligned w/ human nature, so to speak. AP "limitations" solve a lot of ATP problems.
I couldn't wait & ended up taking the Mac pill. M-series ARM SoC battery life is GOATed. Have high hopes for Intel x86 development in subsequent Lunar Lake iterations, though.
You think it's good that they're scanning conversations & doxxing people to law enforcement?
In Japan, they live by the exotic, untranslatable concept of "meiwaku", which means "dont cause problems for other people".
There shouldn't be bans on art no matter what, really.
Assuming the network truly decentralizes, network schism/fracture is inevitable. Will see the same thing that happened to the Fediverse prob. happen here. Only worse insofar as +90% of users will most likely remain on Bsky.
Just going to make self-hosting & local models more important. Rly, self-hosted everything. Reminds me of Apple when they intended to scan everyone's iCloud photos ostensibly to recognize & report CSEM. All these things disingenuously done to "protect children".
All these morons demanding safety controls + "guard rails" on this tech are to blame. Congrats, everyone: now you get zero privacy & your messages get scanned, indexed, & reported to law enforcement.
Safetyism and the populist fake “corporate accountability” everyone pretends is leftist has exactly one direction: fascism. futurism.com/openai-scann...
Nah.
没事，朋友
OP specifically asserting that Russia "hit" the EU delegation & outright suggests that it was an intentional act. The EU delegation wasn't hit. Their building wasn't hit. The building their offices were located in experienced shockwave damage from a nearby explosion.
They keep trying to get rid of him.
who the fuck is Headphone Jack
OP is upset that she has to train basic skills to someone who makes 3x her income. You're "sick" of…training younger ppl who have their careers ahead of them bc you're…what? Jealous of their yet-unrealized possible long-term potential? Bc they're men? I don't rly understand your post, here.
Nothing quite like quiet mornings to myself w/ a hot cup of coffee + some soft music to listen to.
Tomorrow I get to train someone who makes 3x my income on how to computer Ahh yes, the millennial dream
+ one way to "worsen the horror" too is to not seek to address any underlying causes behind attacks on the public & instead focus on blaming the tools, weapons, or components used to actualize such attacks.
"yeah, if we could just turn the country into a mass-surveillance police state w/ our ideological & sociopolitical views strictly enforced by law then liberal democracy would function great."
Unfortunately, dead non-EU-delegation-civilians wouldn't make any real news nor headlines. Nobody would really bring it up as they might when the people whose safety is jeopardized aren't govt. officials.
Russia didn't hit the EU delegation. The EU delegation building was damaged by a shock-wave from a munition. Nobody from the EU mission was killed nor injured, based on every report that I've read. Thanks, though: your misinformation motivated me to do some morning reading!
• They wouldn't be turbofans if electrically powered. • I've personally seen some of these concept aircraft flying out in the wild & spoken w/ the crews testing them. The tech exists & there are aircraft being flown w/ it—but, tech is far from mature & a much longer way off than journalists imply.
Feel like these are the same ppl arguing against AI bc a mentally-ill kid circumvented its nominal "guard rails" & leveraged it to consult him on suicide. Something tells me that these ppl wouldn't pass "the breakfast question".
How to land a small helicopter on a ship's helipad. 🚁🚢 #helicopter
There is no causation between guns & mass-shooters, just as there is no causation between aircraft & hijackers flying aircraft into buildings. You don't stop terrorist attacks by restricting aircraft, just as you don't stop mass-shootings by restricting guns. Tools used != underlying motivations.
need people to stop making technology inspired by Star Trek (smooth, sleek, no buttons, all screen, voice activation, extremely minimal, wireless) and start making technology inspired by Star Wars (clunky buttons, indicator lights on everything, ugly, indestructible, fat cables that emit steam)
Windows machine: · sleep state issues, · ran hot, · would randomly start cooking itself w/ lid closed while supposedly asleep or hibernating, · no captive portal support for public WiFi. Running Linux fixed literally all of that & things "just worked".
FYI Yemen has the second-highest gun ownership rate on Earth, several active conflicts, and I am pretty sure they still don't have regular deliberate mass shootings of children. Children have been caught in the crossfire, but their last school shooting was 1997.
I find ATmosphere's self-hosted infrastructure components: · Not very easy to setup nor get running, · not well polished, · no great diversity in choices between components, · not guaranteed to be supported by Bsky. &, this + nature of the network adds little incentive to self-host.
I wish I was married.
Shoot down an aircraft in international airspace?
Patlabor 2: The Movie
If I want a LLM to coach + support me through suicide, h/e, I don't want it to tell me "no" nor give me constant push-back. I want to be able to delete / dir on my machine if I want to. I don't want computers or programs dictating what I can or cannot do or putting up a fight about it. I'm the user.
Idt OpenAI has (nor should have) any liability here. The dude "misused" the product. He went out of his way to circumvent otherwise p. stubborn "guard rails" & content use-case policies. Ootb ChatGPT will not coach you nor support you through suicide. You may get it to w/ great time + effort.
I don't think it's faulty. Again, I can make ChatGPT say almost anything. You can circumvent & break down its "guard rails" w/ time + prompts. Personally, Idt it SHOULD even have guard rails. If there was the option for a "jail broken" LLM sans "guard rails" I'd choose to use that one.
Okay? I can make ChatGPT say almost anything I want it to say. Ultimately, it's not real & it has no locus of its own. It's a soulless tool. It lacks malice. This = tantamount to letting go of the wheel of an automobile while speeding down the highway & then blaming the car for the subsequent wreck.
Primarily Apple account is an @outlook.com email. Opened an @icloud.com email w/ the desire of making it my primary Apple account login. But, it just set the @icloud.com email as an "alias" of the @outlook.com email. I can't set the @icloud.com email as primary nor delete it from being associated. 😑
What did I do to you? 🥴
Very sane response.
People aren't allowed to like something?
Two things can be true at once.
I don't understand how Americans (whose entire govt. are basically blood-thirsty, terroristic jingoists) get the courage to show up literally anywhere anything positive is said about China & levy criticism + condemnation. Americans: you are THE evil empire! You have no place to utter any opinions.
"China has not fired a single bullet in international warfare in over 40 years, while during this time the US has invaded, bombed or carried out regime-change operations in over a dozen Global South countries. If there is any state that poses a known threat to world peace and security, it is the US"
AIs are kind of good at doing the things C-level officers do because they generate milquetoast statements based on no information, make up crap when they have to, and sound almost enthusiastic about almost caring for the things you care about.
That's a huge shame. Russia has made massive contributions to space exploration & manned space flight. They've been a big help in the recent past with both manned & cargo launches to/from ISS, e.g.. Any loss of mankind's space-faring capabilities is a blow. Can also represent threats to safety.
Kinda pointless & performative endeavor when my browser already (& always has) blocked ads, tracking, cookies, etc.. I still reject the cookies & tracking anyway, whether they happen or can happen or not.
The 'click to reject cookies' business has not improved my experience of using the Internet even slightly. Just another pop up to get rid of alongside newsletter sign-ups and notification requests.
One can accurately place causation on mental illness while also arguing for greater access + thoroughness of mental healthcare. Actually, Idk anyone who wants less healthcare availability/affordability/accessibility. Everyone wants robust, widespread, accessible, & affordable (free?) healthcare!
It didn't land on a target or platform, I don't think. It intentionally "landed" at a predefined location in the Indian Ocean & blew up after collapsing into the waves after "touch down".
That's a lot of paper. Bro clearly wasn't touching grass.
I have given up entirely on the dream of owning a home or owning land or anything like that. It's just not going to happen, & therefore there's no reason to even care or worry or otherwise be disappointed by it. C'est la vie~
Behind every company doing layoffs and running teams ragged is a series of expensive enterprise software subscriptions nobody uses that will get renewed each year without question.
social media is awesome you get to slowly get to watch people go completely insane and become worse
Mostly, I see people who already don't like LLMs existing whatsoever, or don't like OpenAI specifically as a company (for their own w/e reasons), disingenuously grasping onto incidents like this as ammunition to further the grievances they already had.
I don't really think there is a problem. I don't think OpenAI is expressly to blame for any of this. I don't think there is any issue with LLMs either in this regard. If you're mentally ill & gonna' sui that's not the fault of LLMs as a technology.
Good thread: https://mastodon.social/@Gargron/115074631868019158 @Gargron was dogpiled by Bluesky shills but he's right about everything. Imagine claiming that you're a decentralized network and then adding a geoblock because of some draconian "law" in some random country.
N2P (like A21/M7) will be super comfy.
Honestly, there's nothing I hate more than a computer/program telling me "no". That's not how this works.
(these same ppl will call you a "fascist", btw)
Progressive, liberal democracy-enjoyers on Bsky: "yeah if we could just liquidate 40% of the country's electorate then everything would be great & the system would work perfectly"