Hacker Newsnew | past | comments | ask | show | jobs | submit | davidgerard's commentslogin

NEWS UPDATE: Kent's AI girlfriend has had a realisation https://social.vlhl.dev/notice/B3gZZsCZxDtZJ88XxI

transcript https://paste.xinu.at/6atmCN

> We finally got the future where people will write sad breakup country songs about their tractor leaving them instead of sad breakup country songs about their wife leaving them.


Which specific claims are "outdated"?

Your answers have all been that it'll be solved in the fabulous future. Quite possibly! But until it happens, it hasn't happened.


the work is currently happening, which I have mentioned in other comments, since you indicated you have read them


At this point it sounds like a Duke Nukem Forever thing and it will be funcional around 2038, and be totally obsolete at this point.


There have been some people trying to think about DID migration, but also core protocol designers saying that's problematic and probably not going to happen. We'll see, Bsky CTO has said real politik has its place, so rally anything can be on the table if you can convince the majority. I think we could have something equivalent to a "hard fork" should the need arise


it was "currently happening" in May 2023, but I look forward to the working examples!

also, that's not a list of which specific claims are "outdated"


An official read-only mirror was released last week

> also, that's not a

you didn't respond to some of my points and questions, I'm leaving it as an exercise for the reader at this point


It's very much a Not Invented Here of Mastodon and the Fediverse.

Bluesky is a good user experience insofar as it's centralised.

Mastodon is a bad user experience insofar as you're forced to be aware of the decentralisation.

If you want successful decentralisation, Mastodon has that out of the box. You can stand up a Mastodon, Akkoma, GotoSocial etc on a $5/mo VM and you're an equal participant immediately. Or you can join someone else's server.

ActivityPub is underspecified and Mastodon just ignored a lot of it and so the actual protocol is an unholy mishmash of the two. It mostly works though, by the process of people beating on it until it works.

With Bluesky, you have a centralised service and a lot of people saying "decentralised!"

AT Proto is theoretically decentralised in the fabulous future and points of absolute and financial centralisation keep turning up.

I spend all day posting to both, fwiw. They each do a particular job. But the "decentralisation" in Bluesky is fake. Or at best, simply not feasiblly true.


Last I looked its how to sign on page said "first create a keypair". That's certainly a good way to avoid the problems of success ...


Bluesky growth spurts are always when Musk or (less frequently) Zuckerberg step on their dicks again and more people come over from their services. In between are slow declines.


A relay or appview needs a ton of resources. Blacksky finally created the second ever real-world usable appview instance after 2.5 years.

Also, the open source version of the appview doesn't work at Bluesky scale. You need a proprietary database for sufficient speed.

AT Proto is completely decentralised, except for all the structural and financial points of absolute centralisation.


The relay is not that bad, the only really bad part is building an index, and most apps on the atmosphere have no need to index bluesky records, so the economics for them look very different.

The work towards permissioned data and group-shared data will make it so apps can choose their own levels of "decentralization" of "federation" on atproto primitives. For example, two diametric options

1. An app that is not open source code, but still does all the same atproto credible exit stuff. Naturally leans into winner-take-all

2. An app that is tied to community, think something like Discord, where most servers don't care about what other servers are doing. Each community could run their own version and only care about their data. This is raspberry pi hostable.


As I said below, there's hands on and there's hands on. The operator is unambiguously the operator, even as they automate to some degree.


I'm getting two genuinely held narratives here

1. Some people call Openclaw an RCE in a tin: "<RCE in a tin>, operator pulls pin, tin goes boom"

2. Some have a very different framing: "There's no such thing as the RCE in a tin, it's all fake news, therefore it couldn't go boom"

Well either way people need to be more responsible, both sides agree to that much. But the safety implications of boom vs not boom might be a problem for a while.

The really uncomfortable position is : "This thing has real but unreliable capabilities, which is exactly the most dangerous configuration."

On the upside, once you've Named The Problem, there's several ways forward from there, at least.


I do not believe the third position is supportable, no. This incident does not supply evidence.


Yeah, I get it, that's plausibly outside your personal Overton window; we all come at this from different angles. I take it you reject the operator's confession out of hand then?

That said, the theory you're putting forward today seems to be a bit flimsy and convoluted for a mere $209 in tokens. Were they ever even spent? I don't think it should survive Occam's Razor.

If the meme-coin you mentioned enters the timeline before 2026-Feb-08, plausibly that resurrects the theory.

But who knows, I'm not the crypto-hunter. Maybe you've seen crazier stuff!


https://basescan.org/token/0x3ed514d115c6ec0f1b3d48dfcb3878f...

First timestamp: 2026-02-13 01:05:59 UTC

So that means it stays implausible for me as yet, from my personal perspective.


author here - there's hands on and there's hands on, a Waymo operator is mostly getting the robot out of stuff beyond its if-then loop but also they are in any reasonable sense the human operator.

Like, I'm sure our bro automated as much as possible that wasn't more directly relevant to the crypto bit.

The original PR is innocuous at a glance, but then an innocuous PR is how Jia Tan started.


author here, the video and post content are basically the same

full credit to Ariadne Conill for finding the crypto angle, that makes sense of the whole thing


except she didn't find anything about 'bro', just that the bot is using crypto?


you should probably read down the thread, and also find out who Ariadne Conill is


I can't see the crypto token, but everything about this reeks of someone will announce a token shortly.

EDIT: oh there it is


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: