@hrefna@hachyderm.io
@hrefna@hachyderm.io avatar

hrefna

@hrefna@hachyderm.io

SRE at Google. Queer. Poly :potion_polyamory: Trans :verified_trans: :nonbinary_potion: Engineer. Ace :flag_ace: Member of AWU-CWA. #ActuallyAutistic :rainbowinfinity: #UnionStrong

Opinions my own. Does not suffer fools gladly.

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

hrefna, to random
@hrefna@hachyderm.io avatar

If you are talking to someone who is not currently using:

  • A password manager
  • MFA on most or all accounts that allow it
  • An up to date operating system on all of their devices

Please stop yourself from recommending:

  • A VPN
  • Tor
  • Tails (or any variations/equivalents)

Just… rewind a bit and help with the first items first.

hrefna, to random
@hrefna@hachyderm.io avatar

for various surgeries:

hrefna, to random
@hrefna@hachyderm.io avatar

Three different problems:

  1. How do I get everyone who agrees with me into a room?

  2. How do I get everyone in a room to agree with me?

  3. How do I get everyone who doesn't agree with me out of the room?

I think a lot of people on the left confuse these.

(2) is the bread and butter of union organizers, but the look, feel, and compromises required are radically different from (1), and both of them are quite different from (3) except in trivial cases.

hrefna, to fediverse
@hrefna@hachyderm.io avatar

Do people seriously not understand how much gatekeeping there is in the statement "you need to understand the of the [before you should be here]"?

The Culture of the Fediverse™ is by design and necessity something that can and should evolve. If it is going to succeed it must evolve.

You can't say "everyone is welcome here" and then turn around and say "but only if you adjust to a set of cultural expectations that no one will explain to you or write down."

hrefna, to random
@hrefna@hachyderm.io avatar

Overall I'm kind of just of the view that blocking Meta's hypothetical server preemptively is just the most… performative, useless gesture

To annoy everyone who reads this for a moment, however:

It does encompass what many leftists (and liberals) seem to think of as "taking action." A nice moral test. Easy to pontificate about. Feels good. Let's you talk in grandiose terms. Easy for an individual to do. Largely ineffective both short and long term.

I swear we are our own worst enemies.

hrefna, to random
@hrefna@hachyderm.io avatar

"Choosing an instance isn't difficult, I think people are overstating that."

Okay, now find an instance that:

  • Isn't defederated from some significant portion of my friend group.

  • Isn't defederated from by the same.

  • Doesn't defederate from major news sources running on their own instances.

  • Has reasonable moderation.

  • Will actually be here tomorrow and not just fold.

  • Has reasonable uptime.

  • I can trust to not to keep an unencrypted copy of the database on a personal device.

hrefna, to random
@hrefna@hachyderm.io avatar

Regardless of ones precise position, I want every single person who is talking about the need for "opt-in, not opt-out" with a bridge to take a moment and reflect on how we handle replies here in mastodon-land.

Which is neither opt-in nor does it give you an opt-out.

I can name a wide variety of features like that here, and I'd love to see more energy directed at those issues.

We barely provide controls, or even guidance, around federation here and what we provide is not nuanced.

hrefna, to random
@hrefna@hachyderm.io avatar

To reiterate: my problem with consensus approaches is that they present a "false consensus" of the world.

Whatever biases are in the source lists? Are present in the output. Often amplified.

Whatever priorities are in the source lists? Are present in the output. Often amplified.

The idea is to remove biases, but that only works if biases in your source lists are normally distributed

Which they aren't. Thus: amplified

But this is obscured and it is presented as "more fair."

1/

hrefna, to fediverse
@hrefna@hachyderm.io avatar

I don't know who needs to hear this, but the does not need a dating app. Please please please do not create a dating app here at least until we fix some fairly substantive issues around trust and safety.

hrefna, to random
@hrefna@hachyderm.io avatar

"Death of the Author" never originally meant "ignore the problematic aspects of the author" or "forget about the author."

It meant that the author's opinion of their own work was not authoritative and, in fact, that there could be no "definitive" interpretation. It was about how each individual's interpretation could be considered authoritative in its own right

It wasn't supposed to be about "we can ignore the author being terrible or trying to push a narrative." It works at a different level

hrefna, to generativeAI
@hrefna@hachyderm.io avatar

Maybe, and hear me out, if you need to steal the work of others for your to exist, then your generative AI should not exist.

hrefna, to random
@hrefna@hachyderm.io avatar

I'm absolutely begging so-called leftists to stop spreading anti- propaganda without at least minimal fact checking.

Absolutely begging people.

If you want counterexamples of this they are literally everywhere.

We can start with my union:

https://www.alphabetworkersunion.org/our-people

or the union we're under, CWA:

https://cwa-union.org/about/leadership

I get that there are a long history of problems in unions, and also you don't make things better by spreading lies and half truths about them.

hrefna, to random
@hrefna@hachyderm.io avatar

Between and and let's also look at another threat model that I think people don't fully appreciate with

How much do you trust the blocklist source—not its upstreams, but the actual place you get it from—to do what they are telling you it does?

How much do you trust the maintainer to not perform a MitM attack?

How much do you trust others who have access?

If a MitM attack were performed, how would you know about it? How would you catch it? How quickly?

1/

hrefna, to random
@hrefna@hachyderm.io avatar

I love* how much preparedness starts with "have money."

I also love* how many people who give disaster advice completely ignore how much money they are actually talking about.

I really love* all of the people with no experience in disasters confidently talking about this as if it were something they had done a thousand times.

  • not actually
hrefna, to random
@hrefna@hachyderm.io avatar

"Small instances are the best, only small instances. Mid sized at most."

Also: and they should have sub-hour 24/7 oncall.

Also Also: and they need to have a sub-hour public response time.

Also also also: they need to hire a comms team

Also also also also: What do you mean those things cost money?

hrefna, to random
@hrefna@hachyderm.io avatar

Protip: Showing up in people's mentions who you don't follow and who don't follow you in order to give them "advice" when they have not asked for it is almost always hostile behavior and will be read as such.

Especially if your "advice" assumes they don't know what they are doing.

You aren't entitled to their time or attention and you shouldn't act surprised when you encounter hostility as a result.

hrefna, to random
@hrefna@hachyderm.io avatar

The SCOTUS is very likely to reject Trump's immunity bid, and I'm getting really tired of doomposting irrespective of if they do

If you are in despair sit with it for a moment. Your feelings are valid. But then the question is:

What next?

"I'm going to yell about it on mastodon" isn't going to solve anything. It might bring momentary catharsis, but only momentary

What next?

Eventually we all need to take meaningful action

Freezing is rarely a healthy response

So let's talk about it.

1/

hrefna, to random
@hrefna@hachyderm.io avatar

One of the things about scoring algorithms is that they are prone to bias from a lot of sources you don't think about unless you actually work in this space.

For example, with blocklists, let's say you have five trusted sources and you use a majority evaluation.

Your recommends $foo for a block.

Now you add two more trusted sources. Both are good, high quality sources, but they don't list $foo. Now $foo doesn't meet the inclusion criteria.

What does that mean?

1/

hrefna, (edited ) to fediverse
@hrefna@hachyderm.io avatar

If you are looking for a project to work on I highly, highly recommend checking out @jenniferplusplus's project: http://letterbookhq.com / https://github.com/Letterbook/Letterbook

It has all of the marks of things I want to see in the fediverse going forward.

hrefna, to mastodon
@hrefna@hachyderm.io avatar
hrefna, to fediverse
@hrefna@hachyderm.io avatar

Draft of my (to-be) FEP for audience targeting and filtering is up at https://github.com/larpconnect/featherpub/blob/main/feps/8f9c/index.md for those that are interested.

I haven't submitted as a FEP yet (and won't for a while), but that's the ultimate goal for it.

hrefna, to random
@hrefna@hachyderm.io avatar

Tired: "70% of bugs are caused by memory safety issues"

Wired: "70% of bugs are caused by scope creep"

Inspired: "70% of bugs are caused by overworking and underfunding developers."

hrefna, to random
@hrefna@hachyderm.io avatar

One of the core fundamental failures of consensus is the lack of definitive signal in the source list.

Instances get added to an instance's list because:

  • They are CSAM
  • They had a bad interaction with the moderator
  • Because there was a moderation disagreement
  • Because they are shutting down
  • Because they have a different policy toward CW or adult content
  • Because someone else blocked them who a moderator trusted
  • Because someone else blocked them and everyone's busy

1/

hrefna, to random
@hrefna@hachyderm.io avatar

No, googlers in general are not "worried that [we] are slowly being replaced by AI."

I haven't done a formal poll, but it would shock me if the number turned out to be significant.

That's not how this works, that's not how any of this works. What kind of clickbait nonsense is this.

hrefna,
@hrefna@hachyderm.io avatar

Layoffs are happening because our leadership is short-sighted.

Because everyone wants to invest in AI, and they feel like they need to turn over every rock to push money that direction.

Because of the capitalist dystopia and capitalists are going to capitalist.

Because googlers have been propagandized and mislead by the former into being anti-union.

Not because we're being "replaced by AI." That's not only not a thing, it is at the "not even wrong" level of not a thing.

hrefna,
@hrefna@hachyderm.io avatar

Layoffs are happening because upper management has a herd mentality across tech companies and because we've become a conventional company more focused on the next quarter than on the long term.

Layoffs are happening because our leadership team are all billionaires or near-billionaires who have no idea what it is like to be rank and file any longer.

Layoffs are happening because our union isn't strong enough to fight back.

Layoffs are NOT happening because we're being replaced with AI.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • JUstTest
  • GTA5RPClips
  • DreamBathrooms
  • InstantRegret
  • magazineikmin
  • khanakhh
  • Youngstown
  • ngwrru68w68
  • slotface
  • everett
  • rosin
  • thenastyranch
  • kavyap
  • tacticalgear
  • megavids
  • cisconetworking
  • normalnudes
  • osvaldo12
  • ethstaker
  • mdbf
  • modclub
  • Durango
  • tester
  • provamag3
  • cubers
  • Leos
  • anitta
  • lostlight
  • All magazines