lollobomb

Now, can we please get the long-awaited full system backup feature? We pay you for a reason!

(sarcasm obvious, but indeed full system backup is sorely missing. Seedvault is crap, not an option)

I just stumbled completely randomly (I mistyped a search term) on this website. It's like, jump back to the 90s. AFAICT is a small family website of a random old American couple, with pictures of their family get-together, anecdotes of their history, genealogy, etc. Just some mundane content of a mundane family, and yet it strikes my chords somehow. Absolutely not high-tech, but I felt I had to share it. I wish the internet could be a bit more like this once again. Thanks.

Heh, I could do it for my subordinates (and I don't need to, I made pretty clear with them that I have zero tolerance for this shit and they seem to comply), but for other teams it's not so easy, the environment is pretty brutal in terms of politics, if I start sabotaging the "SUCCESS" of some dumb BD, the manager will comply with me and the CTO.

This quote from the original blog post resonates with me:

> The room had been arranged in such a way that saying so was not a contribution; his managers were too invested in the appearance of momentum to want the appearance disturbed.

Yes, I know, I should learn to be more subtle. I just don't have the energy for this stuff. I am tired.

Fuck, yes. This.

I work in an "AI-first" startup. Being "The Expert", my work has become 90% reviewing the tons of crap that confident BD people now produce, pretending to understand stuff that has never been their domain, proudly showing off their 20-pages hallucinated docs in the general chat as the achievement of their life.

"Heads up folks, I wrote this doc! @OP can you review for accuracy and tone pls?"

And don't hit me with the smartass "just say no", it's not an option. I tried that initially. I have a pretty senior position in the org, I complained to the CTO which I report to, and with the BD managers as well, that I do not have bandwidth to review AI-produced crap. After a couple of weeks, CEO and leadership in an org call spelling out loud that "we should collaborate and embrace AI in all our workflows, or we will be left behind". They even issued requirements to write a weekly report about "how AI improved my productivity at work this week". Luckily I am senior enough to afford ignoring these asks, but I feel bad to all my younger colleagues, which are basically forced to train their replacements. I am not even sure at this point whether this is all part of the nefarious corporate MBA "we can get finally rid of employees" wet dream, whether it's just virtue-signalling to investors, or if CEO and friends genuinely believe their own words. I have the feeling leadership (not only in my org) has gone in AI-autopilot mode and just disappeared to the sunny tropical beaches they always wanted to belong to.

I would happily find another workplace at this point, but you know how the market is right now, and anyway, I have the feeling that this shit is happening pretty much anywhere money is.

Everyone feels smart now, and it's a curse.

God, how I hate this. It's making my life miserable.

Exactly this. The fact that you got downvoted tells a lot about how much political propaganda has infiltrated HN.

Very relevant: https://news.ycombinator.com/item?id=34486824

That is not really true, or at least it's very difficult and you lose accuracy. The problem is that the definition of "Open Source AI" is bollocks since it doesn't require release of the training set. In other words, models like Qwen are already tuned to the point that removing the bias would degrade performance a lot.

Mind you, this has nothing to do with the model being Chinese, all open source models are like this, with very few niche exceptions. But we also have to stop being politically correct and saying that a model trained to rewrite history is OK.

Sorry, this is utter BS, or at least it doesn't apply to all situations.

I played exactly this script during my first interviews, never giving the first number, etc. It worked well for me exactly one time: the very first job I had out of academia where I did actually have to negotiate a salary. All the other times it resulted in waste of time and loss of sanity.

I am now very upfront with my expectations, or at least my non-expectations: "My current salary is X and I have Y perks, and I am in a good team; I wouldn't even consider moving for less than this, regardless of how cool your team is, and if you don't beat these numbers substantially I already know that my current employer will make me a counteroffer, and then you lose time but I win regardless." (of course not literally, but the message is this).

Less interviews, less time wasted.

I use Graphene OS and I like it a lot, but 1) I have the feeling that, with Android's Decree coming, they are counting their days left to live. Unfortunately they built an amazing OS on very shaky foundations, it's not their fault, it's the mobile OS ecosystem that sucks. And 2) They (or, better, their benevolent dictator) tend to be very silly when it comes to threat modeling, as in "my way is the only one that makes sense". Personally, I prefer to use a browser like Firefox that allows me to block every annoying ads and to customize my experience as I want, rather than a super-secure fully isolated browser like Vanadium that a) does not replace Chrome anyway for many websites that require strong attestations (e.g. Wise's verification works on GOS with Chrome but not with Vanadium), and b) it's still based on Chromium, so still built on shaky Google foundations. With Mozilla's questionable choices over time, I keep my fingers crossed for Ladybird or Servo, or similar.

Because Threads is Meta's attempt at bullshitting Mastodon users in welcoming a wolf among the herd. Search for "Fedipact": Meta is de facto cut off from many Mastodon instances.

As for another data point, opening a Mastodon account for me was as trivial as choosing one random link from a webpage, choosing a username, password, and entering my email. If you get frozen on the choice of instance, that has more to do with your mental process than with an effective difficulty for the average user.

If you really cannot go beyond your inclination, and since you are a a long-time Internet user and a nerd, why not host your own instance?

This and the recent HN entry on "open social" smells to me like an astroturfing campaign from BlueSky.

https://news.ycombinator.com/item?id=45388021

Protip: BlueSky is "decentralized" only in name. Don't. With all their drawbacks, ActivityPub and Nostr are way better.

Yes, 100% this. Plausible deniability for everyone does not hurt everyone, it rather protects everyone, even if you don't use it.

The submission is a bit old but let me try anyway since I see some user claiming to be Graphene community manager here. Let me first reiterate that GOS is an amazing project and that I am super grateful for your work. That said, I think the #1 missing feature is the lack of a robust backup solution. Las time I checked, there was an ongoing discussion about shipping an ad-hoc backup system for GOS. ANy uodate on this? Thanks!

Hey, thanks, I missed the news! Will try again!

Can you please clarify the Revolut part? Just to understand, you are saying that you are able to perform NFC payments via the Revolut app which you installed on your Graphene OS through the official Play Store? Where are you based? (asking because I start having the doubt that it might be geo-dependent)

Can you please clarify the Revolut part? Just to understand, you are saying that you are able to perform NFC payments via the Revolut app which you installed on your Graphene OS through the official Play Store?

Can you please clarify the Revolut part? Just to understand, you are saying that you are able to perform NFC payments via the Revolut app which you installed on your Graphene OS through the official Play Store? And you are based in Poland?

OK, and then these HN users who report being able to pay via NFC with Revolut on Graphene OS are lying? Sorry, I am confused :|

Yes, I know the issue, but my question was more: is Revolut one of such banking apps?

I am a long time GrapheneOS user, amazing project. One thing that is not clear to me is the support for NFC payments. Las time I checked, NFC payments on Graohene didn't work at all, but I am reading on this thread that some users do manage to pay via NFC? Did Iget this right? Mind explaining how?

I do not use banking apps (I only use banks that allow me to log in via browser using a 2FA which is not a proprietary app, like a FIDO key or other physical dongle), but do I get it right that Revolut would allow me to pay via NFC in this case? Is this something geo-dependent?

[2019]

Wow, this is extremely cool! Only problem, the JS slows my Firefox almost to freezing, is it normal?

I use this and it's great. Only problem is when: 1) you want something outside of github (from my experience, already gitlab and codeberg can be buggy here, although very rarely), and 2) when you need a specific release channel (example: Firefox Beta, which requires a bit of work). But overall it works great. Now, one has to consider the security aspects: stores like Google Play (and, to a lesser extent, F-Droid) do perform some antimalware checks. It's not bulletproof, but it gives a bit more trust in case the dev goes rogue or is compromised. BUT you have to trust the store. With Obtainium, you have to trust: 1) the app's developer 2) Github/Gitlab/Codeberg 3) Obtainium's developer. So, it depends what's your threat model. I'm looking forward to seeing wider adoption for Accrescent!

It is so sad that there are so many vultures eager to make a profit out of desperate people with terrible diseases. This is why ethical guidelines and FDA regulations are for, after all. Of course it would be great to judge case-by-case (the scientist in the article is a real hero!). I would like to see a detailed cost-to-benefit study on the topic of allowing unapproved treatments for terminally ill patients.

This is nice, I tried a plugin for pandoc in the past but didn't really work well.