fwn

7 ms·

1.
by fwn 3d ago
This comment specifically was also auto-collapsed for me, without being marked as flagged or dead. It might help to ask moderation about this. Could be an artefact of brigarding or something similar. I also wouldn't worry about individ
2.
by fwn 4d ago
No. And no alignment, guardrails or slowdowns will ever secure our insecure infrastructure. Bad infrastructure decisions are a risk independent from AI. The only way to secure infrastructure is to actually do the work needed to secure it. O
3.
by fwn 4d ago
If we focus too much on the world domination fiction, we really might, in a moment of lapsed attention, end up plugging essential infrastructure into the public internet, hoping some dream of global alignment might save us. That's like
4.
by fwn 4d ago
The biggest AI risk, by far, is the concentration of power in a few companies, located in the US. There is a lot of fear marketing about our text generators turning into Terminator. But other than the centralization of power, such fears are
5.
by fwn 4d ago
I made the same transition, and it felt very seamless. I didn't look into the Obsidian first-party sync service, since I was already using Syncthing at that point. For even more contrast with Evernote: the fantastic Obsidian applicatio
6.
by fwn 6d ago
You seem to have conceded the "is it deterministic" argument only to sidestep by declaring determinism irrelevant. Your original claim was that LLMs are deterministic "in the same sense" as brains. We can write down an L
7.
by fwn 6d ago
No, that is not at all something we can just state as a fact. Whether the brain is deterministic is an open question that just inherits the good old, probably unsolvable determinism debate. The LLM pseudo-randomness from above is engineered
8.
by fwn 6d ago
LLMs are deterministic, though. Much like the video. AFAIK using the same input tokens, weights, and numerical operations will lead to the same probability distribution for the next token. It uses pseudo-randomness to enable temperature, et
9.
by fwn 10d ago
I bought a Mac mini M4 simply because it was very cheap and the M4 chip offers great performance. It is currently mounted on the back of my TV and serves as a media station. It runs macOS, but I have no particular interest in the operating
10.
by fwn 11d ago
Irrespective of political affiliations, I felt that Mullvad addressed the outrage in a mature and non culture war manner. Because of how they handled the situation, I have a lot more respect for the company. It's great that they didn&#
11.
by fwn 13d ago
(Not the patent) I'm very much pro letting children roam freely, but police intervention for walking or driving with your own child is clearly overreach. The good intention does not change that. And even as a German who already spent a
12.
by fwn 15d ago
I think that this take is unfair to the parent. Many VLOP requirements currently lack best practices, and it's clear that affected companies borrow from each other to mitigate compliance risks. The commission even seems to encourage kn
13.
by fwn 16d ago
Neither the media nor politicians are interested in hyping up the issue right before the state election on Sep 20. So it's probably just what you suggested: pretending that nothing happened.
14.
by fwn 16d ago
I see my style of thinking and writing everywhere in the world now. What happened? On a more serious note, and to tie the fun AI comment into the actual discussion: I suppose the types of medium through which people try to express themsel
15.
by fwn 17d ago
Chat Control 1.0 is not at all voluntary for those controlled by it. For example: there is no setting in Facebook Messenger where users can enable or disable Chat Control. This is like saying that private prisons are voluntary because the c
16.
by fwn 17d ago
If this area of research interests you, look into Thompsons "Designing Responsibility: The Problem of Many Hands in Complex Organizations" from 2017 or Bovens ur-werk "The Quest for Responsibility" from 1998. Diffusion o
17.
by fwn 19d ago
Unless you set up something like a brick-and-mortar Trustpilot competitor, you are likely to only ever read online about user feedback on anything. But this does not say a lot about the quality of the feedback. I am always surprised that pe
18.
by fwn 21d ago
Organizations very much rely on websites for the information they share. I don't even know how most data would need to be sliced to fit into tweets. Like, how would you tweet exchange rates consistently and accessibly? https:/&#x
19.
by fwn 22d ago
AFAIK, you cannot run Claude without one of their mandatory system prompts at all. If we wanted to compare model responses, we would give all models system prompts with model names, thereby fixing the Kimi misattribution. The reason Kimi of
20.
by fwn 24d ago
Imagine a headline reading, "People are starving in [city]." Someone might reply to that news by pointing out that starvation is as old as humankind. ...older even. Would that remark be on topic? Would it be noise?
21.
by fwn 29d ago
OpenRouter does not offer private inference, so your plaintext is shared with companies you may not even know by name. Tinfoil OTOH claims hardware attestation and confidential computing, which is a pretty strong and (in theory) verifiable
22.
by fwn 29d ago
Mozilla PR department seems to be vanilla Claude. tl;dr of the article: Rather than building a closed AI stack (like Google - but they don't name it) Mozilla is pursuing partnerships. In this case, they are partnering with Exa.ai to of
23.
by fwn 1mo ago
Whenever I start reading an article and notice traditional AI markers, I throw it into my AI summarizer to see if it's worth reading. Fighting fire with fire. For example, this article (allegedly) makes two main observations to justify
24.
by fwn 1mo ago
A friend of mine had a younger brother who sometimes played Worms World Party (the version of my time) with us. Since he was much worse, we implemented a handicap by not using the more lethal weapons against him. I still remember him giggli
25.
by fwn 1mo ago
> By disabling signature verification you open yourself up to man-in-the-middle attacks and supply chain attacks. I am no expert, but AFAIK Android does not disable signature verification when installing from "unknown" sources.
26.
by fwn 1mo ago
Reproducible builds are nice, and I would love to have them for Signal. But I think I disagree with most of what you've written. Enabling "unknown sources" does not make my device less secure. The setting is also disingenuous
27.
by fwn 1mo ago
> [Signal is] obsessed with [...] user tracking, via app store stats. Not sure about that. You do not need Google Play Store to download or update Signal. I use Obtainium to update directly via signal.org for a very long time now. JSON:
28.
by fwn 1mo ago
> When you start denying IAMP to your customer in the name of "encryption" at that point it becomes privacy theatre, instead of privacy, irrespective of how nobly activist their intensions are. How would the lack of IMAP compro
29.
by fwn 1mo ago
I have no experience with style imitation, so perhaps I'm missing something here. I asked Kimi K3 to rewrite your comment in the style of Ernest Hemingway: > I would like to get the old models back. If anyone can show me the way, I
30.
by fwn 1mo ago
I am on the paid plan since mid-2015 and nowadays I regularly get ads for concerts in the UI. The ads cannot be disabled and there is no amount I can pay to stop them. I even wrote to support, only to get a sloppy reply telling me no: becau

More ›