Aurornis

10 ms·

181.
by Aurornis 2mo ago
Anthropic and OpenAI don’t have that many employees compared to most big employers. Their employees are highly compensated. The ratio between the housing the occupy and the taxes paid by their employees has to be one of the most favorable r
182.
by Aurornis 2mo ago
> are generally understood to have a civic responsibility to make contributions to their local community, above and beyond their tax obligations Literally what taxes are for. If there are expectations that a company contributes to civic
183.
by Aurornis 2mo ago
Yes. There are pockets of hardline anti-LLM policies for various reasons. SourceHut bans LLM generated code from their platform https://news.ycombinator.com/item?id=49461724 The Asahi Linux project bans LLMs. Unfortunate as
184.
by Aurornis 2mo ago
Their crime is compensating their employees too well, which this author thinks makes them responsible for higher rental prices? The other complaint is that the author doesn’t think they see any benefit from all the taxes the companies are p
185.
by Aurornis 2mo ago
> but no indication of what that means. That’s the point. There are 27 member states that each have their own regulations that need to be complied with. The article has more details than your comment implies, such as the fact that some s
186.
by Aurornis 2mo ago
> What if the model is hopelessly obsolete, and thus no demand, but I want that specific model? You can still find a lot of old and completely outdated models on OpenRouter. The providers can scale serving of models up and down as demand
187.
by Aurornis 2mo ago
> I think the biggest reason is to own the stack so your model can't be changed out from under you, The concern would be future regulations that prohibit you from buying a hosted version of the model. Even that could be bypassed wit
188.
by Aurornis 2mo ago
Gigabit is more expensive, consumes more power, and generates more heat. If a device doesn't need the speed, 100M is a better choice. Even the most high bitrate streaming services have peak bitrates under 80Mbps, with averages being mu
189.
by Aurornis 2mo ago
For the users of the project, the reasons don’t matter. They’re free to do as they want for any reason they choose. Users of the project can (and have been) moving to forks that are operated as open, community projects in the continued spir
190.
by Aurornis 2mo ago
> Unfortunately that 1.0 work is happening in a private repository. That’s even more reason to consider it non-viable. There is no reason to encourage or support this type of behavior in a project. It is their right to do with it as they
191.
by Aurornis 2mo ago
> The problem with httpx as a dependency is that it's currently working towards a 1.0 release which will be full of breaking changes. The httpx maintainer closed off access to issues and discussions on the repo, has been ignoring PR
192.
by Aurornis 2mo ago
> what about effect size and remission rate? This feels a bit like cherrypicking Trying to pull out NNT feels like cherry picking because it sounds really bad to people who don’t know how other drugs look on this measure. If your point i
193.
by Aurornis 2mo ago
> The other problem with SSRI treatment is that individually they are only barely better than placebo. They're among the worst classes of drugs as far as NNT, These numbers are really misunderstood when taken out of context. SSRIs h
194.
by Aurornis 2mo ago
> After a few weeks I was tapered off them. > I felt no different before, during or after them. SSRI onset is slow, which is one of the common problems with treatment. In studies where they survey both the patient and their family mem
195.
by Aurornis 2mo ago
Bambu Studio is open source and on GitHub.
196.
by Aurornis 2mo ago
> And AI is not acting like a manager who wants to fire you because of gut feeling or possible stock options. Imagine that AI CEO is owned by employees, You’re just describing a company owned by employees who manage themselves. The AI is
197.
by Aurornis 2mo ago
> KickStarter launch, distributing Bambu Studio binaries to every backer, reviewer and influencer without source code . I don’t think you understand how Kickstarter works. The product ships after the kickstarter ends, not the moment it
198.
by Aurornis 2mo ago
Several of the claims in the article are also not true. Bambu Studio was open sourced on GitHub when the printers shipped to Kickstarter backers in 2022, not a year later after pressure from the community like he claims. Bambu also has an o
199.
by Aurornis 2mo ago
> A classic management failure is not paying enough attention to something. This is usually not a problem for computer systems. Humans are bandwidth-limited. Teams of AIs are far less bandwidth-limited. That's a feature, not a bug.
200.
by Aurornis 2mo ago
That would make more sense. Thank you.
201.
by Aurornis 2mo ago
The article says the SFC had a fundraiser and raised over $250,000 to hire a lawyer. There's a confusing part of the article where he says that what he believes is a violation is something that a judge isn't going to care about. I
202.
by Aurornis 2mo ago
> Kühn said; they often rely on mechanisms that an actual judge is not going to care about. In this case, Bambu Studio would pop up a request to download "a little more stuff" with the classic choices of "Yes" or &quo
203.
by Aurornis 2mo ago
Presumably the efficiency numbers they're quoting are for the high concurrency state they were serving. RAM was probably the bottleneck for the amount of context they were offering. I assume it would run a little faster with lower conc
204.
by Aurornis 2mo ago
GPT-5.6 Luna is also served for free to the entire world with a tokens per second rate nearly 10X higher. > ... I'm at a loss for words here No need to be so dramatic. I think it's great that they're developing chips, but
205.
by Aurornis 2mo ago
I like this font and the license it comes with: > The OFL allows the licensed fonts to be used, studied, modified and redistributed freely as long as they are not sold by themselves. The fonts, including any derivative works, can be bund
206.
by Aurornis 2mo ago
Ox Alpha is a smaller model and it was running very slowly. Chinese AI accelerators are coming along, but nVidia’s lead is huge.
207.
by Aurornis 2mo ago
> I would love to see an Opus-4.8-level local model but TBH I just haven't got there yet. The models I've tried so far _are_ good but they aren't able to solve tough technical challenges, regardless of harness/prompti
208.
by Aurornis 2mo ago
> Closest thing to the “it just works” experience that Mac was always known for I see this repeated a lot, but the people I know who tried Omarchy were always fixing something or dealing with breakage from an update. The best argument I’
209.
by Aurornis 2mo ago
> Wouldn't the security team and the agents just go and fix the security holes? Of course they’re going to react to what is reported and fix it. That’s a given. The concern is the development process that is leading to these types o
210.
by Aurornis 2mo ago
> I'm enjoying working with Ox in a way that I'm just not enjoying talking to the 5.0 Anthropic models. That’s very valid, but right now every other model I use is easier to talk to than Opus 5.0 Opus 5.0 has an impenetrable wa

More ›