aetherspawn

It’s not like people don’t know the references, they just don’t want to engage in a public argument with you.

Old Testament: Lev 18:22, Gen 19:1-7, Judges 19:22-23, Jude 7:7

New Testament: 1 Cor 6:9, Romans 1:26-27

And your accusation about Hypocrisy in eating pork etc:

Laws of Moses superseded: Gal 3:23-25, Romans 10:4, Col 2:14, Col 2:16-17, Romans 6:14-15

by New Testament principles: Acts 15:28,29 and Acts 10:9-16

Hey stranger, welcome to 2026. It’s somewhat different to what you’re used to in 2035. We do things differently here.

This stuff is a nightmare for new manufacturers and is usually lobbied-for by large OEMs or to keep startups out of the market or as a patent trap

The most recent regulatory disaster that blew up a bunch of startups was mandatory lane keep assist for trucks in overseas western markets, which meant all new startups needed fancy steering racks which are very much not off-the-shelf, and it virtually tripled the cost of the software stack too

I don’t think this is true. All the models prior to Fable were honestly dumb as rocks, and Fable is too sometimes, but at least it’s helpful now and not a hindrance.

The future of AI most definitely involves making something twice as good as Fable that is virtually its own employee, and not on reducing inference costs because to be honest Fable isn't actually that expensive.

The real utility behind an AI model (imagining that it can be made twice as good as it is now) would be being able to scale a small business up and down instantly without hiring (to implement a new feature or whatever), which is costly and time consuming these days.

Can the 30 second embedding time be done beforehand and sent to the browser?

Inference is nice and quick after that.

Is it as good as Fable..? Fable is the first model that mostly writes without the AI slop format for me, and so I can comfortably actually copy and paste most of what it spits out.

OpenAI models have always been the worst in my experience for verbose, slop formatted responses, with each generation increasing in sloppiness.

Just because something doesn’t perceive to be lightweight doesn’t mean that it isn’t. Many modern web frameworks are extremely lightweight (ie Svelte), but it’s complicated how they achieve it, and the source code looks like sludge.

Mr. Robertson: That's the way it has to be. That's the way it's always been. You should understand that better than anyone.

Temporal GPT: The snake that trains on its own tail, forever and ever?

I imagine not just anyone gets to clean a million dollar toilet.

Surprised models still output tools as text when for ages we’ve been able to constrain the output at the inference engine level and constrain the model what tools, parameters etc are available

Edit: found it, it’s called Grammar-Constrained Decoding (GCD)

It makes sense to me that distributing across more parameters results in models that can be quant more heavily (information theory - more bits available)

I wonder if anyone has figured out how the information is compressed and calculated the amount of information an LLM can hold depending on its size

80 tok/s which is kind of a lot for GLM. My experience running 80 tok/s on other LLM is that it ~seems faster than cloud inference, but that obviously depends what you use, in my case ChatGPT.

Disagree, I assumed it was a 10 year old story because it was written in the personable style of the old internet…

Compared the form ID in original submission and after redirect and it looks ok…

It’s like an angsty book written by a teenager with only a middle-school level understanding of the world: throwing a few completely random words in there to sound smart.

Hard to believe that something that writes so terribly is so good at mathematics, given that writing non-slop must be at least some part formulaic.

Came here to figure out why this Ad has 30 upvotes…

I don’t know about the advice on market size… nearly every 2nd founder says go into beach head markets and stay away from markets with high competition.

The ones in beach head markets seem to succeed more often on average (anecdotal)

Ah yes but because it’s a MoE 48GB active model, then it’s possible that we might be able to run it locally in specialised setups such as 256GB unified memory.

Many MoE models (seem?) to only require enough memory to load the active expert.

I wish they would release the requirements to run on llama.cpp with any announcements of open models.

A bonus would be tok/s on common hardware.

More detailed response below, but I believe the best way is to use a router that can filter DNS requests (ie Unifi), and set the upstream of that to your ISP.

You can set the DNS server of your entire LAN using Network options without setting it on the individual devices. It’s up to the end-device whether it uses it or not, but 99%-ish do.

Heck. I wonder if they surveil Outlook in western countries.

Now that Cloudflare Email is out of beta someone should build a self hostable serverless email stack based on it. Then we can deploy our own email with reliable delivery, but own our own data and only pay for actual usage as well.

A good way to work around this is to have a small server in your house as the first hop for ad blocking, and use the second hop (set the DNS servers of your DNS blocker) as your ISP.

If you use a Unifi router you can use it for this purpose (and caching) by broadcasting its IP as the default DNS server using Network options. It caches/forwards/filters any requests according to how you setup DNS in Firewall settings.

Use your ISPs official DNS so that you get the shortest path possible from the ISPs handoff location to the CDN (and overseas trunks), not a generic DNS that doesn’t know about your ISPs layout.

ISP: 1ms to Cloudflare

Cloudflare: 10ms to Cloudflare

Thank you for your attention to this matter.

Edit: will clarify, this advice applies to countries with good privacy laws and no national surveillance i.e. not the USA

I can run most models 100B and under on my MacBook Pro with Ultra 3 and 128GB of RAM at 25-70tok/sec-ish, and it was around $5k.

I don’t think I can run GLM 5.2 since it requires around 256GB of memory and the inference is probably too slow, but the future (planned) Apple M7 may be able to. The leaks say it will support up to ~700GB of RAM.

The models under 100B are kind of dumb as a brick and aren’t that useful unless you’re really bad at coding imo. They can’t really be trusted not to hallucinate so they’re not even good for data processing.

LLMs find more vulnerabilities than people because people time is heaps more expensive than LLM time, that’s it.

We’ve always been able to find heaps, we’ve just never had the right structures to put in the effort and renumerate people for looking (even if they don’t find anything).

I don’t mind playing 2nd from front. On a lot of maps this leaves 6/8 roles that I like. On others, it leaves 4/8 roles.

If people really liked playing those roles, why am I always forced to play them? Shouldn’t everyone else be jumping to grab them? Lol

I just want to say that I have a friend that floats around OS 25-30 and he tells me exactly what’s required to reach that level.

I’m 100% aware that most high OS players on Front are using interface mods to manage their units and send back hurt units automatically to get healed.

In fact, he’s not even that much better than me. We still do 2v2 lobbies together, and we have good and bad games and carry each other.

I know what other people are doing. I’ve been playing for over a year (200 games?), and BA on Spring for years before that (500 games, I think). I’m not naive or particularly bad at the game. It’s just not how I want to play, and when I play those positions I’m at a disadvantage because I know other players are “legal cheating”.

And don’t gaslight me lol, every single game I could see the other players falling back their individual units in big crowds 1 hit before death, and I’m telling my friend about it and he’s like “oh yeah everyone above OS 20 uses this mod, which is basically a cheat”. That’s kind of when I decided to stop playing.

Also, the community wasn’t like this in Spring BA! That community was the nicest ever despite the game being basically identical, engine and all. This is a BAR issue.

I played for a year, got my friends into BAR, one of them is OS 25 (that’s high) and still plays. We still share this viewpoint with each other which we both agree with.

Hey PtaQ, I appreciate the comment and that BAR is putting effort into improving the community, but I raised a pull request to edit the community code of conduct to clarify that players shouldn’t be banned simply for allowing a game to run its natural course, and virtually every dev disagreed with me and favoured the idea of kick banning a player who won’t surrender immediately after front is lost because “the game is clearly won”.

If you want to make the community better, delete all the rules that encourage players to gang up on other players for subjective reasons. These kinds of rules punish non-verbal players, who easily fall prey to verbal players (I’ll just call them bullies) who can use the rules as a justification basically to harass someone whilst getting branded a hero.

Also, a large majority of players have fun, even when they’re losing. Asking those players to accommodate a hyper-competitive few who want the tightest loop possible to winning and restarting is anti-fun. Both from the perspective of the person winning (I get so disappointed when I win front and they just surrender before I get to blow up their base) and from the perspective of the non-front players who are pressured to quit as soon as front is lost.

Won plenty of games in more casual lobbies (kind of rare) where the front player, after dying, was happy to just receive T2/3 units to control, and due to being able to focus on just controlling a few units, won the game.

My OS floated around 10-15 and I virtually always ended up having to go front in every lobby, which is the position I hate the most because I don’t like playing my strategy game like an RPG - that is, micro-ing just a dozen of units and optimising every rocket or bullet hit or whatever, which is kind of what front is like.

Front has zero opportunity or resources available to build any kind of economy, and once the T2 units start coming through from the other side they feel very expendable. As the front player you build the same 1 or 2 units every single game and never really get to strategise.

What also enraged me is that the back players would have the nerve to make the front player “pay” for their T2 constructor units after working so hard to keep everyone alive, despite everyone knowing the front player has zero resources at any given time because it’s all going into units that are being meat grinded.