The CIV50 would be interesting in the game ECO (https://play.eco/) where players specialise in roles and need to work together to survive while working on a common goal of stopping a meteor. There can be some overlap but ultimately players can't reasonably fulfil every role alone.
DNS resolvers getting sued should stop resolving domains/properties of the companies suing them too when they are forced to take these kinds of actions/steps off the back of a legal decision like this.
While I agree in spirit, this also makes the resolver away from being able to claim they're just a neutral service which just delivers existing records without being opinionated
Fair point. I'd say "We're neutral, unless you attack us, in which case we stop being neutral about you, in particular". I'd still use them if this were the policy.
I can't help but imagine that the sort of user choosing to use a service like this in lieu of their ISP default is also likely to support this bit of pain.
...also, it's probably their own PlayStation, that their kid uses. Speaking from personal experience.
I mean if they're getting sued for failing to block proxy/vpn services they're being told to be opinionated and they can't be a neutral service. "You asked us to be opinionated about traffic, so we're doing that. Do you want us to go back to being neutral?"
That ship sailed. The court unequivocally determined that the resolver is not a neutral party, and their ruling explicitly demands that the resolver not be neutral.
That's risking yourself getting flashbanged on a plane for suspicion of human trafficking and sacrilegious creation of fictions and engaging in Antaractican nuclear programs.
The problem is that these types of companies are started by naive idealistic techs who only know how to play by the rules and can't figure out why people would misuse their products
They were very well described by the "rubber hose crypto" XKCD
(Not that I don't think a DNS company stays at the same level of complexity - both legal and technological - as a VPN company, but still)
Or, get this, anything that is useful can be misused, so maybe we should go after the people misusing things rather than locking all of society in a cage to pre-emptively prevent misuse.
"The problem is that these types of companies [kitchen knife manufacturers] are started by naive idealistic [craftsmen] who only know how to play by the rules and can't figure out why people would misuse their products [by stabbing people]."
Hopefully the leadership of public resolvers isn't as a childish as you are.
That kind of action would penalize their users far more than it would hurt the evil corporation that slighted them. Not to mention they'd lose users. People who choose quad9 likely aren't too keen with DNS tempering.
Isn't the intent of rejecting, or boycotting, a company supposed to cause a disruption in the first place? If the people boycotted a company into bankruptcy, because said company refused to change their behavior, would the people be in the wrong because it might hurt the users? No.
While this is generally true, it's _a little_ less true the larger the model is.
Also, quantization techniques have improved - the I in IQ3 stands for imatrix - Importance Matrix - it is a bit more surgical in what it cuts. The result is a model where the most important weights are even Q6 or above, the least important Q2 or even below, overall it takes the space of a Q3 but with better results.
A lot of words, but no benchmarks. I'm a little tired of all the "finger in the air" vibe checks. You can say all you want, but you'll only know once you put out some numbers. Which people have in other threads, and 4 bit 27B beats Flash at 3 bit
This is the conventional wisdom, but in practice what matters is how reliably the model performs on your tasks in the real world. I have an R9700 and an RTX 5060 Ti and I've been running an IQ3_S quant of 27B on the 5060 Ti vs a Q6 quant on the R9700. I still manage to get stuff done with the IQ3_S quant.
I was an early adopter of Zed (private alpha mid-2022) and it's crazy how far they have come in a relatively short space of time. Sadly I stopped using Zed when the push of AI features started to happen (same with Warp terminal) and have since used Gram more. I may have to give Zed another run as I believe you can turn all the AI features off now?
I have had an eye on this for a while (found via pi.dev) but I don't really have a solid use case for it, but the idea/concept of is appealing where the price is not. I can buy a £100-150 mini-pc with better hardware to run 24/7 for my own VMs extending my homelab (granted my ISP doesn't put any restrictions on me, I know many others can't say the same).
And even of those models trained for tool calling and agentic flows, mileage may vary depending on lots of factors. Been playing around with smaller local models (Anything that fits on 4090 + 64gb RAM) and it is a lottery it seems on a) if it works at all and b) how long it will work for.
Sometimes they don't manage any tool calls and fall over off the bat, other times they manage a few tool calls and then start spewing nonsense. Some can manage sub agents fr a while then fall apart.. I just can't seem to get any consistently decent output on more 'consumer/home pc' type hardware. Mostly been using either pi or OpenCode for this testing.
The opposite is true for many of us. Having used various models, tools, agent orchestration etc.. All of these sensationalist posts of 'When I use Model X with Tooling Y I can build the world!' just don't resonate and it becomes draining to constantly have all this pushed through my eyeballs and having to filter it out.
LLMs have their uses but it isn't as great as everyone makes out to be nor is it as bad as others make out it is. Every week its X model has new SWE bench and is the best in class frontier blah blah - yet its actually just much the same as the week before. Quarter to quarter you could argue there is more of a diff between capabilities and performance but the LLM news cycle is much shorter than that.
This is what is great about it, the community posting hyper-creative (sometimes cursed) solutions for fun! I usually use AoC to try out a new language and that has been fun for me over the years.
> not like they changed anything significant about the product itself
They had already done this months ago, they changed the licence and stripped a lot of the code of their admin UI/object browser. So a lot of the features vanished for people overnight if they updated. This is what the OP is linked to - a fork at the point that they did this. This was work already done, feature already widely in use, they decided to take it out of what was available to the community.
So the product has significantly changed and offering had been reduced. That in addition to the stopping of publishing images - both without notice - caused a decent bit off community 'wtf'.
reply