There's this weird tension where small models are getting a lot better, but the gap between the frontier and "what I can run on my gaming GPU" at home isn't necessarily getting smaller. That said, Qwen-3.8-27B-Q3 runs on my 9700 XT and performs well enough for tool calling in agents, which is a critical floor in local model performance imo.
Do you use qwen for any coding? Do you make a frontier model orchestrate it? Are you running with llama.cpp? Sorry for the barrage of questions, but I've been interesting in local AI for a while, yet I've struggled to set up anything usable.
My Qwen testing has been with Hermes trying to do mundane (to me) things that hosted models understandably refuse to do, even open-weight ones like Deepseek V4.1 Flash. Example: I tried to get it to find a torrent for $tvshow and add it to qbittorrent (connected to hermes via MCP already). It certainly tried and I think frontier models would have done better. My biggest issue was context space more than anything else. I had to drop down a quant level to increase it to 128k. On 64k it was constantly compacting.
Adding complexity would imply you're doing both. In this case it would be more correct to say you're trading the complexity of key distribution and access control via keys for the complexity introduced by tailscale ssh.
if you are one of the ones who is valued enough to keep their job, you are strong enough to go get hired at another FAANG. Do you actually still want to be there after the bloodbath that the follow-up round of layoffs causes?
when you start somewhere after their bloodbath, you don't have to deal with all the crap you would if you stay through a bloodbath (i.e. you aren't on the hook for all the legacy shit your team owned before it had a haircut)
>They could network their smart glasses and leverage their ubiquity to help people find lost pets — but strictly only lost pets, not people — but then Something Labs might ship a tool to make a person's face look pet-like enough to turn Find My Pet into Find My Ex.
>Google, Apple, and Microsoft are each in a stronger position to win consumer AI because they're guardians over the foundations of the user's experience.
From everything I've seen, they're all too large and dysfunctional to make a competent play in consumer AI. Anyone in there trying to work on it is too far removed from end product and is answering to too many masters.
>But what I wonder: can we legitimately grind on Physics or curing cancer?
If we can enable a feedback loop, yes absolutely. But feedback loops for things in the physical world like these are measured in months per cycle usually.
There are really fringe perspectives out there that imagine capitalism as a living entity of which we individual humans are simply parts of the whole. Governments that represent the economy over the people are exactly what you'd expect if that were true.
We can lower our standards instead
reply