I'm a big fan of the idea of gathering a few million dollars in donations for GPU hardware and building a job queue software factory for Apache Foundation and other non profit software custodians. Maximize hardware utilization, work your way through the value stack of replacing software with open replacements. This is a distributed computing problem fundamentally: jobs, queues, workers, staging and moving underlying data around (imho). In goes requirements, specs, context, etc, and out comes MIT licensed code for all to build on and extend. Similar for a reverse engineering factory.
"Put the software factory and reverse engineering machine in the rack bro."
I'm really curious what this will end up doing to video games. There's a lot of modding scenes that are starting to get mods that are essentially just a different game stapled onto the first one with ai assistance.
I think this is going to kill fast-paced competitive gaming. The cheat scene is going to go bonkers, to the point where competitive games will require the server to become fully authoritative (i.e. tick-based server-heavy games like MMOs, or full on cloud streaming gaming). Neither are compatible with very fast paced games.
Oh no, people might jailbreak the os on their TV or refrigerator and remove all the ads...
We can't have that...
The companies won't give up their scams without a fight.
They'll double down with secure boot and encrypted everything.
Let's imagine that virtually any software can be reverse engineered with enough tokens.
The bottleneck is the human effort to deploy the tokens.
I imagine that bottleneck is much, much, less of a constraint for openAI, anthropic, and the ilk.
I'm a big fan of the idea of gathering a few million dollars in donations for GPU hardware and building a job queue software factory for Apache Foundation and other non profit software custodians. Maximize hardware utilization, work your way through the value stack of replacing software with open replacements. This is a distributed computing problem fundamentally: jobs, queues, workers, staging and moving underlying data around (imho). In goes requirements, specs, context, etc, and out comes MIT licensed code for all to build on and extend. Similar for a reverse engineering factory.
"Put the software factory and reverse engineering machine in the rack bro."
DeepSeek just open-sourced its inference engine - https://news.ycombinator.com/item?id=50163277 - October 2026
Can someone explain how does this work behind the scenes?
Also, how much would be typical AI token costs of doing this for various apps, simple to complex using some reference AI model?
Thanks.
the future is going to be wild
I'm really curious what this will end up doing to video games. There's a lot of modding scenes that are starting to get mods that are essentially just a different game stapled onto the first one with ai assistance.
I think this is going to kill fast-paced competitive gaming. The cheat scene is going to go bonkers, to the point where competitive games will require the server to become fully authoritative (i.e. tick-based server-heavy games like MMOs, or full on cloud streaming gaming). Neither are compatible with very fast paced games.