How are people doing all these decompilation and recompilation projects using LLMs? What kind of LLMs are they using? What kind of hardware is needed for this?

Finally an actual use case for AI with real impact that isn't just a chatbot like chatgpt.

We would not have ps5 emulation this quickly without the help they provided in reverse engineering.

Let me remind you why:

Quote from: reddit.com/r/LocalLLaMA/comments/1wt5f4e/looks_like_the_era_of_subsidised_compute_isLooks like the era of subsidised compute is coming to an end. The old ChatGPT Pro $200 20x plan will be halved. The new $500 plan will have similar limits as the (old) $200 plan.

Quote from: reddit.com/r/LocalLLaMA/comments/1wth4iz/anthropic_just_dropped_the_greatest_advertisementAnthropic just dropped the greatest advertisement for GLM ever.

reddit.com/r/LocalLLaMA/comments/1wtjp1j/another_case_of_censorship/

Many in that sub write that running Qwen3.8-27B is good enough for 80-95% of their use-cases.

Quote from: artificialanalysis.ai/articles/aa-agentperf-localAA-AgentPerf-Local: Benchmarking local AI agents on laptops and workstations

Announcing AA-AgentPerf-Local, our open-source inference testing tool for local AI models - test how fast agentic AI can run on your own laptop or workstation, and browse our list of serving configurations to plan your next agent setup

[..]

View the initial results here:

So, running LLM AI models locally (and privately) is becoming a bigger and bigger thing and now even ArtificialAnalysis made a post about it and tested several hardware configs.

More about running AI models locally: notebookchat.com/index.php?topic=315954.0 ("Your own ChatGPT, offline: AI without the cloud on your laptop") (yes, straight to the valuable comments section).