Jan Wildeboer 😷:krulorange: on Nostr: I *want* nerds to run local LLMs (Large Language Model, generative „AI“) in their ...
I *want* nerds to run local LLMs (Large Language Model, generative „AI“) in their homelabs. I *want* them to explore how they can be put to productive use. Running LLMs locally will lead us to better use of these systems. It will lead to „AI“ that is focused, limited, domain specific and allows for realistic calculations of resource use/waste. #UnpopularOpinion, maybe, but I stand by it. LLMs can be used in good ways. „Big AI“ isn’t one of them.
Published at
2026-06-05 12:37:52 UTCEvent JSON
{
"id": "f035e6aa938c98f4e2789bb1e337f87fd079860bc51442cfce02a79decd9d850",
"pubkey": "b7471ee072dd4baed58f0ae76edb7f1f85ae8662e143887b50f8648e8d9e0825",
"created_at": 1780663072,
"kind": 1,
"tags": [
[
"t",
"unpopularopinion"
],
[
"proxy",
"https://social.wildeboer.net/users/jwildeboer/statuses/116697535143479017",
"activitypub"
],
[
"client",
"Mostr",
"31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr",
"wss://relay.ditto.pub"
]
],
"content": "I *want* nerds to run local LLMs (Large Language Model, generative „AI“) in their homelabs. I *want* them to explore how they can be put to productive use. Running LLMs locally will lead us to better use of these systems. It will lead to „AI“ that is focused, limited, domain specific and allows for realistic calculations of resource use/waste. #UnpopularOpinion, maybe, but I stand by it. LLMs can be used in good ways. „Big AI“ isn’t one of them.",
"sig": "08d3f27984371bf17a3db1ab369fe1cb6177f54e66bed19bc55157f017cfd626ead1cc99b3b3c6a5ac6810d1bb59e62c649dade4b427633f10c11c57e038cfeb"
}
