Activity
Mon
Wed
Fri
Sun
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
Nov
What is this?
Less
More
9 contributions to Super Intelligent Labs
WEEKLY Q/A WITH Luke - ASK ME ANYTHING ABOUT Ai 👇
What are we working on or what do you have questions for in AI?
1 like • 3d
Never built a pc before, thinking to build one for potentially run some local model mostly just for brainstorming with obliterated models and using them to instruct other open weights to help me code certain features/ reverse engineering things. I don’t mind getting a mac mini but I would rather get the PC instead if it doubles as a gaming PC for my rare gaming days…not heavy on gaming Any good resources to start understanding what’d work the best for this situation? Thought of asking it here before i check w opus 5.5 Context : I live in canada, can spend about 2k cad, if it goes more then that, i would wanna see the ROI for max 3k-3.5k
1 like • 15h
@Mike Doolittle @Luke Brinton this is awesome. Thanks for sharing
Qwen4-Max preview: what’s out and what’s not
ok so Alibaba just posted that Qwen4 is coming, and Qwen4-Max is already in preview Qwen is Alibaba’s AI model family, kind of like the next season of a game you already play what I know right now - Alibaba says it’s better at coding, agents, and front-end builds - they say early access is on their Token Plan, plus QwenWork, Qoder, the Qwen app, and QwenStudio - Token Plan is Alibaba’s access plan for their models, and I haven’t checked the price yet real talk before you get hyped - I checked the link and it’s not live for me yet, it should be soon - I have NOT tested it, everything here is from Alibaba’s own post - it’s a preview, so it can change I’ll edit this post with the exact steps the second the link works Alibaba’s post: https://x.com/Alibaba_Qwen/status/2120335993643684315 Qwen Cloud models page (Qwen4 isn’t listed here yet): https://www.qwencloud.com/models?region=ap-southeast-1&sort=newest
1 like • 3d
Any official benchmarks yet? If its even remotely closer to opus 5.5 let alone fable… I’m cancelling my claude
Brain dump on future of Ai & risks attached to local models
What good with chatjimmy ai? It is adorably fast. Almost like what gpt was 2ish years ago. I think something like this when upgraded (a fast,cheap,reliable ai) is the future. I also know this community is really well versed in local AI models. I would love to hear your guys' perspective on how stuff going south could be prevented, because it's always good to have something running fast that gives you good results, and it's all local. Potentially, in the future, you might have something maybe running on just your Apple Watch. Who knows what? There are not always the best people around you, and someone somewhere in the world might just spin up a billion agents to have a perfect plan of mass destruction ready. I know governments are planning for those shenanigans already, but what other things are potentially a big risk, and then who is planning or doing something to prevent it? Let me know what your guys' perspective is on that part too, because local models would definitely dominate a good market in the future when corporates realize they don't need to use freaking Copilot for drafting stupid emails. I suffer from those.
1 like • 11d
@Luke Brinton I think you mentioned in one of your earlier videos, or I probably saw it somewhere else, that it just might be a part of that AI bubble when it bursts: companies that are borrowing money to buy GPUs and RAM and lending those RAMs to AI companies. The money they took to buy those in the first place, they put the RAM as collateral. Once local models get so crazy cheap and you can easily run them anywhere on any kind of device, and then they give you good enough results, this could be potentially true in the future: you don't really need as much RAM and computation. I think companies like even ChatGPT, I know they're evil as hell, so they might even be planning something like this. They're just waiting for this part of the AI bubble to burst, and then they could just buy those collaterals for really cheap and then just have dominance.
1 like • 11d
@Luke Brinton atp it's almost like electricity: it has reached that stage where people are finding its potential and building things upon it. For example, with electricity, no one knew that you'd get televisions in the future or these data centers that we have rn. But I'm curious to think of what looks like the future, because software would no longer be a commodity. You'd potentially have something hooked into your Meta glasses that you wanted to book a flight. It would just create a UI for itself, and then you can just go from there. You don't ever have to open any app. You just tell it what you want to see, what you want to do, and it had that full UI built right there. It wouldn't even take more than a second. That's how I think that would look, but then that's something I can think of. What part of my thinking kind of lags, as in, thinking of a different perspective that I could look at and better have an idea of the future? If I ever plan to build something, I don't build some AI slop that people would just write code in a day and have that same thing open-sourced (because I know that'd be it for a lot of the things happening). In these upcoming months, I saw someone posting about their game on Twitter, and then the next day someone had it on fucking Roblox, cloned almost one-to-one. Let me know what you think of it and what building solutions for the future look like. I can definitely think of helping companies and people who are in their 40s understand AI, use those local models, and prove that ROI of investing in AI. I know my company spent almost $1 million, a bit more than that, on our copilot and other licenses, but then I have people in my team that barely use it. They still have as much use as myself, so I just end up sending them my tasks that can just run on their computer for hours. I just get the results back, and I'll have my copilot, with probably Sonnet, audit the results and then just do minor tweaks, and then I'll have my things ready.
20 uncensored models you can run on your own computer
uncensored just means the model refuses less, and what you make with them is on you all of these are on Hugging Face (the site where people share open AI models) 1. Qwen3.8 27B Uncensored https://huggingface.co/HauhauCS/Qwen3.8-27B-Uncensored-HauhauCS-Aggressive-MTP-GGUF 2. Huihui Qwen3.8 27B Abliterated https://huggingface.co/huihui-ai/Huihui-Qwen3.8-27B-abliterated 3. Qwen3.8 27B NEO-CODER Heretic https://huggingface.co/DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF 4. Gemma 4 26B A4B Heretic https://huggingface.co/mradermacher/gemma-4-26B-A4B-it-heretic-GGUF 5. Muse Glimmer 30B Heretic https://huggingface.co/0bserverx/Muse-Glimmer-30B-Heretic-Uncensored-GGUF 6. LFM2.5 24B A2B Uncensored https://huggingface.co/AtomicChat/LFM2.5-24B-A2B-Uncensored-GGUF 7. Cydonia 24B v4.3 https://huggingface.co/TheDrummer/Cydonia-24B-v4.3 8. Dolphin Mistral 24B Venice https://huggingface.co/bartowski/cognitivecomputations_Dolphin-Mistral-24B-Venice-Edition-GGUF 9. GPT-OSS 20B Heretic https://huggingface.co/DavidAU/OpenAi-GPT-oss-20b-HERETIC-uncensored-NEO-Imatrix-gguf 10. Hermes 4 14B https://huggingface.co/bartowski/NousResearch_Hermes-4-14B-GGUF
1 like • 11d
Realizing how slow they’d be on my 5 yo 8gig ram laptop and how I can’t install it on my 32 gig ram work computer 😔🥀
1 like • 11d
@Luke Brinton immm lari man, if i end up making a twin to corona. You got my back right?
YOOO WHATS GOOD EVERYONE! HOW ARE WE USING HERMES DROP USE CASES BELOW
How does Hermes help you with your business or just coding wise and why do you use it?
0 likes • 11d
I am well versed in Telegram bots and those API connections, but I've never used Hermes at all. I've heard about it for a couple of things. I couldn't get into that iOS test beta either, but I'd love to try Hermes because I'm using Muse for a couple of things, and it works fine. If there's a way that my data doesn't go to Mr. Zuckerberg, it'd be even better, like cake on top of a cherry, right? Let me know if there is something you guys know I could get started with. I'd love to start with something that doesn't cost me much, and then, if I see the ROI, I can potentially hook a cheap model like Sonic 5.5 up, or potentially even something that I can run that doesn't increase my AI bill.
1-9 of 9
Manjot Singh
2
6 points to level up
@manjot-singh-4294
Lowk an AI larper atp … but the more you larp, the less you larp

Active 15h ago
Joined Oct 31, 2026