2damoon

5 inference backend in 1 single app. Access to all models and methods. For MacOS. Control Qwen Thinking and chose LLM or VLM. kruks.ai github.com/scouzi1966/ves… #llm #apple #LocalAI #Qwen3

Powerful post!

I'm tempted to buy $BABA here and hopefully invest the gains on private AI gear. We should all do that. I'm not against Frontier labs at all. They are technically awesome! But we need choices. Small businesses need access to good-enough AI. Open Source and proprietary software has co-existed fro many years now. In fact, AWS and Azure strived by re-packaging OSS and bundling them into marketed services.

I'm starting to wonder if there will be some over-investment here. Sure, AI requires memory and lots of it. The question is how much and at what costs? Unless there's a wide adoption breakthrough outside of Coding agents, we may see memory prices fall back to nomal. finance.yahoo.com/technology/ai/…

@tobiflamme Let's connect and build TOBI! github.com/scouzi1966/mac…

🎉 I just hit 34K followers! From 0 to 34K through consistency, replies, and showing up every single day. If you’re building too, let’s get acquainted 🤝

@Sainoleno GM Saino. Let's connecy and build! github.com/scouzi1966/mac…

Looking for active mutuals today? 🔥📈 Say 👉 "GM" 🌞 Connect with supportive people and grow your network 🤝💚

@ZunnuMetaX Let's connect and build! github.com/scouzi1966/mac…

Small account? Big dreams? 🚀 Drop "NEXT" 👇 Let's connect and grow together 🤝🔥

@jun_song $20k is what a small rack mount server costs in a data center now from Lenovo, Dell, HP etc. With 128 Gb of memory and 2 small Xeon CPUs. It's not overpriced. These DC versions are more robust etc but still.

Mac Studio M5 Ultra 768GB price prediction: Calculation assuming they maintain the post-hike pricing: > M3 Ultra full chip 96GB base price: $6,799 > MBP 64GB -> 128GB RAM upgrade cost: $1,600 By simple math: $6,799 + $16,800 + $500 (price bump for base 2TB SSD) = $24,099. This is just a simple calculation. To actually run a 5-6TB model, I expect you'll have to pay at least $30k+. My prediction is it will land somewhere around $35k to $40k.

@atanu__07 Let's connect and build it! github.com/scouzi1966/mac…

Hey Guys Looking to connect with more builders here. If you're into: • AI & ML • Software Engineering • Startups • Web Development • Side Projects • Building in Public • Audience Growth Say hi and let's connect #AI #BuildInPublic #WebDevelopment #IndieHackers #TechTwitter

@ZunnuMetaX Yes I do!

Want 500+ organic followers? 🌱 Comment “Yes” 👇✨ Meet active mutuals and boost engagement 🚀🤝

@ChainChaserVN Let's connect and build! github.com/scouzi1966/mac…

Bro to Bro: build your x account now 💜 Just say “Hello” and gain 2000 more mutuals here.👇

@james84_ Let's connect and build! github.com/scouzi1966/mac…

Bro to Bro: build your x account now 💜 Just say “Hello” and gain 1000 mutuals here 👇

@james84_ Let's connect and build. github.com/scouzi1966/mac…

Gain 800+ followers today 🙏 Say Hello 👋 Let's follow you immediately⚡💚

@kevinxu It makes IT management more complicated than it should be.

can someone pls explain to me like i'm 5 what $NOW actually does

@icanvardar The trend is modular monolith now. Microservices introduce inter-service API latency, operational complexity, increased Cloud costs and wasted CPU cycles due to overhread burden of distributed systems.

monolith was always the best choice against microservices and deep down everyone knows it

@ivanfioravanti Priced an 128 GB MacBook pro 2TB yesterday. Max Cpu/gpu with nano texture. Went up about $1300 CAD from what I remember.

MacBook Pro 16" with max specs and 3 years Apple Care+ in Italy is now ~12.000 Euro or ~14000 USD 👀

If there is such a demand, why not sell them at a premium to the local Chinese AI providers? It does not make a lot of sense.

@FirstSquawk Qwen must be getting too much attention

ANTHROPIC ACCUSES ALIBABA OF USING 25,000 FRAUDULENT ACCOUNTS TO PROBE CLAUDE AI MODELS Anthropic⁠ alleges that individuals linked to Alibaba Group⁠ used nearly 25,000 fraudulent accounts to extract information about Claude’s capabilities and safety systems, escalating tensions over AI model security and intellectual property protection.

@danielnewmanUV It will make more net profit than the stock price a year ago. You were buying a foward PE of 1 without knowing it.

$MU forward PE based on its new guide is now around 7. 😎 ‘But it’s gone up so much.’ 😆

@vulcanbench I like this repo. I'm applying against local opsen source models on Mac.

Did a deeper dive into my v1 benchmarking tasks with Opus, and found out why my first benchmark didn't go quite according to plan. For anyone interested in the analysis, and the fix, see screenshots below. Fixes applied in five phases, all now live on Github. Here's an overview of what was changed: - Removed 9 unsolvable Python scaffolds, re-spec'd 2 as anchors, added the spec gate (harness/spec_check.py + validator WARN + scripts/check_spec.py), fixed the root-cause generator - Pruned 12 zero-discrimination Double one-liners (kept 2 anchors/language) - Added 2 genuinely hard tasks: py-expr-eval (operator precedence) and go-parallel-map (concurrency, -race) - Discrimination report: per-task split + per-model-pair McNemar separation in vulcanbench report - Composition guards, curation-discipline docs, README updated

@DudeWhoInvests caught it perfectly overnight

They do call it Microsoft 365 😂