Whitepaper Reading: Local AI & On-prem Models
Local AI & On-prem Models
Running AI locally is accelerating because small, efficient models now deliver strong results without cloud latency, cost, or privacy trade-offs, making compact machines with unified memory ideal for inference and agent workflows; as a result, Apple hardware has become a default local-AI platform, with Mac minis and Mac Studios selling out due to demand outpacing supply.
Topics
Local Agents Don’t Need Scale — They Need Better Framework: effGen shows that ~1.5B models can match GPT-4 on agent tasks if the system is built for them. Tool use, memory, planning, and task split matter more than model size, making strong local agents practical today. https://arxiv.org/abs/2602.00887
Kimi K2.5 - Visual Agentic Intelligence Is the Next Step: moves beyond text-only agents toward systems that see, reason, and act in visual worlds. A useful reference for where multimodal agents are heading.
https://arxiv.org/pdf/2602.02276Mac Hardware Is Becoming a Real Inference Platform: Apple is pushing hard on on-device AI with M-series GPUs and private learning. Experiments like RDMA over Thunderbolt 5 already enable multi-Mac setups with 15TB+ shared VRAM. Mac Studios are sold out, and local model clusters on Apple hardware are closer than most think.
(i) https://machinelearning.apple.com/research/neurips-2025 (ii) https://machinelearning.apple.com/research/exploring-llms-mlx-m5, (iii) https://www.jeffgeerling.com/blog/2025/15-tb-vram-on-mac-studio-rdma-over-thunderbolt-5
Thank you Sky9 for Capital for offering us their office!
The Organizers:
AI+
AI+ is the premier community of 50,000+ AI founders, researchers, and builders. Founded in the San Francisco Bay Area, AI+ has held over 200 events and helps to drive adoption for some of the best AI companies worldwide.
AI+ also organizes AI+ Renaissance, the headline AI conference in San Francisco, with speakers including founders and executives from Parallel Web Systems, Wispr Flow, Replit, Windsurf, Neo4j, MindsDB, and 40 leading figures in AI.
Sky9 Capital
Sky9 Capital is a global VC with ~$2B AUM investing from early to growth stage in technically driven founders building category-defining companies. Sky9 has 35+ exits/IPOs with a strong focus on AI infrastructure, hardware, and deep tech.
Registered in Singapore with presence in the Middle East, North America and Asia, Sky9 has backed 150+ companies including TikTok, Moonshot AI, Rox Motors, and Webull, and acts as a long-term partner providing technical insight, global market access, and scaling support.
Whitepaper Reading Club
Whitepaper Reading Club is a community of 700 builders, founders, researchers, in Singapore, Malaysia, San Francisco, Bangkok, New York and now Lagos. We come together in person every month to discuss the latest blockchain projects (website)
Bittensor - discussion (Jul 2025)
AlpenGlow (Aug 2025)
