Now that details are emerging about Sam’s ouster I’ll resurface what I sent out last night. Sam’s an anomaly of a CEO in AI: he’s a classic 2010s era Silicon Valley popped-collar startup founder deeply entrenched in the Bay Area tech scene through YC.
@mattlynley
-
Entrepreneurial inspiration from successful tech business launch
By
–
What @hblodget did with BI is a big inspiration for why I tried to launch my own thing and it was such an awesome experience
-
ElevenLabs Achieves $1 Billion Valuation in Funding Round
By
–
Looks like @CallumBurroughs and @r1ddhi @TechInsider confirmed my scoop on ElevenLabs $1b+ valuation from last night. Supervised readers, you heard it first! 🙂
-
New Technology Platform Expands Beyond Gaming Applications
By
–
so, not just gaming, even though they are pitching hard on the gaming stuff
-
Apple GPU Shaders Enable ML Operations in New iPhone
By
–
For those following the GPU stuff with the new iPhone at the apple show its shaders that it's referring to are the same operators (mps) for its ML tasks.
-
Apple’s unified AI architecture strategy and development approach
By
–
Something something something unified architecture something https://t.co/oHpI8UZy6e https://t.co/HEGWFQubqa
— Matthew Lynley (@mattlynley) 7 septembre 2023Something something something unified architecture something https://
supervised.news/p/apples-chanc
e-to-remake-ai-development
… -
Nvidia CUDA Lock-In and Greedy Models Hardware Strategy
By
–
Covered some of this re: Nvidia’s (PyTorch) cuda lock-in, which is mostly for what I’d call “greedy models”—“low”-end hardware will be pretty good for a lot of inference (T4, A10, CPUs). Part of what makes Nvidia’s backing for PaxML interesting as well. https://
supervised.news/p/greedy-model
s-and-nvidias-open-source
… -
OpenAI Enterprise Tier: Business Model Strategy Emerges
By
–
With the release of fine-tuning for GPT 3.5-Turbo, and the announcement of a new Enterprise tier, its business model is starting to take shape. While they're positioning it around control/privacy/etc, this is really about rate limiting and reliability.