"Dating a model" can mean one of two very different things, depending where you live in California

Co-sign.

Self recommending must-read for Canadians. “The path Canada is on, economically and culturally, is no longer sufficient to make us a flourishing world class nation.”

Just remembered the story about a computer scientist who had his bike stolen and tried to explain binary search to a cop

Memories.ai is building the visual memory layer for wearables and robotics techcrunch.com/2026/03/16/me…

🚨 BREAKING: A Google researcher and a Turing Award winner just published a paper that exposes the real crisis in AI. It's not training. It's inference. And the hardware we're using was never designed for it. The paper is by Xiaoyu Ma and David Patterson. Accepted by IEEE Computer, 2026. No hype. No product launch. Just a cold breakdown of why serving LLMs is fundamentally broken at the hardware level. The core argument is brutal: → GPU FLOPS grew 80X from 2012 to 2022 → Memory bandwidth grew only 17X in that same period → HBM costs per GB are going UP, not down → The Decode phase is memory-bound, not compute-bound → We're building inference on chips designed for training Here's the wildest part: OpenAI lost roughly $5B on $3.7B in revenue. The bottleneck isn't model quality. It's the cost of serving every single token to every single user. Inference is bleeding these companies dry. And five trends are making it worse simultaneously: → MoE models like DeepSeek-V3 with 256 experts exploding memory → Reasoning models generating massive thought chains before answering → Multimodal inputs (image, audio, video) dwarfing text → Long-context windows straining KV caches → RAG pipelines injecting more context per request Their four proposed hardware shifts: → High Bandwidth Flash: 512GB stacks at HBM-level bandwidth, 10X more memory per node → Processing-Near-Memory: logic dies placed next to memory, not on the same chip → 3D Memory-Logic Stacking: vertical connections delivering 2-3X lower power than HBM → Low-Latency Interconnect: fewer hops, in-network compute, SRAM packet buffers Companies that tried SRAM-only chips like Cerebras and Groq already failed and had to add DRAM back. This paper doesn't sell a product. It maps the entire hardware bottleneck and says: the industry is solving the wrong problem. Paper dropped January 2026. Link in the first comment 👇

I can't believe humans were gifted a planet full of famine, disease, and death and invented fusion, vaccines, moon rockets, and so much food that most people die from obesity.

Compliance shouldn't cost $50k. Today we're launching Universal Basic Compliance — a free foundation for companies that need to prove they’re trustworthy. • Compliance discovery • Contract requirement analysis • SOC 2 controls library • Trust Center • Cyber protection All free. Try it: koop.ai

Introducing Starlight Air, the world lightest space solar panel. Order today at terawatt.space/ Engineered by Starpath in California.

The software industry is apparently dying but job postings for software engineers are rapidly rising!

your timeline convinced you AI is in a bubble. talk to a boomer above the age 35 for 5 minutes. most people don’t even know what claude is.​​​​​​​​​​​​​​​​ kind of wild when you zoom out.

Careers are dead. Jobs are dying. Opportunities arising.

Project LUCI = AI Pin Real-world memories 🦞 We're turning AI from something you prompt into an assistant that actually knows you. Powered by Memories AI for visual memory and OpenClaw for real-world action. Join the waitlist.

The universe gives us a lot to work with…we will never run out of thought-provoking content to share.

People who say “we don’t want t-shirt factories in America” are mistaken about the very nature of technology. What they mean is that we don’t want Capex intense, heavily labor dependent, low margin manufacturing. But all of those attributes—capex, labor use, margin—are subject to technology, and therefore innovation. By exporting the industry, you are excluding the possibility of American technologists changing the *way t-shirts are made* in such a way that it is very profitable to do here. These sorts of changes are the inevitable direction of technology. Why would you count our civilization out of that?

1/7 We needed power for building gas stations on the Moon and Mars. Suppliers quoted us $100/W for solar and 14 months lead times. So we built our own. Space-rated solar, scalable, and 90% cheaper. Now it powers our mission. And yours. Launching today 🧵

Quote RT this tweet, get 1k free credits to use @memories_ai API, sign up fot the hackathon here: luma.com/hp67t6tf

Our portfolio company Memories.ai, now publicly launching their API to enable the next generation of multi-modal AI agents. Memories.ai is building the visual memory layer for AI, enabling machines to see, remember, and understand like humans. With ultra-low hallucinations and an effectively UNLIMITED CONTEXT WINDOW, their video understanding capabilities surpass even leading LLMs. This unlocks intuitive interfaces where AI agents can process video, audio, and visual inputs seamlessly—transforming industries from productivity to video editing. Explore the API: memories.ai/docs VC-backed founders can access up to $10,000 in credits.

I’m Shawn, founder of Memories.ai, former researcher at Meta and CS PhD at University of Cambridge. Today we’re launching Memories.ai: we built the world’s first Large Visual Memory Model - to give AI human-like visual memories. Why visual memory? AI to date is chat-based, which has great applications, but humans are not only chat based. Humans have visual-memories, and are visually driven. Memories.ai enables AI agents, softwares, and even robots to see and remember the world the way humans do. We just raised an $8m seed round led by @susaventures to build the visual-memory layer for AI. Why is Memories.ai so groundbreaking? 👇

We’re thrilled to announce that @memories_ai raised $8M led by @SusaVentures to build the foundational visual memory layer for AI — empowering machines to see and remember like humans… only better. And we’re headlined in TechCrunch today!