Keep on to blur preview images; turn off to show them clearly
Beautiful essay by Andrew: Key quotes: "If open-source/open-weight models continue to grow in market share, then some companies that are pouring billions into training models might not see an attractive financial return on their investment. Additionally, algorithmic and hardware improvements are making it cheaper each year to train models of a given level of capability, so the “technology moat” for training frontier models is weak."
Artificial Intelligence @amazon. RL, General Purpose Agents, OSS AI All views personal!
It's remarkable how Jiang&Hu era is getting retroactively glorified because Xi Bad. I remember vaguely what the discourse was like back then. The coming collapse, the inevitable middle income trap, pollution, suicide nets, one child policy… But I guess it felt nice for Noah.
We're in a race. It's not USA vs China but humans and AGIs vs ape power centralization. @deepseek_ai stan #1, 2023–Deep Time «C’est la guerre.» ®1
RT @theo: Why are people overreacting to this so badly? “This was created by monkeys” is a silly way of saying “I can’t fathom a human mak…
🇦🇺 Co-founder: @AnswerDotAI & @FastDotAI ; Prev: professor @ UQ; Stanford fellow; @kaggle president; @fastmail/@enlitic/etc founder https://t.co/16UBFTX7mo
The Last Night has always been designed to seriously map out the cultural war coming this century. What I failed to predict was that the anti-AI side would turn out so caricatural. We wrote it much more philosophical & reasonable.
Founder @oddtalesgames Directing The Last Night @TLN_Game Art Direction, Cinematography, Tech Art. Atoms, Bits, Memes, Genes. Freedom, Futurism, Humanism.
What a wonderful podcast to listen to on a beautiful Friday morning!!! Building the AI Hardware Engineer with Matthias Wagner, Co-founder of Flux As someone who studied Electronics and Communications in 2000s, this all sounds so magical. It used to take insane amount of effort to design something that is manufacturable. Matthias mentions things are still very slow and he is determined to change it. May be "Devin for Hardware" moment is here. We are moving from manually routing wires to prompting smartphones into existence. Below is a quick summary: 1/ The Stagnation Software tooling has evolved insanely fast (VS Code, Cursor, Copilot). Meanwhile, hardware design tools have been stuck in the 90s. Matthias Wagner, co-founder of Flux, realized this gap: The supply chain is ready for anything, but the design tools are bottling us up. 2/ The New Paradigm: FluxFlux is doing for electronics what Figma did for design—but with a native AI engine. It’s not just a browser-based CAD tool. It’s an agentic environment where you collaborate with AI to build physical things. 3/ From Search to Action Previously, 50% of hardware engineering was Googling datasheets and checking DigiKey for stock. Flux’s AI agent handles this. You ask for a "cheaper replacement for this chip," and it checks real-time inventory, pricing, and pin compatibility in seconds. 4/ Prompt to Product The real magic happens when you move from tasks to projects. Matthias demos this: "Design a smart speaker with a screen, battery, and wake-word detection." The AI doesn't just chat; it executes. a) Break down the spec b) Find components (ESP32, Amps) c) Wire the schematic 5/ The "Grand Vision" This is the killer insight: "If you can prompt a poem into existence with ChatGPT, why can't you prompt hardware into existence?" Matthias isn't just talking about PCBs. He means the entire device. 6/ Build your own hardware: "Why go to Amazon and search for 2 hours... if you can just describe what you want and get it made and sent to you within a week?" We are approaching a future where bespoke hardware is as accessible as ordering a book. 7/ The Economics of One We’ve been trained to believe cheap = mass produced. AI flips this math. If the design cost drops to near-zero (because an AI agent did it), you only pay for materials. Custom, one-off electronics become economically viable for the first time in history. 8/ Who is this for? It’s not just for making a custom iPhone (yet). It’s the hidden layer of the world: A) Vending machine controllers B) Custom traffic lights C) Farm automation tools These are the massive industries waiting to be revolutionised by "Prompt-to-PCB." 9/ The Future We are witnessing the birth of the AI Hardware Engineer. Just as AI is changing who can write code, it is about to change who can build the physical world around us. 10/ Dive Deeper Flux is bridging the gap between software speed and hardware reality. You can signup and try. (Summary generated with Gemini. I can not summarise it better than this. Thanks @demishassabis ) Check out the full podcast here
Artificial Intelligence @amazon. RL, General Purpose Agents, OSS AI All views personal!
Localization is an underhyped feature of image edit models. Now, an image can look like it was taken anywhere in the world. I uploaded the Notion NYC photo and asked Nano Banana Pro: “fully localize this to Tokyo, Japan” It changed the language, people, and even the stoplight!
Partner @a16z AI 🤖 and twin to @omooretweets | Investor in @elevenlabsio, @krea_ai, @bfl_ml, @hedra_labs, @wabi, @WaveFormsAI, @ViggleAI, & more
