Alex Gaziev

I’ve collected all the snapshots of the products we studied in one place on @playbook_hq — a very cool Dropbox alternative for creatives (and one of our clients). Feel free to explore and grab anything that inspires you. 👉 playbook.com/s/evilmartians…

.@AnthropicAI will support stateless remote MCP servers with “just HTTP” as transport. We’re quite excited about this! So we gave a “lightning talk” at an MCP meetup hosted by @paulgb @CloudflareDev yesterday.

Demo of plasma stellarator filmed at CERN on an employee’s phone.

Mechazilla has caught the Super Heavy booster!

a video player that lets you browse through faces

lmao don't use bezier curves to smooth out your bandwidth graph

Introducing Willow, our new state-of-the-art quantum computing chip with a breakthrough that can reduce errors exponentially as we scale up using more qubits, cracking a 30-year challenge in the field. In benchmark tests, Willow solved a standard computation in <5 mins that would take a leading supercomputer over 10^25 years, far beyond the age of the universe(!).

Who knows how to resolve regional ISP issues with DNS? I've got some regions where folks are seeing a DNS_PROBE_FINISHED_NXDOMAIN. This seems like an area where it's hard to find experts!

The paradox of poetry and information theory

in a couple weeks, i built a nuclear fusor in my bedroom – with zero hardware experience the secret? Claude sonnet 3.5 + projects a glimpse into the process below

If each step of an ai agent is 95% accurate. None of the 30 step work flows will work. Going from 95-> 99.9 is a similar last mile problem as with self driving cars. Easy to hack up a prototype. Hard to make it work reliably at scale.

What have you always wanted to create? Bring your vision to life in 30 days. We’ll fund it. At Playbook, our mission is simple: to help creatives be more creative. That’s why we’re launching the $100,000 Playbook Imagination Fund –– to let your creativity fly without financial worries getting in the way. Whether it’s a web archive of typographies or re-branding your local coffee shop, we’re here to support your most original and impactful ideas. Read more here (playbook.com/playbook-imagi…).

Cycle3D High-quality and Consistent Image-to-3D Generation via Generation-Reconstruction Cycle Recent 3D large reconstruction models typically employ a two-stage process, including first generate multi-view images by a multi-view diffusion model, and then utilize a feed-forward model to reconstruct images to 3D content.However, multi-view diffusion models often produce low-quality and inconsistent images, adversely affecting the quality of the final 3D reconstruction. To address this issue, we propose a unified 3D generation framework called Cycle3D, which cyclically utilizes a 2D diffusion-based generation module and a feed-forward 3D reconstruction module during the multi-step diffusion process. Concretely, 2D diffusion model is applied for generating high-quality texture, and the reconstruction model guarantees multi-view consistency.Moreover, 2D diffusion model can further control the generated content and inject reference-view information for unseen views, thereby enhancing the diversity and texture consistency of 3D generation during the denoising process. Extensive experiments demonstrate the superior ability of our method to create 3D content with high-quality and consistency compared with state-of-the-art baselines.

Evaluate LLMs in real-time with Street Fighter 3 🔥 Quite an interesting and creative approach to evaluate the quality of LLMs, by making them ‘fight’ each other in Street Fighter 3 in real-time‼️ (h/t @datitran) LLM-colosseum: github.com/OpenGenerative…

Presenting a novel approach that harnesses generative text-to-image models to enable users to precisely edit specific material properties (like roughness and transparency) of objects in images while retaining their original shape. Learn more → goo.gle/4deVgj5

Jagged Intelligence The word I came up with to describe the (strange, unintuitive) fact that state of the art LLMs can both perform extremely impressive tasks (e.g. solve complex math problems) while simultaneously struggle with some very dumb problems. E.g. example from two days ago - which number is bigger, 9.11 or 9.9? Wrong. x.com/karpathy/statu… or failing to play tic-tac-toe: making non-sensical decisions: x.com/polynoamial/st… or another common example, failing to count, e.g. the number of times the letter "r" occurs in the word "barrier", ChatGPT-4o claims it's 2: x.com/karpathy/statu… The same is true in other modalities. State of the art LLMs can reasonably identify thousands of species of dogs or flowers, but e.g. can't tell if two circles overlap: x.com/fly51fly/statu… Jagged Intelligence. Some things work extremely well (by human standards) while some things fail catastrophically (again by human standards), and it's not always obvious which is which, though you can develop a bit of intuition over time. Different from humans, where a lot of knowledge and problem solving capabilities are all highly correlated and improve linearly all together, from birth to adulthood. Personally I think these are not fundamental issues. They demand more work across the stack, including not just scaling. The big one I think is the present lack of "cognitive self-knowledge", which requires more sophisticated approaches in model post-training instead of the naive "imitate human labelers and make it big" solutions that have mostly gotten us this far. For an example of what I'm talking about, see Llama 3.1 paper section on mitigating hallucinations: x.com/karpathy/statu… For now, this is something to be aware of, especially in production settings. Use LLMs for the tasks they are good at but be on a lookout for jagged edges, and keep a human in the loop.

Cats Helping With the Dishes, Kling AI by Pets Pixels Studio

This is art.

The future that is happening. Automation powered by GPT-4o generates Figma designs based on PRD. @figma

pinch to draw