nolan

Stripe, MIT CS

[nolan @nolangclement]

@samgoodwin89 @peteredm1 @foldkit @alchemy_run https://t.co/wpWoUZ5en9

[nolan @nolangclement]

Big changes to my @foldkit + @alchemy_run starter repo. It's now a live chat app. Cloudflare workers / DO powering the messaging and chats are persisted to Neon, BetterAuth integration too

Impressed with how it feels to build with agents in this stack, feels unsloppable

[nolan @nolangclement]

@samgoodwin89 @peteredm1 @foldkit @alchemy_run Done, keeping Neon as default since its a more seamless experience to deploy without setting up billing, but swapping is now a 1 line diff (provide a different layer)

[sam @samgoodwin89]

It's Effect all the way down. Foldkit on top of Alchemy:

Frontend -> Backend -> Infrastructure

[nolan @nolangclement]

@peteredm1 @foldkit @alchemy_run Definitely considering it, both work with Hyperdrive so it’s super simple to swap in. Thanks!

[nolan @nolangclement]

https://t.co/7KV1hdAQ4g

[nolan @nolangclement]

Here's something I've been working on: a WIP starter repo for building apps with @foldkit and @alchemy_run. Effect end-to-end

https://t.co/2DWAOCskhy

[Artificial Analysis @ArtificialAnlys]

Jamba model launch takeaways - Potentially a new leader for ultra-long prompt use-cases (RAG)

‣ First open-source model of this size to combine MAMBA state-space model architecture, Mixture-Of-Experts (MOE) and the transformer
‣ 256k context window, more than 2X the size of the next largest open-source model (Code Llama 70B's 100k) we measure
‣ High expected throughput tokens/s as with its MOE architecture, 12B of its 52B parameters are active at inference. For shorter prompts, expect faster than Grok-1 & Llama 2 but slower than Mixtral 8x7B
‣ Presents a potentially very attractive offer for long input token prompts / RAG as throughput scales with input token size due to MAMBA architecture. A21 declare 3X Mixtral 8x7B's tokens/s at 128k context window lengths

Congratulations @AI21Labs. We look forward to benchmarking these declared speeds, particularly over long input token lengths 👀

[nolan @nolangclement]

@cwolferesearch @galipaydin Thank you!

[nolan @nolangclement]

@U71161612580151 Next house behavior

[nolan @nolangclement]

This would literally destroy all standardized tests….

Maybe that’s a good thing

[Andriy Mulyar @andriy_mulyar]

Elite hackers have gotten gpt4all to run on a ti-84 calculator. AP calculus exams will never be the same again.

https://t.co/xMWVTjaT5h

[nolan @nolangclement]

@cory Hi Corey! I’ll be at MIT would love to meet up.

[nolan @nolangclement]

@ernestobrivera Yo

[nolan @nolangclement]

@YKouloumbis Let’s go!
https://t.co/ARQKHrbKVY

[Yanni Kouloumbis @YKouloumbis]

Launching @OpenAI Shap-e support on @openrouter and @windowai! Generate high quality 3D assets in < 30 seconds, and build sustainable apps where users can pay for their own usage.

[nolan @nolangclement]

Finally, support the demo launch on Product Hunt: https://t.co/lVE5XwNtEU

[nolan @nolangclement]

Introducing: 3D model generation in your browser with @openrouter

Generate high quality assets from a text prompt in under 30 seconds with OpenAI's shap-e

Try it out now for free here: https://t.co/pv5ENbXToD

Credit @YKouloumbis for building the infra

[nolan @nolangclement]

In addition to 3D model generation, https://t.co/CiMRnVbEDF supports many of the most popular LLMs, with image model support coming soon.

Join the discord (https://t.co/SvvUXpVrIq) and contribute to a growing ecosystem of self sufficient, keyless, model agnostic AI apps.