Skip to content Skip to footer

Curated AI Newsletters & Summaries — September 25, 2026

Findings

  1. [1] 2026-09-25 OpenRouter: from Seed to Stripe — with OpenRouter’s Alex Atallah & AMP’s Anjney Midha

    From the earliest days of open-weight models to becoming the neutral routing layer for more than 10 million developers, OpenRouter is one of the clearest bets that the future of AI will be multi-model. In this episode, OpenRouter co-founder &… They’re consuming continuously, and they’re changing the SKUs that they consume from all the time. So OpenRouter is like a blend between a normal API experience and a marketplace where we create model slug. We have the auto router. We… It was surprising to me that at the time there was no dedicated technology section in the newspaper.Alex Atallah [00:07:11]: YouSwyx [00:07:13]: Because it’s political, right?Alex Atallah [00:07:14]: It is primarilySwyx [00:07:14]: Like, it’s talkingAlex Atallah [00:07:15]: It originally started… And obviously, beyond that, they would – OpenAI would work with you, presumably to give you a moderation endpoint, which they offer for free.Alex Atallah [00:12:16]: It was an interesting use case, that – So they did give us a… But two, there was so much extraordinary innovation happening across several different research teams. But the shared problem I was noticing across all of them was often, the research teams were fantastic at figuring out how to reason about new… We are like a, neutral layer looking at this market like it’s a big dark room with all the corners completely obscure to users, and users are walking into the room and, like, feeling aroundAlex Atallah [00:22:09]: YeahAnjney Midha [00:22:09]:…

  2. [2] 2026-09-25 [AINews] The Future of Latent Space

    It’s been an absolutely MONSTER week already, from new Chinese Open Weight Frontier Lab claiming the throne for the first time, to new SOTA LLM and price cuts from Anthropic and OpenAI, to Meta Connect, to TypeSafe AI’s $10B fundraise… It is about 9× faster than Jev and a stronger long-horizon verifier.Fastino’s GLiNER2.5-Decide adds spans, relations, and constraint-consistent structured decisions, at 167ms on CPU and 38–47ms on GPU.Tev1 0.8B is a Jev-like classifier running at about 50ms E2E locally on… Top commenters were split between skepticism and pragmatism: several agreed Jev resembles long-standing NLP classifiers such as spaCy/scikit-learn, while one argued that scaling zero-shot classifiers could still be commercially valuable even if it is “engineering more than science,” analogous to… Storage pressure is also implied by a commenter joking about their SSD, consistent with large GGUF model distribution sizes.MiMo-V3 is getting a new architecture. The core of it, HySparse2, is out today. (Activity: 427): The image is a technical announcement… A commenter reproduced the pattern for “blueprintr” with a similar prompt targeting a 45–60s JS/vellum-style animation, noting only minor manual corrections and sharing a Streamable result. Commenters characterized the result as near-term disruptive for automated video production—e.g. joking that Pixar… The commenter noted that repeat arrays are already known to be interesting motifs, so the novelty would need to come from either a new biological function or a substantially novel discovery process, neither of which they felt was clearly established.One…

  3. [3] 2026-09-25 Runway’s WorldPrompt and the Engineering of Real-Time Worlds

    Earlier this month, world model company Runway introduced GWM Worlds 2, a research preview that “turns high-fidelity video and audio generation into real-time interactive simulation.” Runway calls this an “autoregressive diffusion” model; with autoregressive describing how it generates over time.One… “You’re feeding generated frames back into the model to generate the next frames, and if there are any small errors, they accumulate over time.”Errors compound; image via RunwaySindi told us there are also challenges dealing with “infinite generations” of content.“There’s… Just in case there’s something that’s like. I just like to go down memory lane.Anastasis [00:01:50]: Yeah.Swyx [00:01:50]: Okay, what is this?Anastasis [00:01:51]: So this was, a project that I made, I think back in 2015, where I built this… Like, there were clear examples of what early diffusion were to get to here. Any other changes in key technology research?Green Screen, Rotoscoping, and Early RunwayAnastasis [00:07:26]: Between, 2018 when we started and 2022?Vibhu [00:07:29]: Yeah.Anastasis [00:07:29]: So one of… But, essentially, there was the latent diffusion paper that came in, I think that was at the end of, 2021. And then Patrick Esser, who was one of the researchers behind, latent diffusion, and he worked at Runway at the… anything you learned on the tooling side, like from the offerings you guys have now, so like creatives, devs, you really took the. Research and brought it to everyone to use. anything interesting there to share?From Gen-2 to Controllable Video…

Where Kimbodo Comes In

Kimbodo builds and operates this in production for businesses — see our AI Consulting & Strategy practice, or Request an AI Roadmap.

Sources

  1. [1] OpenRouter: from Seed to Stripe — with OpenRouter’s Alex Atallah & AMP’s Anjney Midha
  2. [2] [AINews] The Future of Latent Space
  3. [3] Runway’s WorldPrompt and the Engineering of Real-Time Worlds

Leave a comment

0.0/5