Arie Meir, PhD x-Google AI Product

With RL techniques like GRPO, it’s become easy to fine-tune small models to outperform foundation models on vertical tasks. We’re open sourcing Osmosis-Apply-1.7B: a small model that merges code (similar to Cursor’s instant apply) better than foundation models. Links in comments! Using foundation models for high specificity, low complexity tasks like code merging can be overkill - instead, it’s better to use a specialized model instead for higher performance and latency. We fine-tuned the model on 100K real-world code commits where the model needs to merge a code edit snippet into an original code snippet. On the test dataset (subset of CommitPackFT), foundation model performance ranged between 0.77-0.93 reward score. In comparison, Osmosis-Apply-1.7B achieved a 0.98 reward score - while also being 3X-5X more affordable than the next cheaper model AND ~10X faster! Why this matters: the best AI agents use multiple models to ensure task specialization and reliability at scale - on top of cost and latency benefits as well. Links to the model and full write-up (with reward function, hyperparameters, etc.) are in the comments - reach out if you’re interested in using reinforcement learning!

240

18 Comments

Voice is a super interesting modality right now - maybe the first modality we're seeing move to open source models across a number of scale ups / enterprises. Reliability concerns, high costs, and open source model performance are pushing engineers to do their own fine tuning vs. relying on third-party vendors of proprietary models. Many of these orgs have already been collecting their own first-party data and now with third-party vendors like Extrian, David AI, etc they can train really high quality models. RL has been insanely hyped, but it's been unclear how long it will take scale ups and enterprises to actually lean in. Voice AI might be hitting that inflection point faster than expected.

45

7 Comments

"Build the event you wish existed" That's what we at Monadical did last week at #NYTW. We wanted an AI Engineers discussion for Engineers. There's always a lot of events out there for VCs to network, or for startups to learn about G2M, but very little on lessons learned from actual engineers in the field. So that's the event we hosted. Big thank yous to our awesome panel, Roy Pereira, Ben Cohen and Corey J. Gallon. Here's the key takeaways and the AI tools we're using. 🚀 All three panelists independently called AI Agents the most transformative LLM application they’ve used. They specifically called out Claude 3.5 Sonnet for its accuracy and reliability. 🪨 We dug into how LLMs are “jagged”, not general. They can be shockingly good at some tasks and completely fail at others. Everyone agreed: good evaluations are critical (and hard.) 🧪 Corey noted how public benchmarks and reality are two different things. Most public evals are saturated or gamed. ♊ Ben emphasized that AI projects are actually two projects: building the tool and building the evaluation process. 🧱 We explored how falling dev costs may impact startup defensibility and labor demand. Roy shared that founders are already shifting strategies in response. ⚒️ In a world of daily AI launches, the panel discussed how they decide what’s worth attention, and what’s just noise. They called out tools like Goose, Aider, Claude Code, and Monadical’s own Cubbi, which helps run agentic workflows safely in dev environments. (links in the comments). CTA: What would you want to hear in an AI Applied Engineering talk you attended?

33

5 Comments

I’m thrilled to share that White Star Capital is backing AMI - Advanced Machine Intelligence in their historic $1.03bn seed round 🚀 - Led by the legendary Yann LeCun and Alex LeBrun. AMI isn't building another LLM, they're building World Models. Why this matters: *Beyond Text: Grounded in JEPA architecture to solve physical-world problems. *Global Powerhouse: A dream team from Meta FAIR, OpenAI, and DeepMind. *Sovereign Tech: A credible European-led alternative to US Big Tech. Excited to be part of this journey along with the rest of the WSC team Eric Martineau-Fortin, Matthieu Lattes, Bérénice Moustial and Victoria Pozzi Rocco Belforti. 🌍✨ https://lnkd.in/eiCKZKSQ #VentureCapital #AI

86

2 Comments

AI isn't just powering software. BrightAI just raised $51 million to automate inspection and other hardware-based AI capabilities. As Alexa von Tobel, co-founder of Inspired Capital put it: “So many people are so focused on the future of digital AI, but we’re excited about this new layer of AI: the physical world AI.” We are in the early innings of a series of AI-fueled innovations! https://lnkd.in/enhmafvH

47

I’ve been working on the theory of why small, shippable factories are the perfect inversion to the centralized prefab factory model. It all collapsed into one line of math (credit to Gilles Retsin): Factory Efficiency = Usage × (Throughput × Product Value ) / ( CapEx + Deployment OpEx) The idea is that the closer you can drive Usage to 100 % while keeping the denominator tiny, the faster the flywheel spins.  A shippable microfactory tackles this by primarily focusing on a high Usage factor. Because the microfactory can be packed up afterwards and reused, its CapEx gets amortized over multiple projects, effectively raising its overall Usage across the year. If one project alone doesn’t fully occupy the factory, it can simply roll over to another job to maintain high Usage. Its CapEx is often an order of magnitude lower than a centralized factory (< $1M vs. $10-50M min), so the breakeven throughput is more achievable on a small pipeline of work. The real breakthrough here is portability. There is near-zero stranded capital and almost no idle time. To flesh this theory out, I wrote 3000 words on the topic for Brad Hargreaves and Thesis Driven. The article breaks down: - why “fixed factories, shipped goods” is being inverted to shipped factories, fixed goods - what a sub‑$1 M robotic cell does to CapEx per home versus a £45 M off‑site plant - early production/cost data on a shippable microfactory from Mollie Claypool, Gilles Retsin, Sam Baker and the Automated Architecture (AUAR) team - the hybrid reality of robots tackling the heavy structural components and human crews assembling faster and cheaper - technical constraints to pull this model off and future considerations on the optimal business model (own vs. rent) Major shoutout to Gilles Retsin for helping me pull this together. https://lnkd.in/eRTnbZPa

54

2 Comments

We’re thrilled to share that Usul (YC S24) has raised $3.3M in Seed funding led by Scout Ventures with participation from BVVC, Teamworthy Ventures, True Ventures, Y Combinator and more. Usul is the modern platform for defense contract capture and federal growth. Our customers have won hundreds of millions in defense contracts for capabilities across Air, Land, Maritime, AI/ML and more. When Oliver Gomez, Joonghyun Lee and I joined forces during Stanford’s Hacking 4 Defense class, we interviewed 300+ defense companies to intimately understand how difficult it is for national security solutions to get deployed quickly at scale. In normal capital markets this is called dead weight loss. In warfighting, this costs lives. We're building an elite team of hackers and defense enthusiasts in San Francisco. If you’re interested in powering the next 1000 defense primes, email us at founders@usul.com or dm me. Read more about our journey in Erin O'Brien's Tectonic article below in the comments. Lastly, thank you to our amazing cap table for joining the ship. Bradley C. Harrison, Cody H., Nate George, Zach Beecher, Michael Keane, Joe Musselman, Gene Ebersole, Nick Nowlin, Thomas Lehrman, Stephen Schmalhofer, Michael Montano, Gustaf Alströmer 🇺🇦, Jared Friedman, Steve Blank, Jacqueline Tame, Jack Shanahan, Peter Newell, Steve Weinstein, Jeff Decker, Ivan Kirigin, Jeff Heitzman, Keith W. Gibson, CFCM, SCPM, John Tincoff, Michael Malis, Austen Allred, Tom Hsieh, Donald Donckers

448

77 Comments

Honored and excited to partner with Drew Baglino and the incredible team at Heron Power as they re-imagine the electric grid for the modern world. Power in America is not zero sum. We deserve access to cheap and abundant electricity and we should win the AI race. And we must build the capacity to manufacture critical technology here at home. When faced with existential challenges, we deploy technology against them. And we win.

54

1 Comment

Building for the nation’s most critical technologies means rethinking how we approach human repair - from the battlefield to the lab bench. Harpoon is proud to back Somite AI in their $47M Series A, led by Khosla Ventures. Somite is developing the AI blueprints for functional human cells - combining cell therapy, large-scale simulation, and a novel foundation model for biology. Their platform represents a bold leap forward in programmable cell manufacturing, with the potential to unlock next-generation treatments for organ failure, degenerative disease, and more. We’re excited to support Micha Y. Breakstone 🇮🇱🎗️ and the Somite team as they push the frontier of regenerative medicine and build the infrastructure for truly scalable cell replacement therapies.

76

5 Comments