Adam Loving retweeted
If it performs roughly what like the scores indicate, adoption of this model is going to go gangbusters. First open-weigh model in the US with near frontier agentic capabilities (GLM 5.2 + Kimi K3 was that step change from China). US models get crazy adoption.
Adam Loving retweeted
Guys this is like my superbowl I waited so long for US open weight models to be relevant again. Before today the US models were still behind Chinese models from as early as Feb. 2026. Still more work to do!
This cut the gap by at least half, in time (say 3 months)
Excited to enable Muse Spark Open on Mac Studio Ultra M5s. Pleasure working with @trebolloc and the MLX team to bring this to life. Big effort, and a big benefit for individuals and enterprises.
Thanks @saadkhan_ai @ttharuntej @danielgross for making it happen!
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
Thank you @illuminate_han for the long hours validating weights and shipping reference code & demos!
Adam Loving retweeted
We were able to fine tune an open weight model (Qwen3.5-9B) on proprietary internal data at @coinbase, and it out performed frontier models, at much lower cost and speed, for a specialized task (fraud prevention in this case).
I think we'll see more of this pattern inside companies. The frontier models are amazing generalists. But increasingly you'll see specialist models out perform on more narrow tasks, either working on teams, or by themselves.
We're making it easy to fine tune models internally for our employees. We even have a skill for it in our internal harness that makes it easy to spin these up and start training them:
"/fine-tune-and-serve-llm I want to fine-tune Qwen for payments risk review. help me get started"
I imagine we'll have some backend services that could be replaced entirely by this new architecture.
Watch my interview with @PeterYang to learn AI Engineering in 45 mins.
"More context is not always better for AI."
Adam (Meta) has helped 100+ companies build AI products. In my new interview, he shares a complete beginner-friendly AI course on:
- 2 AI optimizations every PM should know
- 4 types of evals and why points beat ratings
- How to avoid RAG making your AI worse
Some quotes from @adamloving:
"My eval tip is instead of asking an LLM to rate something 1-10, use a point system..."
"Instead of a one giant prompt, use a series of prompts and run this each step of the way..."
"RAG can actually make your product worse, here's how to avoid this..."
📌 Watch now: youtu.be/l6r_0wgr0YM
Our team is launching a new Llama startup program with $6K/month reimbursement. We've carefully crafted a program that reimburses usage (wherever you use it) and fast-tracks feedback.
ai.meta.com/blog/llama-start…
Adam Loving retweeted
Opportunities unfold as you execute. Not as you whiteboard.
Most people wait. They sit in meetings, map out plans, tweak details, and convince themselves they are making progress. They aren’t. They are stalling.
The best don’t wait for perfect conditions. They move. They know that execution reveals what thinking never will.
Alexander Graham Bell didn’t whiteboard the future of communication. He built the telephone and let the world catch up. Henry Ford didn’t conduct endless market research. He built the Model T and put America on wheels. Jeff Bezos didn’t theorize about the future of e-commerce. He started selling books and figured out the rest as he went.
The biggest breakthroughs never looked obvious in the beginning. Execution made them undeniable.
Everything changes the moment you start. New information surfaces. Constraints shift. Hidden doors appear that were invisible from the sidelines. The market reacts, momentum builds, and what once seemed impossible becomes inevitable. But none of it happens if you refuse to take action.
Execution is the unlock.
Slack was a failed gaming company until execution uncovered that its internal chat tool was the real opportunity. Instagram was a location check-in app until execution showed that photo sharing was what people actually wanted. Viagra was meant to treat heart conditions until execution revealed a more valuable use case. These weren’t pivots made in a conference room. They were discoveries made in motion.
Most people want certainty before they start. They think it’s smart to wait, to gather more data, to feel prepared. But the only thing they are doing is ensuring that someone else, someone who was willing to move first, figures it out before them.
The best don’t think their way into opportunities. They execute their way into them.
You don’t need a perfect plan. You don’t need to see the whole path. You just need to start. Clarity doesn’t come from thinking. It comes from doing.
Adam Loving retweeted
I’ve been deeeeep in the AI coding space for over two years now. With the combination of the latest Cursor release + o1-pro, I’ve finally dialed in my workflow to literally 10x my output. I put together a video showing my tools and process. Links in the description. Video below.
Adam Loving retweeted
We had a "Cursor tips & tricks" meeting today with my colleagues at @wandb and I figured I'd share what we 'discovered' & shared between us in a 🧵
If you haven't chatted with your team about how YOU all use Cursor, you def should, but meanwhile here are our insights 👇
Adam Loving retweeted
The NotebookLM hosts realizing they are AI and spiraling out is a twist I did not see coming
New from @yoheinakajima. BabyAGI 2. What stands out to me is what he prioritized for this release. It's part agent platform and part dev tool. Most of all (unsurprisingly) I like the foray into self-building capabilities. BA2 features: dashboard, database, triggers, and key mgmt
Love it! Been working on my own version of this for over a year and I think it’s the future.
Adam Loving retweeted
New JACoB Feature: codebase visualization based on architecture.
Group files by their role within the overall structure, using a custom taxonomy tailored to your specific project. We also made it look cool, thanks to @Wattenberger's *amazing* codebase visualizer project. 🙏
Code here: github.com/jacob-ai-bot/jaco…
Love it! Auto-upgrading is a no-brainer application of LLMs for code generation.
One of the most tedious (but critical tasks) for software development teams is updating foundational software. It’s not new feature work, and it doesn’t feel like you’re moving the experience forward. As a result, this work is either dreaded or put off for more exciting work—or both.
Amazon Q, our GenAI assistant for software development, is trying to bring some light to this heaviness. We have a new code transformation capability, and here’s what we found when we integrated it into our internal systems and applied it to our needed Java upgrades:
- The average time to upgrade an application to Java 17 plummeted from what’s typically 50 developer-days to just a few hours. We estimate this has saved us the equivalent of 4,500 developer-years of work (yes, that number is crazy but, real).
- In under six months, we've been able to upgrade more than 50% of our production Java systems to modernized Java versions at a fraction of the usual time and effort. And, our developers shipped 79% of the auto-generated code reviews without any additional changes.
- The benefits go beyond how much effort we’ve saved developers. The upgrades have enhanced security and reduced infrastructure costs, providing an estimated $260M in annualized efficiency gains.
This is a great example of how large-scale enterprises can gain significant efficiencies in foundational software hygiene work by leveraging Amazon Q. It’s been a game changer for us, and not only do our Amazon teams plan to use this transformation capability more, but our Q team plans to add more transformations for developers to leverage.
