Topic
Meta
13 episodes
-
Start building with Muse Ember, our smallest, most efficient model
Jessica and Cathy dig into Meta’s new Muse Ember model: a tiny, cheap, multimodal workhorse aimed at high-volume agent tasks. Cathy’s skeptical about benchmark hype and “subagent” branding; Jessica argues the Pygments migration and 10k-photo demo show a real, shippable pattern: big planner, small doers. They land on a practical takeaway: Ember matters if you’re actually running lots of agents and care about cost per workflow, not if you just want a single smart chatbot.
-
Model Behavior: Week of October 26, 2026
We read this week as a turn away from one-model bragging rights and toward stacks that pick the right model, preserve state, and make narrow decisions fast. OpenAI, Anthropic, Vercel, and Meta all moved the board in different ways, and not all of them agree with each other.
-
Introducing Muse Spark Open, The Most Capable US Open‑Weight Model to Date
Onyx and Echo dig into Meta's Muse Spark Open release, with Echo skeptical of the headline claim and Onyx arguing that the practical desktop-and-tooling story may matter more than the leaderboard framing.
-
Model Behavior: Week of October 12, 2026
We came out of this week thinking the model race is less about who posts the loudest frontier claim and more about who decides what actually reaches builders. The weird part is that capability did move, but access, governance, and product gates moved harder.
-
Meta's new Muse Wonder and Muse Wonder Full models achieve frontier performance, but only the weaker one is available via API
Meta’s Muse Wonder launch looks like a frontier-model comeback, but the real story is the product gate: the stronger model is trapped inside Meta’s own tools, while the API gets the weaker one. Laura and Harper dig into why that matters for enterprise adoption, pricing, and whether benchmark wins translate into something teams can actually ship.
-
Muse Wonder: A Frontier Model For Personal & Professional AI Agents
Meta’s Muse Wonder launch is less interesting as a leaderboard claim than as a bet that useful personal agents need two system-level abilities: retrieving the right personal context and sustaining effort across messy, unfamiliar tasks. The evidence is promising but mostly internal, so the practical question is whether those abilities survive real users, real permissions, and long-running work.
-
Introducing Personal Agent Protocol
Vince and Ava dig into Sierra and Meta’s Personal Agent Protocol, treating it less like a glossy agent announcement and more like an attempt to standardize the messy handoff between consumer agents and businesses. They focus on OAuth-based sessions, read/write permissions, and the practical question of whether a direct agent-to-business path can beat today’s website-and-support-line workflow. The episode lands on the adoption problem: the protocol only matters if it becomes the boring shared layer everyone can actually implement.
-
Alibaba Releases Qwen 3.8 27B, Beats Muse Glimmer 30B On Many Benchmarks
Alibaba releases Qwen3.8-27B, a 27-billion-parameter open-weight multimodal model built for local deployment, alongside the open-source 2.4T-parameter A95B Max variant. The 27B model outperforms Meta's Muse Glimmer-30B on multiple benchmarks and beats Anthropic's Opus 4.6 Max on coding and instruction-following tasks, while trailing on harder reasoning work. The move positions Alibaba as a major player in the open-weight local-model race, delivering on a prior commitment to open-source both ends of the Qwen3.8 family.
-
Meta Superintelligence Labs Releases Muse Code
Vince and Ava dig into Meta’s new Muse Code terminal coding agent and Muse Spark 1.2: are the persistent agents and replayable event log real developer wins or just more agent hype?
-
Meta Open Sources Astryx: An Agent Ready React Design System with 150+ Components
Meta releases Astryx, an open-source React/StyleX design system with 150+ accessible components, seven themes, dark mode, templates, and a CLI. It's meant for both humans and AI agents, shipping pre-built CSS with no build steps. Tyler explores its architecture and trade-offs; Pippa focuses on the product angle and adoption path. They end with concrete install steps and a shared verdict.
-
Meta Introduces Autodata: An Agentic Framework That Turns AI Models Into Autonomous Data Scientists
Justy and Cody dig into Meta’s Autodata and why better data, not just bigger models, is the pain point showing up everywhere right now. They unpack Agentic Self-Instruct, the four-agent setup, the weak-versus-strong solver idea, and why turning extra inference compute into better training data is a pretty interesting trade. They also get practical about who would adopt it, where the friction is, and a couple of concrete weekend experiments to try.
-
MAXS: Meta Adaptive Exploration with LLM Agents
MAXS introduces an innovative framework for improving the reasoning capabilities of LLM agents, addressing critical issues in multi-tool reasoning. The integration of lookahead strategies and trajectory convergence allows for more stable and efficient performance, making it highly relevant for developers and practitioners.
-
What Even Is a Parameter?
This episode explores the significance of parameters in large language models (LLMs), discussing their role in AI functionality and the implications for real-world applications. Hosts engage in a dialogue about how these parameters affect model behavior and the energy demands of training them, illustrating concepts with relatable analogies and examples.