its a good call but agent selector will replace it
@swyx
-
Tiny Project Updates Subscribe Development Progress
By
–
(subscribe for updates… this is still in development https://
latent.space/p/tiny) -
Tiny Teams as Co-op Multiplayer Model for AI Engineers
By
–
> "If the AI Engineer was the single player game, Tiny Teams are the co-op multiplayer game, capable of far more adaptability, resilience and “damage per second”." by far the top private feedback about AIEWF 2025 was that we hit gold with the Tiny Teams focus. Everyone wants to
-
Voice AGI interruption capabilities and conflict resolution mechanisms
By
–
sure but that would be insufficiently different to semantic vad. i think a voice AGI -would- interrupt you. and in an interrupt conflict read all signs – face, stutter, etc – to continue or interrupt the interrupt all theoretical until some mad lad actually does it. just a cost
-
Disappointment with Voice Mode: AGI Promise vs Reality
By
–
where was THIS voice mode we were promised
— swyx 🐣 (@swyx) 15 juillet 2025
surprising how much this made me feel the agi vs the… horribly quantized to within an inch of its life 4o that we gotpic.twitter.com/YaM5AdFpP0where was THIS voice mode we were promised surprising how much this made me feel the agi vs the… horribly quantized to within an inch of its life 4o that we got
-
Pareto Frontier Meme Impact on Frontier LLM Cost Metrics
By
–
one pleasant outcome of memeing #paretofrontier so hard that it became figure 1 for gemini 2.5 is that now every frontier LLM is being launched with an X axis that represents "cost". as reasoning model verbosity and tool use inefficiency can vary up to a factor of 650x, always
-
Six-Way Parlay: AI Startup Acquisitions Predictions
By
–
six way parlay. i want Mistral founders to Apple, Mistral team to Meta. Character remainco to Perplexity. Quora to Perplexity. Ideogram to Figma. Gamma to Notion. read it back
-
Clipped multiagents section from Latent Space podcast
By
–
clipped the multiagents section https://
x.com/latentspacepod
/status/1944846464120754529
… -
OpenAI Using Claude Suggests It’s Genuinely Superior
By
–
even openai using claude means the alternative explanation is probably more true: claude is just that damn good and it isnt just mass hysteria