"auto" assesses each query and adjusts the query planning and iterative search effort it receives, a faster, lower-cost path for simple questions and complex ones get multi-step retrieval, without manual tuning.
Let the query decide: Introducing "auto", a new retrieval reasoning effort mode in Foundry IQ that recovers 97.6% of the evidence recall of the highest effort level while using 46.9% fewer query planning tokens.
techcommunity.microsoft.com/…
1
46
More instinct codes folks if anyone is interested
app.instinct.com/invite?t=jm…
105
Why does every agent framework eventually grows into
• AGENTS.md
• Tool schemas
• Memories
• Skills
• Evals
• Structured outputs
I do not think that its accidental. Engineering has always been a conversion of inference into specification.
againstentropy.dev/essays/th…
1
117
Was working on an something and the models called out something back on me
Procrastination-as-infrastructure (n.)
1
81
4) Survives the productivity test
Unlike social media scrolling, designing a better knowledge system feels productive.
And that is why probably more destructive.
1
15
Coding agents revealed something I hadn't appreciated before. Search is no longer just a developer feature.
It's becoming a programming primitive. And winners might look very different from one owning the best retrieval models
Some thoughts here - againstentropy.dev/essays/se…
1
82
Lots of people kept asking me if SaaS is dead. Reading production postmortems pushed me toward a different conclusion. Software maybe isnt what companies eventually pay for.
They pay for operational confidence. That's what this essay explores
againstentropy.dev/essays/th…
49
Everything old is new again
nibblestew.blogspot.com/2026…
Great points...I pressure on memory would bring back the culture of packing more in less
87
I have always been amazed by magentism and gravity. Never been able to explore and explain them well untill now
againstentropy.pages.dev/exp…
try it out - some of the best times to learn stuff.
73

